Interesting article about Dell's manufacturing
Mark Weaver
Wasn't aware that all Dell desktops sold in the US are assembled here. And
that Dell laptops are actually built by Dell itself overseas--not contract
producers. Interesting. Interesting, too, what it takes to achieve the
level of efficiency necessary to keep making computers locally.
imout...@mac.com
I mean, there can't be more than a half-hour of labor into each unit.
ah, from the article:
"The labor costs of a PC are 'roughly 10 bucks,'"
ah,
"And I.B.M., which created the PC market in 1981"
...fuckers.
Snit
BKidnTvHq7K...@comcast.com on 12/20/04 11:19 AM:
> http://www.nytimes.com/2004/12/19/business/yourmoney/19dell.html?adxnnl=1&adxn
> nlx=1103564655-xWXb5LYieftAMQ4SIBxuLg
>
> Wasn't aware that all Dell desktops sold in the US are assembled here. And
> that Dell laptops are actually built by Dell itself overseas--not contract
> producers. Interesting. Interesting, too, what it takes to achieve the
> level of efficiency necessary to keep making computers locally.
My favorite quote:
-----
Research and development is one way Dell tamps down costs. The company
devotes 2 percent of its bottom line to this area, much less than its
rivals. Innovation inside Dell is instead more about how one produces,
packages and markets a product than it is about improvements in the product
itself. "We have some competitors who are spending 5 or 6 or 8 percent on
R&D," Mr. Rollins said, "but our financials suggest our R&D model is the
right model."
Others, however, wonder if those cost savings come with a long-term cost.
According to the Dell supplier quoted anonymously above, when Dell squeezes
the profit out of a market it also squeezes out everyone's ability to
innovate in any meaningful way.
-----
--
Picture of a tuna milkshake: http://snipurl.com/bh6q
Feel free to ask for the recipe.
StormDrain
imout...@mac.com wrote:
> I really wish Apple would build its computers here.
I've seen newer Apple boxes that say Assembled in USA. "Made in USA" is
tough honor to get in the computer industry. Close to 100% of all parts
have to be made in USA: US metal, US plastic, US made components (those
components from US parts) etc. They've been trying to water down "Made
in USA" for years but so far haven't succeeded, hopefully they won't.
> I mean, there can't be more than a half-hour of labor into each unit.
>
> ah, from the article:
>
> "The labor costs of a PC are 'roughly 10 bucks,'"
> ah,
>
> "And I.B.M., which created the PC market in 1981"
>
> ...fuckers.
--
SD
"...merely a preponderance of evidence."
John
> In article <1103590928.5...@f14g2000cwb.googlegroups.com>,
> imout...@mac.com wrote:
>
>> I really wish Apple would build its computers here.
>
> I've seen newer Apple boxes that say Assembled in USA. "Made in USA"
> is tough honor to get in the computer industry. Close to 100% of all
> parts have to be made in USA: US metal, US plastic, US made
> components (those components from US parts) etc. They've been trying
> to water down "Made in USA" for years but so far haven't succeeded,
> hopefully they won't.
>
How new? Like 5 years ago maybe. But ZERO are assembled in USA today.
imout...@mac.com
Company's manufacturing facility in Cork, Ireland, and by external
vendors in Fremont, California, Fullerton, California, Taiwan, Korea,
the Netherlands, the People's Republic of China, and the Czech
Republic."
Apple's 2004 10K.
StormDrain
"John" <nos...@nospam.com> wrote:
Actually very new. Working at the recycling center I saw some new G5
boxes come through with that label...OK, I'm now not sure if it was a G5
box but it was a NEW Mac, not an old one. I remember telling guy I was
working with "hey look at this..., assembled in USA", he wasn't
impressed. :)
Steve Mackay
> In article <1103590928.5...@f14g2000cwb.googlegroups.com>,
> imout...@mac.com wrote:
>
>> I really wish Apple would build its computers here.
>
> I've seen newer Apple boxes that say Assembled in USA.
My MDD G4 says assembled in USA.
> "Made in USA" is
> tough honor to get in the computer industry. Close to 100% of all parts
> have to be made in USA: US metal
Umm, US metal isn't really US metal anymore. The environmental whackos
have killed that off quite well. Most of our steel comes from Europe and
Switzerland. Not sure about aluminum, we don't use much of that. But I
know of no "premium" steel made in the USA. And China has been buying up
our scraps at record pace, and record prices right now.
>, US plastic
Most plastic these days is US plastic. I'm nearly 100% postive that the
Macs are being produced with US plastic. Especially if they are
Polycarbonate.
> US made components (those
> components from US parts) etc. They've been trying to water down "Made
> in USA" for years but so far haven't succeeded, hopefully they won't.
Well, my mother-in-law worked in an embroidery shop in the late
80s that would emroider a hat with a logo and cut out the made in China or
taiwan tag, and sew a Made in USA tag in there. So don't trust everything
you see.
Snit
pan.2004.12.21....@hotmail.com on 12/21/04 1:55 AM:
> On Mon, 20 Dec 2004 17:44:58 -0800, StormDrain wrote:
>
>> In article <1103590928.5...@f14g2000cwb.googlegroups.com>,
>> imout...@mac.com wrote:
>>
>>> I really wish Apple would build its computers here.
>>
>> I've seen newer Apple boxes that say Assembled in USA.
>
> My MDD G4 says assembled in USA.
Easy to get any Mac to say that... just run an AppleScript:
say "assembled in USA"
:)
> Well, my mother-in-law worked in an embroidery shop in the late
> 80s that would emroider a hat with a logo and cut out the made in China or
> taiwan tag, and sew a Made in USA tag in there.
Figures...
--
I am one of only .3% of people who have avoided becoming a statistic.
Lisa Horton
Steve Mackay wrote:
>
>
>
> Umm, US metal isn't really US metal anymore. The environmental whackos
> have killed that off quite well. Most of our steel comes from Europe and
> Switzerland. Not sure about aluminum, we don't use much of that. But I
> know of no "premium" steel made in the USA. And China has been buying up
> our scraps at record pace, and record prices right now.
>
Actually, the steel industry changed, the US steel companies didn't.
After a difficult period of "adjustment", new mini-mills are once again
making steel in the US, profitably.
Lisa
Mark Weaver
"Snit" <SN...@CABLE0NE.NET.INVALID> wrote in message news:BDECCAA2.17E88%>
> My favorite quote:
> -----
> Research and development is one way Dell tamps down costs. The company
> devotes 2 percent of its bottom line to this area, much less than its
> rivals. Innovation inside Dell is instead more about how one produces,
> packages and markets a product than it is about improvements in the
product
> itself.
Yes, exactly. But remember that it was a US company that invented video
tape recording equipment (Ampex, if memory serves). But they put their
equipment in the hands of roughly 0% of consumers. Figuring out how to make
something inexpensive enough that it can be afforded by everybody can be
important as the original invention.
Something similar is happening with personal computers now, but I think most
people with a US-centric or at least western-centric viewpoint don't really
see it. People here sniff at $300 computers and consider them junk, but to
a huge fraction of the globe, $300 is still a LOT of money. Over the next
decade, huge numbers of people who never had computers before are going to
be buying them--and those computers are going to be very low cost by our
standards. They may or may not be very low cost Dells--but virtually none
of them are going to be very low cost Macs (a contradiction in terms).
> "We have some competitors who are spending 5 or 6 or 8 percent on
> R&D," Mr. Rollins said, "but our financials suggest our R&D model is the
> right model."
>
Sounds about right to me. There's a lot of R&D in a Dell desktop, of
course, almost certainly making up more than 2 percent--but it's R&D done by
MS for the operating system and by all the component vendors for their chips
and drivers. How much should be spent in R&D on the computer itself? 2
percent doesn't sound unreasonable at all.
And, in fact, I definitely prefer a desktop with less rather than more
'custom' engineering. What I like about Dell desktops is that all the parts
are standard bits that can easily be swapped out. Beyond that, give me a
quiet-running box that I can open up and work on without tools and make sure
that all the selected parts play well together, pack it up, and send it out.
> Others, however, wonder if those cost savings come with a long-term cost.
> According to the Dell supplier quoted anonymously above, when Dell
squeezes
> the profit out of a market it also squeezes out everyone's ability to
> innovate in any meaningful way.
> -----
Now that's silly. If you're first to market with an innovative product, you
get to enjoy high margins--for a little while. Then everybody else catches
up and, if you're in the innovation/high-margin segment, you'd better be
ready with your next 'new new thing' by that time. That's how the game
works for all kinds of products, not just computers. Of course, over time,
some markets mature and it gets progressively harder to come up with
innovations that consumers are willing to pay a lot extra for (are there
still microwave oven manufacturers, for example, that can command
high-margins because of innovative features?). C'est la vie.
imout...@mac.com
> "Snit" <SN...@CABLE0NE.NET.INVALID> wrote in message
news:BDECCAA2.17E88%>
> > My favorite quote:
> > -----
> > Research and development is one way Dell tamps down costs. The
company
> > devotes 2 percent of its bottom line to this area, much less than
its
> > rivals. Innovation inside Dell is instead more about how one
produces,
> > packages and markets a product than it is about improvements in the
> product
> > itself.
>
> Yes, exactly. But remember that it was a US company that invented
video
> tape recording equipment (Ampex, if memory serves). But they put
their
> equipment in the hands of roughly 0% of consumers. Figuring out how
to make
> something inexpensive enough that it can be afforded by everybody can
be
> important as the original invention.
zzzzz. There's not much of a qualitative difference between Dell and
Apple's pricing, currently.
For most examples you may save a couple hundred, perhaps on better kit,
perhaps not, on Dell's offerings.
> Something similar is happening with personal computers now, but I
think most
> people with a US-centric or at least western-centric viewpoint don't
really
> see it. People here sniff at $300 computers and consider them junk,
but to
> a huge fraction of the globe, $300 is still a LOT of money.
They're better off buying last year's computers, really.
> Over the next
> decade, huge numbers of people who never had computers before are
going to
> be buying them--and those computers are going to be very low cost by
our
> standards. They may or may not be very low cost Dells--but virtually
none
> of them are going to be very low cost Macs (a contradiction in
terms).
these people are totally irrelevant to me as a developer. People too
poor for computers are not software purchasers.
> > "We have some competitors who are spending 5 or 6 or 8 percent on
> > R&D," Mr. Rollins said, "but our financials suggest our R&D model
is the
> > right model."
> >
>
> Sounds about right to me. There's a lot of R&D in a Dell desktop, of
> course, almost certainly making up more than 2 percent--but it's R&D
done by
> MS for the operating system and by all the component vendors for
their chips
> and drivers. How much should be spent in R&D on the computer itself?
2
> percent doesn't sound unreasonable at all.
Yes, it is true that the heavy lifting of Dell's R&D is being done by
Microsoft and Intel. Dell's original R&D is devoted to eg. printers and
laptops. If anything Dell is gold-bricking its R&D costs to make its
gross margins look better (a dollar saved from COGS is a dollar earned
as far as Wall Street is concerned).
> And, in fact, I definitely prefer a desktop with less rather than
more
> 'custom' engineering.
Depends. I prefer new and interesting uses, like when you could power
up the Mac II by pressing a key on the keyboard. That was fucking cool.
Plus like how Macs can be scheduled to power up and down automatically.
Now that sleep is so good I don't use this any more, but back in the
day I was using my Mac II as an alarm clock...
> What I like about Dell desktops is that all the parts
> are standard bits that can easily be swapped out.
Gratuitous incompatibility is to be avoided, of course. But I would
argue that the lion's share of Apple's R&D is going toward components
that aren't user-serviceable, like the G5's memory controller and
integrated motherboard features.
The current G5 is an example of standardization (1394, USB 2.0, PCI,
AGP, SATA, ATA, 802.11g, DVI) where it makes sense, but customization
where needed.
> Beyond that, give me a
> quiet-running box that I can open up and work on without tools and
make sure
> that all the selected parts play well together, pack it up, and send
it out.
Relying on Microsoft for R&D though is a two-edged sword. They haven't
been shipping much but bug-fixes for the past 3 years.
Sony's desktop offerings could be a lot cooler if they had more control
over the OS.
> > Others, however, wonder if those cost savings come with a long-term
cost.
> > According to the Dell supplier quoted anonymously above, when Dell
> squeezes
> > the profit out of a market it also squeezes out everyone's ability
to
> > innovate in any meaningful way.
> > -----
>
> Now that's silly.
I disagree -- profits fund R&D. If Dell is sucking all the profit out
of the x86 hardware market, R&D will be compromised. Plus the
stultifying monoculture that rises, like what IBM tried to impose in
the late 80s.
A triopoly of Dell/Microsoft/Intel would indeed be a sorry way to run a
railroad as far as innovation goes. Without Apple to rip off they have
been totally rudderless.
Snit
1103640997.4...@c13g2000cwb.googlegroups.com on 12/21/04 7:56 AM:
>>> My favorite quote:
>>> -----
>>> Research and development is one way Dell tamps down costs. The company
>>> devotes 2 percent of its bottom line to this area, much less than its
>>> rivals. Innovation inside Dell is instead more about how one produces,
>>> packages and markets a product than it is about improvements in the product
>>> itself.
>>>
>> Yes, exactly. But remember that it was a US company that invented video tape
>> recording equipment (Ampex, if memory serves). But they put their equipment
>> in the hands of roughly 0% of consumers. Figuring out how to make something
>> inexpensive enough that it can be afforded by everybody can be important as
>> the original invention.
>>
> zzzzz. There's not much of a qualitative difference between Dell and Apple's
> pricing, currently.
No, there is not, but Dell does serve the lower end where Apple does not.
This makes some not-so-clever folks think that Dell is less expensive.
>
> For most examples you may save a couple hundred, perhaps on better kit,
> perhaps not, on Dell's offerings.
The sad thing is the regulars in CSMA know full well that their expensive
Mac claims are just a myth... darn near every comparison site shows that
Macs are priced very competitively:
http://myweb.cableone.net/snit/csma/prices
Some of the... um, fine folks... in CSMA want us to believe that they have
some secret knowledge that is not known outside of CSMA, or that all other
sources, including PC Week, are Mac biased. They do not understand why I do
not roll over and agree with them on this one. They will, however, point to
sites where I can buy PC's. I suppose the correct response is to point them
to the Apple store. :)
--
Picture of a tuna soda: http://snipurl.com/bid1
Snit
>
> "Snit" <SN...@CABLE0NE.NET.INVALID> wrote in message news:BDECCAA2.17E88%>
>> My favorite quote:
>> -----
>> Research and development is one way Dell tamps down costs. The company
>> devotes 2 percent of its bottom line to this area, much less than its rivals.
>> Innovation inside Dell is instead more about how one produces, packages and
>> markets a product than it is about improvements in the product itself.
>
> Yes, exactly. But remember that it was a US company that invented video
> tape recording equipment (Ampex, if memory serves). But they put their
> equipment in the hands of roughly 0% of consumers. Figuring out how to make
> something inexpensive enough that it can be afforded by everybody can be
> important as the original invention.
No doubt... the problem Apple has (or at least one problem) is that they do
not have enough G5's to go around. If they had more, they would likely make
a headless iMac or something in the $550 to $600 range and make me happier.
A machine with the same specs as the iMac would be fine... would not even
have to be super-miniaturized, though I would be disappointed if it were not
stylish and practical. Rough price thought experiment:
Current 20 inch iMac: ~$1900
Current 20 inch monitor: ~$1300
Fantasy headless iMac: ~$ 650 to $700 (though the difference is only $600)
But that is for a comparable machine to the 20 inch iMac... which includes
all the software and other goodies I expect to come with the headless
iMac... but also includes a Superdrive, a 160 GB HD, and a 1.8 GHz G5.
So make a lower end model with a combo drive, 80 GB HD, and a 1.6 GHz G5,
and sell that for a hundred bucks less... which brings you to a low end
headless iMac running around $550 to $600 (the difference between the combo
drive and SuperDrive eMacs is $200, so potentially they could go even lower
and make a decent profit, but I will stick with the $100 estimate).
Apple would be able to make a profit on each of these... perhaps more than
the do on the iMac itself (remember, I added $50 to $100 on top of the price
difference between an iMac and a similar monitor - and then only made a $100
drop for the low end).
Since there is only a $100 difference, I would bet most would buy the higher
end, but Apple would have a very good machine for about what Dell does...
when you add all the goodies to the Dell that the iMac (and theoretical
headless iMac) have... iLife, NVIDIA GeForce FX 5200 Ultra, modem, FireWire
400, USB 2.0, optical digital sound out, S-Video, Composite video, speakers,
etc.
And, of course, by the time Apple came out with this, each of the above
numbers would be a bit higher... faster CPU, better video card, and bigger
HD... but my "fantasy" numbers come from the assumption the machine is
released today.
I am sure the folks at Apple have done this type of playing with the
numbers...
----------
Here, just for fun... let's make a similar "headless" Dell to the
theoretical headless iMac
Dimension 300 set up in much the same way: $774... *on sale*! The "regular"
price is $824. So let's just say $800 as a nice round number...
And that is for the low end headless iMac: myweb.cableone.net/snit/dell.pdf
So using prices fairly derived from Apple's site and comparing us to what
Dell actually is doing:
Apple: ~$600
Dell : ~$800
And the Dell has integrated video with shared memory and nothing that really
comes close to iLife! And it still is higher than the theoretical *high
end* headless iMac with the bigger hard drive and the SuperDrive! Heck,
Apple could just add $100 (about 20%!) to the my prices above and still be
*very* competitive! So the headless iMacs would be, say, $650 and $800.
Apple would make a tidy profit and they would likely sell *very* well.
Again: the problem - not enough G5's.
Oh, and while I am sure some of the PC apologists will nit-pick the software
I added to the Dell, the price difference is very clear - though keep in
mind this is a *hypothetical* Apple machine... but it is based on the actual
prices today.
--
"If a million people believe a foolish thing, it is still a foolish thing."
- Anatole France
Mark Weaver
<imout...@mac.com> wrote in message
> For most examples you may save a couple hundred, perhaps on better kit,
> perhaps not, on Dell's offerings.
>
> > Something similar is happening with personal computers now, but I
> think most
> > people with a US-centric or at least western-centric viewpoint don't
> really
> > see it. People here sniff at $300 computers and consider them junk,
> but to
> > a huge fraction of the globe, $300 is still a LOT of money.
>
> They're better off buying last year's computers, really.
>
Well, duh -- of course the low end models *do* use parts that aren't leading
edge (3-year-old video chips, etc) where the R&D has already been amortized.
But they're *new* old parts, not used boxes. You aren't going to supply the
Indian, Chinese, Brazilian markets, etc with hand-me-downs.
>
> these people are totally irrelevant to me as a developer. People too
> poor for computers are not software purchasers.
>
Funny -- MS doesn't see it that way. They can, for example, be bothered to
create a cheap Hindi-only version of their software. Even at a few dollars
a copy, developing markets have a huge potential payoff given that the
incremental cost of producing another copy of a piece of software is near
zero. And in time, it's likely that the PC boom will be primarily overseas
after the market has matured here.
>
> Relying on Microsoft for R&D though is a two-edged sword. They haven't
> been shipping much but bug-fixes for the past 3 years.
>
That doesn't seem to be hurting Dell's sales. Most areas of improvement
don't depend on OS upgrades.
> Sony's desktop offerings could be a lot cooler if they had more control
> over the OS.
>
Wouldn't touch a Sony desktop. Cooler & more customized == PITA.
>
> I disagree -- profits fund R&D. If Dell is sucking all the profit out
> of the x86 hardware market, R&D will be compromised. Plus the
> stultifying monoculture that rises, like what IBM tried to impose in
> the late 80s.
>
Nah, there are still high margins to be made in innovations in the PC market
(which is why, for example, ATI can charge so much for the greatest graphics
card).
> A triopoly of Dell/Microsoft/Intel would indeed be a sorry way to run a
> railroad as far as innovation goes. Without Apple to rip off they have
> been totally rudderless.
>
Right without Apple, nobody would've invented DVDR drives, LCD monitors,
faster chips, bigger hard disks, digital photography, MP3 players, etc, etc.
StormDrain
imout...@mac.com wrote:
I guess this means if they need something built in house the run across
the street to Ireland but if they need to outsource they have to go
cleaaarrrrr out to California.
imout...@mac.com
> <imout...@mac.com> wrote in message
>
> > For most examples you may save a couple hundred, perhaps on better
kit,
> > perhaps not, on Dell's offerings.
> >
> > > Something similar is happening with personal computers now, but I
> > think most
> > > people with a US-centric or at least western-centric viewpoint
don't
> > really
> > > see it. People here sniff at $300 computers and consider them
junk,
> > but to
> > > a huge fraction of the globe, $300 is still a LOT of money.
> >
> > They're better off buying last year's computers, really.
> >
>
> Well, duh -- of course the low end models *do* use parts that aren't
leading
> edge (3-year-old video chips, etc) where the R&D has already been
amortized.
> But they're *new* old parts, not used boxes. You aren't going to
supply the
> Indian, Chinese, Brazilian markets, etc with hand-me-downs.
Why not? It'd be the same pattern for cars, and computers are a lot
easier to ship than cars.
Shit, coming from nothing, a 400Mhz PII w/ Voodoo2 would be incredibly
cool.
> > these people are totally irrelevant to me as a developer. People
too
> > poor for computers are not software purchasers.
> >
>
> Funny -- MS doesn't see it that way. They can, for example, be
bothered to
> create a cheap Hindi-only version of their software.
MS is different, of course, since their job #1 is to prevent growth in
any segment that could build into a threat to their dominance. They'd
rather have people pirate their stuff than buy somebody else's product.
> Even at a few dollars
> a copy, developing markets have a huge potential payoff given that
the
> incremental cost of producing another copy of a piece of software is
near
> zero. And in time, it's likely that the PC boom will be primarily
overseas
> after the market has matured here.
This is true that India and China are untapped software markets, and
I'm sure some localizing and support company can give a successful
developer enough money to buy a very exotic car or something. Yet this
is just marginal income, and the rate of return may or may not exceed
investing the effort in the primary markets.
I guess what is necessary for a vibrant & interesting market community
to develop is a viable middle class. People in grinding poverty may
scrape together a few cents to buy my software but I don't see how they
are going to contribute much in return, while the upper class might buy
more software but again I don't see them contributing towards a
community of users forming.
> > Relying on Microsoft for R&D though is a two-edged sword. They
haven't
> > been shipping much but bug-fixes for the past 3 years.
> >
>
> That doesn't seem to be hurting Dell's sales.
Dell is taking sales from everyone so this may not be unexpected.
Plus the corporate market is different of course since they desire
bugfixes much more than new features.
> Most areas of improvement don't depend on OS upgrades.
Most/all areas of incremental improvement, yes. But where API is
involved OS upgrades are indeed necessary. Back in the 80s the PC toyed
with innovation with such companies as Hercules and AdLib producing
interesting stuff, yet without centralized support from Microsoft these
tech areas sorta wandered and drifted in a drunken walk toward
improvement.
> > Sony's desktop offerings could be a lot cooler if they had more
control
> > over the OS.
> >
>
> Wouldn't touch a Sony desktop. Cooler & more customized == PITA.
:) Sorta proves my point... With the Mac I get cooler and more
customized without the PITA.
> > I disagree -- profits fund R&D. If Dell is sucking all the profit
out
> > of the x86 hardware market, R&D will be compromised. Plus the
> > stultifying monoculture that rises, like what IBM tried to impose
in
> > the late 80s.
> >
>
> Nah, there are still high margins to be made in innovations in the PC
market
> (which is why, for example, ATI can charge so much for the greatest
graphics
> card).
Good example really. Real innovation (outside of incrementalism) on the
graphics card side is gated by Microsoft's control of the DirectX spec
(the vendors cooperate together to maintain OpenGL as an alternative
though).
Plus I think you are making a mistake of placing ATI and NVIDIA as
identical with the "PC Market".
They are leading OEM suppliers to it, of course, but there's not much
about their operations that are tied into PCs per se.
The core of the PC market are the PC box builders starting from Dell
and moving down. I see little to no innovation from the majors, as the
years go by.
Small upstarts like Alienware and Shuttle produce interesting kit that
eventually influences the mainstream, but it's been about 10 years
since a product from Dell, Gateway, HP, Compaq, or IBM turned my head.
When I was in Japan in the early 1990s, prior to the Great Win95
consolidation, things were different. Fujitsu and Sharp were doing
interesting (proprietary) stuff, forcing NEC to respond. The success of
Win9x in Japan stopped this innovation cold.
> > A triopoly of Dell/Microsoft/Intel would indeed be a sorry way to
run a
> > railroad as far as innovation goes. Without Apple to rip off they
have
> > been totally rudderless.
> >
>
> Right without Apple, nobody would've invented DVDR drives, LCD
monitors,
> faster chips, bigger hard disks, digital photography, MP3 players,
etc, etc.
Apple was terribly behind the curve with CDR/DVDR (going with DVD-RAM
for some reason). I'll grant you that.
Faster chips and bigger hard disks are not major innovations per se.
Incrementalism is not meaningful innovation.
The fact remains that Microsoft is busy copying 15 years of work from
NeXT/Apple to ship as Longhorn, not to mention recently succeeded in a
near pixel-perfect copy of its iTMS.
Intel has been trying for about a decade to strip out legacy crap like
Super I/O from the x86 spec...
x86 is good at easy incrementalism but piss-poor at actual evolution
that requires short-term pain, like eg. moving to USB from serial ports
or even PnP PCI from ISA.
If x86 were to disappear there's really nothing from its sorry history
or future direction that indicates I as a primary Mac user would miss
out on anything.
(modulo graphics card advances, tho. Apple in the mid-90s had some
half-assed ideas but it was Matrox and 3Dfx who initially led the
charge to where we are now; then again perhaps SGI should get the
credit for this direction).
Mark Weaver
<imout...@mac.com> wrote in message
>
> This is true that India and China are untapped software markets, and
> I'm sure some localizing and support company can give a successful
> developer enough money to buy a very exotic car or something. Yet this
> is just marginal income, and the rate of return may or may not exceed
> investing the effort in the primary markets.
>
> I guess what is necessary for a vibrant & interesting market community
> to develop is a viable middle class. People in grinding poverty may
> scrape together a few cents to buy my software but I don't see how they
> are going to contribute much in return, while the upper class might buy
> more software but again I don't see them contributing towards a
> community of users forming.
>
What constitutes 'middle class' in developing countries is quite different
than here. middle class Mexicans, for example, may have one small car,
certainly have a TV and VCR or DVD player and probably already use the
Internet (but not necessarily at home--I've seen the oddest Internet 'cafes'
down there -- coin laundry & Internet, excercize gym & Internet, etc). It
appears there are a lot of people down there who'd jump on a $200-$300
computer but not a $1000+ computer.
> >
> > That doesn't seem to be hurting Dell's sales.
>
> Dell is taking sales from everyone so this may not be unexpected.
>
Dell isn't just stealing sales -- the overall PC market is growing (faster
than Apple's as it happens). Growth doesn't depend on innovation
alone--increases in bang (and bang for the buck) can drive growth.
>
> Most/all areas of incremental improvement, yes. But where API is
> involved OS upgrades are indeed necessary.
Well, MS does provide incremental upgrades where necessary to support new
hardware. And there is the 'Media Center' edition which has just been
upgraded. It's basically just XP Pro with TV/PVR extensions, but those have
been actively upgraded.
> >
> > Wouldn't touch a Sony desktop. Cooler & more customized == PITA.
>
> :) Sorta proves my point... With the Mac I get cooler and more
> customized without the PITA.
>
Uh, no, by PITA I mean custom case, & bits customized to fit in the case so
that if you want to repair or upgrade you have to locate & buy higher priced
specialized parts. Or peripherals have to be external and clutter the desk.
That kind of thing. Want a TV tuner / PVR card in your PC? No problem.
Want one in your iMac? AFAIK, the best you can do is to plug in a USB
device (PITA).
>
> Plus I think you are making a mistake of placing ATI and NVIDIA as
> identical with the "PC Market".
>
> They are leading OEM suppliers to it, of course, but there's not much
> about their operations that are tied into PCs per se.
>
No, just almost all their revenues and profits (though PS/2 and XBox are
also important).
>
> Faster chips and bigger hard disks are not major innovations per se.
> Incrementalism is not meaningful innovation.
>
It's meaningful to me. Increases in size and speed may seem dull, but
without them most of the new recent functions of PCs would not be possible
(e.g. stitching multiple 8MP images into a 50 MP panorama, dumping multiple
DVD images to the hard disk, storing an entire CD collection in MP3 form).
Ditto decreases in size and cost--hard disk MP3 players wouldn't exist if
HDs hadn't gotten tiny & cheap.
To be honest that stuff is a whole lot more meaningful to me than, say, a
new OS rendering engine.
Snit
>
> <imout...@mac.com> wrote in message
>>
>> This is true that India and China are untapped software markets, and
>> I'm sure some localizing and support company can give a successful
>> developer enough money to buy a very exotic car or something. Yet this
>> is just marginal income, and the rate of return may or may not exceed
>> investing the effort in the primary markets.
>>
>> I guess what is necessary for a vibrant & interesting market community
>> to develop is a viable middle class. People in grinding poverty may
>> scrape together a few cents to buy my software but I don't see how they
>> are going to contribute much in return, while the upper class might buy
>> more software but again I don't see them contributing towards a
>> community of users forming.
>>
>
> What constitutes 'middle class' in developing countries is quite different
> than here. middle class Mexicans, for example, may have one small car,
> certainly have a TV and VCR or DVD player and probably already use the
> Internet (but not necessarily at home--I've seen the oddest Internet 'cafes'
> down there -- coin laundry & Internet, excercize gym & Internet, etc). It
> appears there are a lot of people down there who'd jump on a $200-$300
> computer but not a $1000+ computer.
I doubt a cheap computer would be durable enough to take that abuse. :)
--
If A = B and B = C, then A = C, except where void or prohibited by law.
Roy Santoro, Psycho Proverb Zone (http://snipurl.com/BurdenOfProof)
imout...@mac.com
> > > That doesn't seem to be hurting Dell's sales.
> >
> > Dell is taking sales from everyone so this may not be unexpected.
> >
>
> Dell isn't just stealing sales -- the overall PC market is growing
(faster
> than Apple's as it happens). Growth doesn't depend on innovation
> alone--increases in bang (and bang for the buck) can drive growth.
All of this growth is totally irrelevant to me. My market already owned
personal computers in 1985.
What interests me as a user and developer is purely advancement in the
state of the art.
> > Most/all areas of incremental improvement, yes. But where API is
> > involved OS upgrades are indeed necessary.
>
> Well, MS does provide incremental upgrades where necessary to support
new
> hardware. And there is the 'Media Center' edition which has just
been
> upgraded. It's basically just XP Pro with TV/PVR extensions, but
those have
> been actively upgraded.
Yes, tablet/media center is one area where Microsoft is innovating.
Whether or not this innovation pans out into something useful to me
remains to be seen though.
> > > Wouldn't touch a Sony desktop. Cooler & more customized == PITA.
> >
> > :) Sorta proves my point... With the Mac I get cooler and more
> > customized without the PITA.
> >
>
> Uh, no, by PITA I mean custom case, & bits customized to fit in the
case so
> that if you want to repair or upgrade you have to locate & buy higher
priced
> specialized parts.
Sure, gratuitous incompatibility is never a good thing. My point was
there's a bigger payoff for the incompatibility with Apple than with
Sony.
> Or peripherals have to be external and clutter the desk.
Heaven forfend! I actually prefer external peripherals since they are
more easily transferable and repurposable than cards.
> That kind of thing. Want a TV tuner / PVR card in your PC? No
problem.
> Want one in your iMac? AFAIK, the best you can do is to plug in a
USB
> device (PITA).
I see peripherals as an analogue to the UNIX commandline way of
thinking; many small independent bits working chained together rather
than one top-down shrinkwrapped integrated solution.
> > Plus I think you are making a mistake of placing ATI and NVIDIA as
> > identical with the "PC Market".
> >
> > They are leading OEM suppliers to it, of course, but there's not
much
> > about their operations that are tied into PCs per se.
> >
>
> No, just almost all their revenues and profits (though PS/2 and XBox
are
> also important).
Sure, the PC business of course supports their R&D. x86 being the
mainstream and all. My argument is that NVIDIA and ATI exist apart from
the PC space though. If the Wintel/x86 platform vanished tonight they
would not be SOL, outside of having to figure out a replacement for
AGP/PCI Express.
> > Faster chips and bigger hard disks are not major innovations per
se.
> > Incrementalism is not meaningful innovation.
> >
>
> It's meaningful to me. Increases in size and speed may seem dull,
but
> without them most of the new recent functions of PCs would not be
possible
> (e.g. stitching tiple 8MP images into a 50 MP panorama, dumping
multiple
> DVD images to the hard disk, storing an entire CD collection in MP3
form).
> Ditto decreases in size and cost--hard disk MP3 players wouldn't
exist if
> HDs hadn't gotten tiny & cheap.
My argument is that there are two kinds of innovation, I* and I+. I*
innovation is qualitative, significant like when Apple does something
interesting. I+ is 'faster chips and bigger hard drives'. You are
correct that x86's iterative I+ innovation will add up to I* over time,
but really since about the introduction of the 440BX platform in 1998
we've seen a very small I+ value for innovation in the Wintel world.
Same crappy BIOS setup UI. Same Apple II-ripoff boot chirp, for god's
sake. Same horrific window damage repair behavior of Windows.
> To be honest that stuff is a whole lot more meaningful to me than,
say, a
> new OS rendering engine.
Depends on what you use your PC for, I guess. I like qualitative
improvements to my user experience.
Faster/smaller/better is useless in isolation, especially in the short
term. My PC buying experience is basically skipping a generation or
two; my current 2.6Ghz box should remain current for several years
more.
This is an age old debate though, and is why you are primarily an x86
user and I am primarily a Mac user. But I'm not a pure Mac partisan,
I'd love to see other computer companies step into the breach and
extend the state of the art in meaningful ways, like eg. Sony grabbing
the ring with the PS3 and shipping a BeOS-like desktop developer
environment. That would rock my world.
Wintel sucks though. Always has and seemingly always will.
Mark Weaver
<imout...@mac.com> wrote in message
>
> What interests me as a user and developer is purely advancement in the
> state of the art.
>
And what I'm saying is that advancement in the state-of-the-art in recent
years has had not very much to do with changes in the OS and computer itself
and much more to do with the computer's role as a network node and digital
media hub. And I include Apple in that. Apple has been rev'ing its OS more
quickly, but that's just doing the same things slightly more slickly. Yawn.
I get far more bang for my buck, when, say I upgrade from a 3MP/3X zoom
camera to an 8MP/7x zoom camera than I would with any marginal usability
improvements (or collection of improvements) in the OS.
>
> Yes, tablet/media center is one area where Microsoft is innovating.
> Whether or not this innovation pans out into something useful to me
> remains to be seen though.
>
I'm not running MCE (though I may install it on a separate partition at some
point to check it out), but I'm starting to believe the whole 'convergence'
idea finally makes sense. There is a computer in the living room that now
plays that role--line out runs to the stereo--MP3s are played that way, it
has a TV/PVR card, s-video connection to the TV and we also connect a
portable DLP projector to watch DVDs (sound through the stereo) or live TV
(via the TV card and stereo). And that PC also has the digital photo /
video collection which can be displayed with the DLP projector (only XGA,
but even so digital photos look great projected onto a 120" screen). And
once in a while we let the kids use the projector for games. A collection
of consumer electronics that could do some, but not all of that (and not as
inexpensively or flexibly as adding the functionality to the computer).
All this stuff seems to me like it *should* be right up Apple's alley with
their 'iLife' orientation. So it strikes me as very strange (OK, stupid)
that they're not really doing it.
>
> > Or peripherals have to be external and clutter the desk.
>
> Heaven forfend!
Forfend? Who are you, Montgomery Burns? George Will? Do you wear a bow
tie ;)
> I actually prefer external peripherals since they are
> more easily transferable and repurposable than cards.
>
Well, fine, but I want a computer where I have a *choice* between internal
and external. For a TV-tuner/PVR that's not going to be removed from the
machine for a long time, if ever, I'd prefer not have the extra wires &
clutter (not to mention, the price-performance of the external devices is
not as good and selection is much more limited).
> My argument is that there are two kinds of innovation, I* and I+. I*
> innovation is qualitative, significant like when Apple does something
> interesting. I+ is 'faster chips and bigger hard drives'. You are
> correct that x86's iterative I+ innovation will add up to I* over time,
> but really since about the introduction of the 440BX platform in 1998
> we've seen a very small I+ value for innovation in the Wintel world.
> Same crappy BIOS setup UI. Same Apple II-ripoff boot chirp, for god's
> sake. Same horrific window damage repair behavior of Windows.
>
Complaining about the BIOS setup and 'boot chirp' to me is like saying, "I
can't believe we're still starting cars with keys after all these
years--little funny-shaped bits of metal--how archaic". Well, I suppose it
is, but who really cares?
As for damage repair, it has improved, but it certainly could be better.
Even so, it's a relatively trivial problem compared to spyware/adware.
> > To be honest that stuff is a whole lot more meaningful to me than,
> say, a
> > new OS rendering engine.
>
> Depends on what you use your PC for, I guess. I like qualitative
> improvements to my user experience.
>
I don't dislike 'qualitative improvements', it's just they're a whole lot
less appealing than new classes of functionality. I wouldn't spend much on
them. If MS were rev'ing XP Apple has with OS X was similarly charging $100
a pop, I wouldn't fork over the money. $69 for a PVR card is a whole lot
more compelling than $100 for marginal 'qualitative user experience
improvements'. To me anyway. And to most people, I'd say.
> Faster/smaller/better is useless in isolation, especially in the short
> term. My PC buying experience is basically skipping a generation or
> two; my current 2.6Ghz box should remain current for several years
> more.
>
Exactly. And much slower machines than that remain perfectly usable for
most things--my daughter has a several-year-old, hand-me-down Athlon 750 in
her room. It runs XP just fine and easily does everything she wants to do
with it--I haven't heard her yet asking about a new one. She wants a lot of
things (some of them expensive and electronic), but a new computer isn't
really even on the list.
> This is an age old debate though, and is why you are primarily an x86
> user and I am primarily a Mac user. But I'm not a pure Mac partisan,
> I'd love to see other computer companies step into the breach and
> extend the state of the art in meaningful ways, like eg. Sony grabbing
> the ring with the PS3 and shipping a BeOS-like desktop developer
> environment. That would rock my world.
Sony, man...they can't even do a hard disk MP3 player right even after
coming to the market very late and having the advantage of seeing everybody
else's efforts. Creating an open platform of any kind really isn't in their
genes.
Walter Bushell
Snit <SN...@CABLE0NE.NET.INVALID> wrote:
Ah, yes. Dell the Wallmart of manufacturing.
--
Guns don't kill people; automobiles kill people.
Walter Bushell
"Mark Weaver" <wea...@nospam-corvusdev.com> wrote:
<snip>
> Now that's silly. If you're first to market with an innovative product, you
> get to enjoy high margins--for a little while. Then everybody else catches
> up and, if you're in the innovation/high-margin segment, you'd better be
> ready with your next 'new new thing' by that time. That's how the game
> works for all kinds of products, not just computers. Of course, over time,
> some markets mature and it gets progressively harder to come up with
> innovations that consumers are willing to pay a lot extra for (are there
> still microwave oven manufacturers, for example, that can command
> high-margins because of innovative features?). C'est la vie.
And microwaves were HIGH technology just a split century ago. It's a
market Apple should look into. ;)
Walter Bushell
imout...@mac.com wrote:
<snip>
> Shit, coming from nothing, a 400Mhz PII w/ Voodoo2 would be incredibly
> cool.
Good enough for all business purposes, especially where labor is cheap.
Stock it with period software or Linux and free software and you are
good to go.
I remember G3 Macs going for $50 or so, with a 30 day replacement
policy, so essentially free.
imout...@mac.com
> <imout...@mac.com> wrote in message
> >
> > What interests me as a user and developer is purely advancement in
the
> > state of the art.
> >
>
> And what I'm saying is that advancement in the state-of-the-art in
recent
> years has had not very much to do with changes in the OS and computer
itself
> and much more to do with the computer's role as a network node and
digital
> media hub.
Yeah, for the past 5+ years we've seen the PC kinda kicking around,
looking for a new reason for people to upgrade their existing kit after
the web access wave crested.
If had the choice between the dual-core 4Ghz x86 PCs of 2006 or the
Longhorn CR, I'd take the Longhorn CR (as I fully expect 2.6Ghz + 256MB
AGP 8x will be more than sufficient for Longhorn).
My 2.6Ghz PC is fast enough for what I use it for. But if I were more
heavily into more productive uses (ie full-day Maya or Photoshop usage)
I would probably take the faster box instead of slicker OS, but we're
talking PC use not workstation use here.
But computers still suck, and it's not about a lack of external
connectivity. For an XP user, moving back to Win2k or even Win98
wouldn't be much of a loss wrt usability (perhaps none, the first thing
I do with an XP install is turn off the fisher-price UI), but as a Mac
user moving back from 10.3.x to 9 or 8.6 would be horrible, with a lot
of usability advances lost.
> And I include Apple in that. Apple has been rev'ing its OS more
> quickly, but that's just doing the same things slightly more slickly.
Yawn.
> I get far more bang for my buck, when, say I upgrade from a 3MP/3X
zoom
> camera to an 8MP/7x zoom camera than I would with any marginal
usability
> improvements (or collection of improvements) in the OS.
These 'marginal improvements' rock. I serve mp3s wirelessly from my g4
cube that sleeps rather quietly between uses. The OS of 10.3.x is so
much more capable than the OS of 1999.
Part of the problem for you is that the /Windows/ OS of 2004 hasn't
improved much over that of the NT5 betas of 1999. In fact I would say
it has /devolved/ in usability and esthetics over this time.
The OS and associated API has an immense effect on the utility of the
machine. Your previous points about increasing hardware performance and
capacities are well-taken, but for the vast majority of what I use a PC
for the basics of CPU performance, HD capacity, and memory capacity are
less important than nuts-and-bolts usability and the overall ergonomics
of the GUI and ease-of-programmability of the API and associated tools
(apache, sql, php, etc).
Steve's take is that the PC itself belongs in the den not in the living
room. With home networking infrastructure I agree with this. PCs can
and will take over the home, but I think more as a centralized media
server than a downsized media component.
Apple *has* taken the first steps with their Airport Express 802.11g ->
stereo converter.
UWB looks to be very interesting in this regard for serving HD video,
which is too much for anything short of 802.11a at the moment.
> Well, fine, but I want a computer where I have a *choice* between
internal
> and external. For a TV-tuner/PVR that's not going to be removed from
the
> machine for a long time, if ever, I'd prefer not have the extra wires
&
> clutter (not to mention, the price-performance of the external
devices is
> not as good and selection is much more limited).
I don't care really since the PCs are in a work cabinet and not
something I really have to look at when using.
I don't know much about the price/performance thing. I buy quality
stuff with I expect to be 10 year + useful life. These expenses aren't
that frequent for me to really look for saving $10 or whatever, I go
for utility.
Plus I just have a general distrust of pre-packaged solutions. They
just tend to suck for some reason.
> > My argument is that there are two kinds of innovation, I* and I+.
I*
> > innovation is qualitative, significant like when Apple does
something
> > interesting. I+ is 'faster chips and bigger hard drives'. You are
> > correct that x86's iterative I+ innovation will add up to I* over
time,
> > but really since about the introduction of the 440BX platform in
1998
> > we've seen a very small I+ value for innovation in the Wintel
world.
> > Same crappy BIOS setup UI. Same Apple II-ripoff boot chirp, for
god's
> > sake. Same horrific window damage repair behavior of Windows.
> >
>
> Complaining about the BIOS setup and 'boot chirp' to me is like
saying, "I
> can't believe we're still starting cars with keys after all these
> years--little funny-shaped bits of metal--how archaic". Well, I
suppose it
> is, but who really cares?
Good example. Once you get a car with bluetooth ignition unlock (like
the Toyoto Prius) you go, 'damn! what a good idea!'. Similar to remote
door unlocks. It's a minor thing, but life is full of minor things, and
I appreciate companies willing to tackle the small-stuff, attention to
detail.
> > > To be honest that stuff is a whole lot more meaningful to me
than,
> > say, a
> > > new OS rendering engine.
> >
> > Depends on what you use your PC for, I guess. I like qualitative
> > improvements to my user experience.
> >
>
> I don't dislike 'qualitative improvements', it's just they're a whole
lot
> less appealing than new classes of functionality.
That's my definition of qualitative vs. quantitative improvement. An
extra 200Mhz, or hell, even going 64-bit isn't really adding many new
classes of functionality. These come when a fundamental breakthrough
happens, like the GUI, a fully-realized high-res print architecture
(Adobe's Postscript -> Apple Talk -> Apple's rasterizer -> Canon's
laser engine), external serial bus, highspeed hassle-free external
expansion (firewire), implementing 802.11, etc.
The next frontier I see for PC's is high-res display. 80DPI is NOT GOOD
ENOUGH! 180DPI will be really really cool and offer a qualitatively
different experience. Note that this advance is gated by both the
necessary incremental technology advances AND OS/API improvements to
correctly handle the higher DPI.
> I wouldn't spend much on
> them. If MS were rev'ing XP Apple has with OS X was similarly
charging $100
> a pop, I wouldn't fork over the money. $69 for a PVR card is a whole
lot
> more compelling than $100 for marginal 'qualitative user experience
> improvements'. To me anyway. And to most people, I'd say.
Sure. 95% of the market apparently agrees with you, just like 95% of
everyone eats fast-food crap every day for lunch and/or dinner. I have
no problem with that. Your loss, not mine.
I also agree that 10.x's improvements are necessarily worth the $100/yr
upgrade price, it can certainly be argued that 10.2 is what 10.0 should
have been, and 10.3 was more of a revenue spin and further bugfix
effort over true qualitative/functional improvements.
The thing is, Apple *is* spending ~$200M/yr on OS improvements. I think
$50/yr is a much more realistic pricepoint, then again Apple is free to
try to charge what the market will bear I guess.
> > Faster/smaller/better is useless in isolation, especially in the
short
> > term. My PC buying experience is basically skipping a generation or
> > two; my current 2.6Ghz box should remain current for several years
> > more.
> >
>
> Exactly. And much slower machines than that remain perfectly usable
for
> most things--my daughter has a several-year-old, hand-me-down Athlon
750 in
> her room. It runs XP just fine and easily does everything she wants
to do
> with it--I haven't heard her yet asking about a new one. She wants a
lot of
> things (some of them expensive and electronic), but a new computer
isn't
> really even on the list.
IMV, though, that's (nearly) the sum total of what Wintel has brought
to the PC party. Vanilla incrementalism on Apple's fundamental
advances, both in hardware and software.
> > This is an age old debate though, and is why you are primarily an
x86
> > user and I am primarily a Mac user. But I'm not a pure Mac
partisan,
> > I'd love to see other computer companies step into the breach and
> > extend the state of the art in meaningful ways, like eg. Sony
grabbing
> > the ring with the PS3 and shipping a BeOS-like desktop developer
> > environment. That would rock my world.
>
> Sony, man...they can't even do a hard disk MP3 player right even
after
> coming to the market very late and having the advantage of seeing
everybody
> else's efforts. Creating an open platform of any kind really isn't
in their
> genes.
yeah, I know. We all have our dreams :) My dream is
Sony/Apple/IBM/Sun/SGI hookup and take over the world. Too bad Sun sold
out to MS tho, and IBM hates Sun's guts...
Mark Weaver
<imout...@mac.com> wrote in message
news:1103774537.9...@c13g2000cwb.googlegroups.com...
>
> Yeah, for the past 5+ years we've seen the PC kinda kicking around,
> looking for a new reason for people to upgrade their existing kit after
> the web access wave crested.
>
> If had the choice between the dual-core 4Ghz x86 PCs of 2006 or the
> Longhorn CR, I'd take the Longhorn CR (as I fully expect 2.6Ghz + 256MB
> AGP 8x will be more than sufficient for Longhorn).
>
But what do you think Longhorn is going to *do* for you that you can't do
now? Or, alternately, what do you do now that you think will become more
productive with Longhorn? Or is it just a question of the aesthetic
pleasure that a new interface will provide?
> But computers still suck, and it's not about a lack of external
> connectivity. For an XP user, moving back to Win2k or even Win98
> wouldn't be much of a loss wrt usability (perhaps none, the first thing
> I do with an XP install is turn off the fisher-price UI), but as a Mac
> user moving back from 10.3.x to 9 or 8.6 would be horrible, with a lot
> of usability advances lost.
>
Yes, well, that's because 9 *really* sucked. No pre-emptive multi-tasking,
no real threading, no real protected memory -- under the hood it was much
more primitive than Windows 95 (and even than Windows 3.x in some ways).
It's amazing that Apple managed to get as much performance and unsability
out of such an archaic foundation as they did.
>
> These 'marginal improvements' rock. I serve mp3s wirelessly from my g4
> cube that sleeps rather quietly between uses. The OS of 10.3.x is so
> much more capable than the OS of 1999.
>
Yeah, I did that for a while until I bit the bullet and ran the wire (I
needed to run video cable anyway, so did audio as well). No OS advancements
were required for either the wired or wireless approaches, though, and both
worked fine. But wireless was not as simple as line-out to line-in.
>
> Steve's take is that the PC itself belongs in the den not in the living
> room. With home networking infrastructure I agree with this. PCs can
> and will take over the home, but I think more as a centralized media
> server than a downsized media component.
>
> Apple *has* taken the first steps with their Airport Express 802.11g ->
> stereo converter.
>
What Steve is missing is that there's no *THE* PC. I think that's a pretty
basic mistake. How many computers are there in the house here? Off the top
of my head, I'm not even sure--one each in the kids bedrooms, the wife and I
each have a notebook, one in the living room, and two desktops with a KVM in
the home office where I work. So seven I guess. I'm not saying that's
typical, but more than one is certainly typical and the idea of putting one
expensive server machine in the den and then trying to wirelessly distribute
the media content to another room where the stereo and TV are is goofy.
Hell, I don't even bother to share MP3s across the network--I just
replicate. Everything I own is about 15G, and that's what--about $7 worth
of hard disk space per machine?
The 'central server' notion is based on an obsolete notion of one expensive
machine with an expensive hard disk per home. The homes that have just one
are not going to go for wireless connectivity to the TV. And the homes that
have more of less state-of-the-art toys (e.g. the target market) aren't
going to have just one computer.
> UWB looks to be very interesting in this regard for serving HD video,
> which is too much for anything short of 802.11a at the moment.
>
Plain old copper wires look a whole lot more interesting to me. Seriously,
why make it more complex?
>
> I don't know much about the price/performance thing. I buy quality
> stuff with I expect to be 10 year + useful life. These expenses aren't
> that frequent for me to really look for saving $10 or whatever, I go
> for utility.
>
A 10 year useful life for electronics? Right. What were digital cameras
like 10 years ago?. Digital music players? Wireless networking? Photo
printers? Computers? DVD players? Cell phones? It doesn't matter if they
*last* 10 years--you'll stop wanting to use them long before then. I
actually *do* have some 10 year old equipment in some of those categories
around--stuff that actually works as well now as the day I bought it, but
it's so obsolete there's no reason power it up. For example, I have a
Pentium 133 notebook in a cabinet somewhere. Works perfectly--if you don't
mind SVGA resolution (hey, at least it's not passive matrix), no integrated
optical drive, and a 700 meg hard disk.
>
> That's my definition of qualitative vs. quantitative improvement. An
> extra 200Mhz, or hell, even going 64-bit isn't really adding many new
> classes of functionality.
Depends--an extra 200Mhz is a lot if you're running at 100Mhz to start with,
but it's trivial if it represent a 10% increase. I generally don't upgrade
until I get a doubling in performance/capacity etc.
>
> The next frontier I see for PC's is high-res display. 80DPI is NOT GOOD
> ENOUGH! 180DPI will be really really cool and offer a qualitatively
> different experience.
My 15.4" notebook runs at 1680x1050. That's close to 150dpi. I could have
had 1900x1200 in the same size screen, which would have been over 150dpi.
It's nice--I like it. Text is noticeably sharper. But qualitatively
different? Nah, not really.
>
> > I wouldn't spend much on
> > them. If MS were rev'ing XP Apple has with OS X was similarly
> charging $100
> > a pop, I wouldn't fork over the money. $69 for a PVR card is a whole
> lot
> > more compelling than $100 for marginal 'qualitative user experience
> > improvements'. To me anyway. And to most people, I'd say.
>
> Sure. 95% of the market apparently agrees with you, just like 95% of
> everyone eats fast-food crap every day for lunch and/or dinner. I have
> no problem with that. Your loss, not mine.
>
No, my gain. I don't have an unlimited budget for electronic doodads--and
right now, money spent on other things (peripherals, projectors, digital
cameras, MP3 players, etc, etc) is simply more compelling than OS
upgrades--even hypothetical ones that aren't ready yet. Can't quite see how
I'm poorer for, say, having spent the money to be able to project a football
game or a movie on a 120" screen than having bought multiple $100 OS
upgrades for the various computers in the house. And I don't expect that to
change. When Longhorn is released in a couple years, an upgrade to a 1080i
native projector will probably look better than OS upgrades for our
computers.
>
> The thing is, Apple *is* spending ~$200M/yr on OS improvements. I think
> $50/yr is a much more realistic pricepoint, then again Apple is free to
> try to charge what the market will bear I guess.
>
It's not a question of a realistic price, really. Revenue has to pay for OS
R&D. It's not that big a target market and not nearly everybody in it will
pay for an upgrade. If selling for $50 brought in more than selling for
$100 (if they sold more than twice as many copies), then they'd do that.
> >
> > Sony, man...they can't even do a hard disk MP3 player right even
> > after coming to the market very late and having the advantage of
> > seeing everybody else's efforts. Creating an open platform of any
> > kind really isn't in their genes.
>
> yeah, I know. We all have our dreams :) My dream is
> Sony/Apple/IBM/Sun/SGI hookup and take over the world. Too bad Sun sold
> out to MS tho, and IBM hates Sun's guts...
>
And if they all hooked up, then what? What is it you think your ideal team
of players is going to do that isn't happening as it is?
Tim Smith
"Mark Weaver" <wea...@nospam-corvusdev.com> wrote:
> A 10 year useful life for electronics? Right. What were digital cameras
> like 10 years ago?. Digital music players? Wireless networking? Photo
> printers? Computers? DVD players? Cell phones? It doesn't matter if they
> *last* 10 years--you'll stop wanting to use them long before then. I
> actually *do* have some 10 year old equipment in some of those categories
> around--stuff that actually works as well now as the day I bought it, but
> it's so obsolete there's no reason power it up. For example, I have a
> Pentium 133 notebook in a cabinet somewhere. Works perfectly--if you don't
> mind SVGA resolution (hey, at least it's not passive matrix), no integrated
> optical drive, and a 700 meg hard disk.
Interesting question...I'm trying to think of what things I had 10 years
ago that I'd still find acceptable to use today, if they were still
working.
Only a handful of electronic items come to mind.
1. Televisions. I had a Mitsubishi rear-projection TV then. When it
died last year and I looked at new sets, the CRT-based rear-projection
sets of today were certainly better, but not so much better that I would
have replaced the Mitsubishi if I didn't have to. (I ended up not
replacing it with CRT-based rear-projection, but rather with a Samsung
DLP).
2. Calculators. I'll take my HP48SX over pretty much any of today's
machines.
3. CD players. I think my home CD player is over 10 years old. I stick
discs in and hit play and it works. I don't know if modern players have
better features, and don't care, since all I use is "play" and
occasionally "skip track". Maybe someday if SACD gets widespread enough
I'll replace it.
4. Game systems. I'm not going to throw out my SNES. It still works,
and "Super Mario World" and "The Legend of Zelda: A Link to the Past"
and "Super Mario Kart" are still great games. Someday, I'll want to
replay them again.
--
--Tim Smith
ed
Tim Smith <reply_i...@mouse-potato.com> typed:
- i've got a 10+ year old laserjet that still works great.
- a 10 year old bookshelf stereo system that i still use
- a ~13 year old cd boom box that gets use when i'm working in the garage.
- a 10 year old digital answering machine
- a ~8 yr old hp48g calculator that rocks
imout...@mac.com
> <imout...@mac.com> wrote in message
> news:1103774537.9...@c13g2000cwb.googlegroups.com...
> >
> > Yeah, for the past 5+ years we've seen the PC kinda kicking around,
> > looking for a new reason for people to upgrade their existing kit
after
> > the web access wave crested.
> >
> > If had the choice between the dual-core 4Ghz x86 PCs of 2006 or the
> > Longhorn CR, I'd take the Longhorn CR (as I fully expect 2.6Ghz +
256MB
> > AGP 8x will be more than sufficient for Longhorn).
> >
>
> But what do you think Longhorn is going to *do* for you that you
can't do
> now?
that's a fair question.... Suck less, essentially. WinXP has a
completely horrid UI (as in ugly, disorganized and otherwise hard to
use), horrid API (C# and the new Longhorn API will be a significant
step into the 1990s for Micorsoft), and hopefully prove to be more
resistant to malware (not that I've had any known problems on my
machines).
OTOH, the PCs from 2006 aren't going to be /that/ much different from
the P4 machine I put together in mid-2003 (especially after I get a
6800-class card for it).
> Or, alternately, what do you do now that you think will become more
> productive with Longhorn? Or is it just a question of the aesthetic
> pleasure that a new interface will provide?
Since I (sadly) essentially live on my PC, aesthetics are somewhat
important to me (who wants to have an ugly piece of crap in their face
all day?), but ease-of-use is more so, of course. These two kind of go
together IMV. Take that fecking yellow XP search dog... Please.
> > But computers still suck, and it's not about a lack of external
> > connectivity. For an XP user, moving back to Win2k or even Win98
> > wouldn't be much of a loss wrt usability (perhaps none, the first
thing
> > I do with an XP install is turn off the fisher-price UI), but as a
Mac
> > user moving back from 10.3.x to 9 or 8.6 would be horrible, with a
lot
> > of usability advances lost.
> >
>
> Yes, well, that's because 9 *really* sucked. No pre-emptive
multi-tasking,
Uh, 8.1 and 9 had PMT for non-GUI tasks, which is pretty much what
Win9x had.
> no real threading,
wrong. 8.1 and 9 had real threading.
> no real protected memory -- under the hood it was much
> more primitive than Windows 95 (and even than Windows 3.x in some
ways).
True to a large extent, false in practice. Win95 was technically a
superior product, but if I had a choice of machines running my life
support I would choose MacOS 9 over Win95.
> It's amazing that Apple managed to get as much performance and
unsability
> out of such an archaic foundation as they did.
True. There was a long series of replacements (Pink/Taligent, Raptor,
Copland) that sorta blew up along the way...
> > These 'marginal improvements' rock. I serve mp3s wirelessly from my
g4
> > cube that sleeps rather quietly between uses. The OS of 10.3.x is
so
> > much more capable than the OS of 1999.
> >
>
> Yeah, I did that for a while until I bit the bullet and ran the wire
(I
> needed to run video cable anyway, so did audio as well). No OS
advancements
> were required for either the wired or wireless approaches, though,
and both
> worked fine. But wireless was not as simple as line-out to line-in.
Apple funds advances in its OS by OS revisions. XP's few real
improvements over Win2k (that I can find) have to do with better
integration with Bluetooth, wireless, and firewire.
OEMs can and do extend basic PC functionality (as they did with Adlib &
Soundblaster), but IME/IMV things work best when the OS provider
manages the innovation (that way we don't get absurdities like sound
card midi ports being reused as game ports).
Fair point I guess. I don't know what Apple is really doing in this
space (I don't think they are targetting x-serves for the home though).
Apple is also working on their Zeroconf stuff though, which *is*
another nifty OS advancement apropos to this.
> The 'central server' notion is based on an obsolete notion of one
expensive
> machine with an expensive hard disk per home. The homes that have
just one
> are not going to go for wireless connectivity to the TV. And the
homes that
> have more of less state-of-the-art toys (e.g. the target market)
aren't
> going to have just one computer.
I don't see a central-server strategy so much as an interconnected
strategy.
>
> > UWB looks to be very interesting in this regard for serving HD
video,
> > which is too much for anything short of 802.11a at the moment.
> >
>
> Plain old copper wires look a whole lot more interesting to me.
Seriously,
> why make it more complex?
Wires suck. Really. Wireless is indistinguishable from magic.
> > I don't know much about the price/performance thing. I buy quality
> > stuff with I expect to be 10 year + useful life. These expenses
aren't
> > that frequent for me to really look for saving $10 or whatever, I
go
> > for utility.
> >
>
> A 10 year useful life for electronics? Right.
Peripherals, not electronics in general. I have a Focus Enhancements
VGA -> composite converter, Sony DV <-> composite converters, USB input
devices, an HP laser printer, a Canon scanner, 17" DVI LCD, a Klipsch
2.1 system, etc that I fully expect to have in front-line service 10
years from now.
> What were digital cameras like 10 years ago?. Digital music players?
Wireless networking?
Yeah, but guess what, we're talking about internal vs. external crap.
Where were the PCI digital cameras, ISA music players etc 10 years ago
that would make your counterpoint remotely valid?
> Photo
> printers? Computers? DVD players? Cell phones? It doesn't matter
if they
> *last* 10 years--you'll stop wanting to use them long before then.
Some things age better than others. That's the beauty of external
peripherals, chances are it will be useful in pieces rather than
agglomerated AIO solutions that age by their weakest link.
> > That's my definition of qualitative vs. quantitative improvement.
An
> > extra 200Mhz, or hell, even going 64-bit isn't really adding many
new
> > classes of functionality.
>
> Depends--an extra 200Mhz is a lot if you're running at 100Mhz to
start with,
> but it's trivial if it represent a 10% increase. I generally don't
upgrade
> until I get a doubling in performance/capacity etc.
Right. That's basically where I+ accumulates to I*, and where I start
looking for a new machine (I look at bus speeds more than CPU speeds
tho).
I've gone from 16Mhz (bus) 68030 to a 50Mhz 601 (the intervening
evolutions were staid x86-like I+ and weren't all that interesting
until Apple's 2nd generation Powermacs came out with better graphics
card support (PCI) to a 100Mhz B&W G3, and waiting for PCI Express to
come before investing in a G5 box, since my current PBG4 is still good
enough for daily use).
> > The next frontier I see for PC's is high-res display. 80DPI is NOT
GOOD
> > ENOUGH! 180DPI will be really really cool and offer a qualitatively
> > different experience.
>
> My 15.4" notebook runs at 1680x1050. That's close to 150dpi. I
could have
> had 1900x1200 in the same size screen, which would have been over
150dpi.
> It's nice--I like it. Text is noticeably sharper. But qualitatively
> different? Nah, not really.
Then it's not sharp enough :) The qualitative difference comes where
the display is indistinguisable from paper, basically.
Depends on how important your PC is in the scheme of things I guess,
and how much crappy stuff you're willing to tolerate.
> > The thing is, Apple *is* spending ~$200M/yr on OS improvements. I
think
> > $50/yr is a much more realistic pricepoint, then again Apple is
free to
> > try to charge what the market will bear I guess.
> >
>
> It's not a question of a realistic price, really. Revenue has to pay
for OS
> R&D. It's not that big a target market and not nearly everybody in
it will
> pay for an upgrade. If selling for $50 brought in more than selling
for
> $100 (if they sold more than twice as many copies), then they'd do
that.
Sure. They figure the pirates won't pay more than $0.00 so the price
they charge now is undoubtedly the revenue maximizer.
> > > Sony, man...they can't even do a hard disk MP3 player right even
> > > after coming to the market very late and having the advantage of
> > > seeing everybody else's efforts. Creating an open platform of
any
> > > kind really isn't in their genes.
> >
> > yeah, I know. We all have our dreams :) My dream is
> > Sony/Apple/IBM/Sun/SGI hookup and take over the world. Too bad Sun
sold
> > out to MS tho, and IBM hates Sun's guts...
> >
>
> And if they all hooked up, then what? What is it you think your
ideal team
> of players is going to do that isn't happening as it is?
Market one of these:
http://web.telia.com/~u28004897/images/data/amiga500.jpg
a low-cost, capable, accessible, *home* computer. PCs suck, Macs
slightly less than Wintel.
I want a standard game-friendly, child-friendly (ie programmable in
BASIC or what have you), tank-tough (rom-boot, robust kernel, solid
API) that is a pure expression of what a HOME computer should be.
Tim Smith
imout...@mac.com wrote:
> > Yes, well, that's because 9 *really* sucked. No pre-emptive
> multi-tasking,
>
> Uh, 8.1 and 9 had PMT for non-GUI tasks, which is pretty much what
> Win9x had.
You are thinking of Win 3.x. In Win 3.x, DOS boxes were preemptively
multitasked, and GUI applications were cooperatively multitasked.
In Win 95, DOS boxes were preemptively multitasked, as were 32-bit GUI
applications. 16-bit GUI applications were cooperatively multitasked
among each other, and as a set were preemptively multitasked with the
32-bit GUI applications and the DOS boxes.
--
--Tim Smith
Snit
reply_in_group-A61...@news1.west.earthlink.net on 12/25/04
12:22 PM:
The devil is in the details: Win 9x completely froze up as much as OS
8/9... of that there is little, if any, doubt. Both could be relatively
stable if you treated them well... but trouble shooting was much easier with
OS 8/9 (nothing compared to the Extensions manager in Win 9x).
imout...@mac.com
resources.
It is my understanding, and I may be off-base but my programming
experience with Win9x tells me otherwise, that the graphics system of
Win9x was indeed not threadsafe, or for some other reason it was quite
common for the Win9x UI to lock up without being able to recover it,
Win9x's PMT or no.
What the exact deal was with Win9x I neither know nor care, but
whatever it was I don't think anyone would argue that Win9x's
multitasking capability was very, very inferior to NT 3.5 & 4. I rather
liked NT4 and if Microsoft had modern DirectX API for it I'd probably
still be using it.
Mark Weaver
<imout...@mac.com> wrote in message > >
> > But what do you think Longhorn is going to *do* for you that you
> > can't do now?
>
> that's a fair question.... Suck less, essentially. WinXP has a
> completely horrid UI (as in ugly, disorganized and otherwise hard to
> use), horrid API (C# and the new Longhorn API will be a significant
> step into the 1990s for Micorsoft), and hopefully prove to be more
> resistant to malware (not that I've had any known problems on my
> machines).
>
Which is to say, except for the malware issue, you're looking for changes
that will primarily affect developers rather than end-users. As for
'disorganized and hard-to-use'--I don't expect major changes there--a major
re-organization would be experienced by most users as problem rather than a
benefit. The Windows UI is, at this point, pretty much like the QWERTY
keyboard -- arguably not theoretically optimal, but standard. Which is
which Linux GUIs tend to follow Windows conventions for the most part.
>
> > Or, alternately, what do you do now that you think will become more
> > productive with Longhorn? Or is it just a question of the aesthetic
> > pleasure that a new interface will provide?
>
> Since I (sadly) essentially live on my PC, aesthetics are somewhat
> important to me (who wants to have an ugly piece of crap in their face
> all day?), but ease-of-use is more so, of course. These two kind of go
> together IMV. Take that fecking yellow XP search dog... Please.
>
The yellow search dog? Don't you just turn it off?
> >
> > Plain old copper wires look a whole lot more interesting to me.
> Seriously,
> > why make it more complex?
>
> Wires suck. Really. Wireless is indistinguishable from magic.
>
Wireless makes sense for some things -- where the relative positions of the
two ends of the connection change frequently. But when we're talking about
a Media-oriented desktop PC and a big screen TV, stereo, etc, we're talking
about components that stay put. In that case, wireless offers no real
benefits--wired is simpler, cheaper, and more reliable.
> >
> > My 15.4" notebook runs at 1680x1050. That's close to 150dpi. I
> > could have had 1900x1200 in the same size screen, which would have been
over
> > 150dpi. It's nice--I like it. Text is noticeably sharper. But
qualitatively
> > different? Nah, not really.
>
> Then it's not sharp enough :) The qualitative difference comes where
> the display is indistinguisable from paper, basically.
>
What's magic about 'indistinguishable from paper'? Already, it's true (or
close to true) that I can read text in the same minimum points sizes as I
can on paper. More than close enough.
>
> Depends on how important your PC is in the scheme of things I guess,
> and how much crappy stuff you're willing to tolerate.
>
Shrug. I spend a lot of time at my computer, but it just doesn't get in the
way of what I want to do. I'm still waiting to hear specific examples of
'crappy stuff' that is currently getting in my way and slowing me down. How
would my computing life be made more efficient if it were
'de-crappified'--specifically?
> > >
> >
> > And if they all hooked up, then what? What is it you think your
> ideal team
> > of players is going to do that isn't happening as it is?
>
> Market one of these:
>
> http://web.telia.com/~u28004897/images/data/amiga500.jpg
>
> a low-cost, capable, accessible, *home* computer. PCs suck, Macs
> slightly less than Wintel.
>
> I want a standard game-friendly, child-friendly (ie programmable in
> BASIC or what have you), tank-tough (rom-boot, robust kernel, solid
> API) that is a pure expression of what a HOME computer should be.
>
Why? We have game consoles that are game-friendly, child-friendly,
bulletproof--but limited in functionality (which is a big part of *why*
they're bulletproof and child-friendly). You could root for game consoles
to become the new home computing platform, replacing PCs and Macs, but I
don't see that at all--the purposes are too different. In particular, game
console manufacturers don't *want* people to buy them as general purpose
platforms--they want them to be used exclusively for the play of games for
which they collect a license fee.
We don't expect all classes of vehicle to eventually be subsumed into a
single category--I don't expect the same of home computing devices either.
imout...@mac.com
> <imout...@mac.com> wrote in message > >
> > > But what do you think Longhorn is going to *do* for you that you
> > > can't do now?
> >
> > that's a fair question.... Suck less, essentially. WinXP has a
> > completely horrid UI (as in ugly, disorganized and otherwise hard
to
> > use), horrid API (C# and the new Longhorn API will be a significant
> > step into the 1990s for Micorsoft), and hopefully prove to be more
> > resistant to malware (not that I've had any known problems on my
> > machines).
> >
>
> Which is to say, except for the malware issue, you're looking for
changes
> that will primarily affect developers rather than end-users.
More like I expect 6.0 to more akin to 10.3.x than XP.
> As for
> 'disorganized and hard-to-use'--I don't expect major changes there--a
major
> re-organization would be experienced by most users as problem rather
than a
> benefit.
There's always 'Classic Mode'. And Micorsoft has not been reticent at
how DIFFERENT Longhorn is going to be.
> The Windows UI is, at this point, pretty much like the QWERTY
> keyboard -- arguably not theoretically optimal, but standard. Which
is
> which Linux GUIs tend to follow Windows conventions for the most
part.
I haven't been tracking Longhorn all that closely, but from what I can
gather the need to differentiate Longhorn from the copycats is actually
driving a lot of this putative improvement.
My general impression from the PDC last year was Micorsoft saying to
the OS folks, "Copy THIS, MF'ers!"
> > > Or, alternately, what do you do now that you think will become
more
> > > productive with Longhorn? Or is it just a question of the
aesthetic
> > > pleasure that a new interface will provide?
> >
> > Since I (sadly) essentially live on my PC, aesthetics are somewhat
> > important to me (who wants to have an ugly piece of crap in their
face
> > all day?), but ease-of-use is more so, of course. These two kind of
go
> > together IMV. Take that fecking yellow XP search dog... Please.
> >
>
> The yellow search dog? Don't you just turn it off?
My point was the design energy that went into that damn dog could and
should have gone into making the search functionality itself more
usable (like it is on 10.3 at least).
> > > Plain old copper wires look a whole lot more interesting to me.
> > Seriously,
> > > why make it more complex?
> >
> > Wires suck. Really. Wireless is indistinguishable from magic.
> >
>
> Wireless makes sense for some things -- where the relative positions
of the
> two ends of the connection change frequently. But when we're talking
about
> a Media-oriented desktop PC and a big screen TV, stereo, etc, we're
talking
> about components that stay put. In that case, wireless offers no
real
> benefits--wired is simpler, cheaper, and more reliable.
I've run enough cat-5 to know that wire drops suck. Wireless is the
future and opens up a lot of functionality possibilies... eg. imagine
an xbox with a wireless modem instead of having to run a phoneline to
the TV center... I used to joke that what we really needed is the
wireless wire. My joke is coming true with UWB and even wireless
firewire.
> > > My 15.4" notebook runs at 1680x1050. That's close to 150dpi. I
> > > could have had 1900x1200 in the same size screen, which would
have been
> over
> > > 150dpi. It's nice--I like it. Text is noticeably sharper. But
> qualitatively
> > > different? Nah, not really.
> >
> > Then it's not sharp enough :) The qualitative difference comes
where
> > the display is indistinguisable from paper, basically.
> >
>
> What's magic about 'indistinguishable from paper'? Already, it's
true (or
> close to true) that I can read text in the same minimum points sizes
as I
> can on paper. More than close enough.
Less eyestrain, better WYSIWYG, perhaps more usable UI design. I'm just
talking qualitative jumps here, btw. The only reason the Macintosh
succeeded on the market was that its crappy 1-bit 512x384 display
offered a clean 72-dpi that developers could easily use for DTP and
other pixel-pushing applications. True, 72-dpi was the heavy lifting to
get displays out of the 80x25 stone age, but unfortunately, Apple has
yet to offer a qualitative improvement over 72-dpi for more than 20
years!
(while eg. Dell and I believe IBM have been offering higher-res
displays for years now, and there is some form of support for this in
Windows, but not sure how well it is).
Actually going resolution-independent and enabling qualitatively better
WYSIWYG would be an immense boost for wanting to upgrade, or even
switch from Windows, to a Mac.
> > Depends on how important your PC is in the scheme of things I
guess,
> > and how much crappy stuff you're willing to tolerate.
> >
>
> Shrug. I spend a lot of time at my computer, but it just doesn't get
in the
> way of what I want to do. I'm still waiting to hear specific
examples of
> 'crappy stuff' that is currently getting in my way and slowing me
down. How
> would my computing life be made more efficient if it were
> 'de-crappified'--specifically?
What's wrong with XP, you say? All we need to do is inspect the feature
list for Longhorn to identify the outstanding issues.
· No fully capable commandline without having to install eg. mingw.
Microsoft's working on this.
· Search really really sucks. Ooops, WinFS was moved past Longhorn.
At least 10.4 looks to be shipping with an improved search (10.3's is
already pretty good compared to XP).
· Fully compositing window server with backing store (Avalon). A
small aesthetic thing.
· Better system-wide antialiasing of text and UI decorations I assume
(Aero Glass). XP's default theme is butt ugly in this department, but
Win32 itself is somewhat better (eg. VisualStudio 7 is more refined
than the system UI proper).
As for where XP falls down for me compared to 10.3, we have:
· Need for polling in the user view of the filesystem.
· Really brain-dead structure and navigation of the directory tree.
Rooting my home folder in \Documents and Settings\ is lame.
· Inflexibility about boot partitions, too much junk in the boot
partition to easily boot between multiple OS's (eg. Win2k and XP)
without having to reinstall every needed app on each OS.
· Sleep sucks compared to OS X's instant wake/sleep capabilities
I don't know how much I value the added utility of MacOS over Windows.
In 1989 it was easily in the thousands when I bought my IIcx (6.0.3 vs
Windows/386) -- sure I could get work done in DOS/Windows but why
inflict that amount of pain on myself???? In 1995 it was still in the
hundreds (7.5 over Win95). In 2004, the only thing XP does for me is be
a PC game loader, DirectX platform, and serviceable Java/OpenGL dev
platform too.
The OS user experience itself is pure crap compared to 10.3 and I am
indeed happy that Micorsoft is shitcanning it completely and reworking
it from scratch for NT6...
> > > And if they all hooked up, then what? What is it you think your
> > ideal team
> > > of players is going to do that isn't happening as it is?
> >
> > Market one of these:
> >
> > http://web.telia.com/~u28004897/images/data/amiga500.jpg
> >
> > a low-cost, capable, accessible, *home* computer. PCs suck, Macs
> > slightly less than Wintel.
> >
> > I want a standard game-friendly, child-friendly (ie programmable in
> > BASIC or what have you), tank-tough (rom-boot, robust kernel, solid
> > API) that is a pure expression of what a HOME computer should be.
> >
>
> Why? We have game consoles that are game-friendly, child-friendly,
> bulletproof--but limited in functionality (which is a big part of
*why*
> they're bulletproof and child-friendly).
Good question. I think kids are too passive consumers these days and
deserve an accessible platform to dig into like I had back in the day
with the Atari, TRS-80, and the Apple II.
> You could root for game consoles
> to become the new home computing platform, replacing PCs and Macs,
but I
> don't see that at all--the purposes are too different.
My PC hardware needs would be met quite adequately by a PS3 that
featured firewire, USB, and DVI for the video. And how.
API-wise, the Java 1.5/OpenGL/OpenAL platform would probably suffice.
> In particular, game
> console manufacturers don't *want* people to buy them as general
purpose
> platforms--they want them to be used exclusively for the play of
games for
> which they collect a license fee.
Yeah, I know. This is why this is my dream. Still, Sony could license
the pro side of the hardware to its licensees while opening up the
hobby side, like it has done with the PS 1 (Net Yarouze) and the PS 2
(the Linux SDK).
> We don't expect all classes of vehicle to eventually be subsumed into
a
> single category--I don't expect the same of home computing devices
either.
Apple's got a problem. I don't need an iMac G5 or dual desktop tower
really. What I need is an updated Amiga 500 with firewire (and/or
external SATA), usb, dvi.
Same thing with Dell, too. The Xbox 2's specs kick total ass on today's
top of the line hardware, really. Microsoft would own the world if it
released NT6 on that PPC box.
Mark Weaver
<imout...@mac.com> wrote in message
news:1104073579.6...@f14g2000cwb.googlegroups.com...
>
> I haven't been tracking Longhorn all that closely, but from what I can
> gather the need to differentiate Longhorn from the copycats is actually
> driving a lot of this putative improvement.
>
> My general impression from the PDC last year was Micorsoft saying to
> the OS folks, "Copy THIS, MF'ers!"
>
That is a problem--it can be more in MS's interest than users' interest to
make changes. Of course, MS can't dictate because the user base might
refuse to move (as has happened with previous, less radical Windows
releases). Radical changes may tend to fend off cloners, but they may also
tend to slow the adoption rate (and hurt MS's revenues).
> My point was the design energy that went into that damn dog could and
> should have gone into making the search functionality itself more
> usable (like it is on 10.3 at least).
Energy isn't in short supply--MS can hire an essentially infinite quantity
of software dev energy. The search changes from 2K to XP were just a bad
call. But better search won't require Longhorn--Google desktop search is
giving them the sense of urgency, and the 'Lookout' search acquisition is
supplying the technology. Won't be long for that--a beta's already out.
> I've run enough cat-5 to know that wire drops suck. Wireless is the
> future and opens up a lot of functionality possibilies... eg. imagine
> an xbox with a wireless modem instead of having to run a phoneline to
> the TV center... I used to joke that what we really needed is the
> wireless wire. My joke is coming true with UWB and even wireless
> firewire.
You don't have to imagine a wireless xbox -- there's already a wifi adapter.
For networking, I have a mix of wired and wireless -- Cat 5 where it's not
too much trouble and Wifi for notebooks and rooms where running wire would
have been a problem. But I'd rather have wire for the non-portable machines
if it weren't hard to run it. Wires are simpler, more reliable, have no
interference issues, no security issues, etc.
> Less eyestrain, better WYSIWYG, perhaps more usable UI design. I'm just
> talking qualitative jumps here, btw.
All that means is 'somewhat sharper'. Given that current displays are
already more than capable of photo realism, I don't see them limiting what
is possible in UI design. If what everybody wanted were scrollbars that
appeared to be carved out of quarter-sawn oak, it could be done now.
> (while eg. Dell and I believe IBM have been offering higher-res
> displays for years now, and there is some form of support for this in
> Windows, but not sure how well it is).
You can pick a nominal DPI you like and Windows scales the fonts, titlebars,
etc. (96 and 120 dpi are the standard settings or you can specify a custom
setting). What I find more useful than that, though, is the CTRL
scroll-wheel font scaling for web pages (Firefox does this especially
nicely).
> Actually going resolution-independent and enabling qualitatively better
> WYSIWYG would be an immense boost for wanting to upgrade, or even
> switch from Windows, to a Mac.
Well, Windows already has the ability to be adapted to different DPIs.
Maybe that's why PC manufacturers offer 1900x1200 and 1680x1050 15.4"
widescreen notebooks and Apple doesn't?
> · No fully capable commandline without having to install eg. mingw.
> Microsoft's working on this.
A developer-only issue. If you want a better command shell, install one.
And surely you're not advocating MS using it's market power to 'bundle' an
advanced shell and put the ISVs out of business? ;)
> · Search really really sucks. Ooops, WinFS was moved past Longhorn.
> At least 10.4 looks to be shipping with an improved search (10.3's is
> already pretty good compared to XP).
Improved search won't require Longhorn--it'll be available from MS, Google,
Copernic, and Yahoo. Competition is good, and I don't see any particular
reason search has to be tightly integrated into the OS.
> · Fully compositing window server with backing store (Avalon). A
> small aesthetic thing.
I'm already dreading the pointless swooping, twisting, and spinning effects
;)
> · Better system-wide antialiasing of text and UI decorations I assume
> (Aero Glass). XP's default theme is butt ugly in this department, but
> Win32 itself is somewhat better (eg. VisualStudio 7 is more refined
> than the system UI proper).
I dislike OS Xs literally eye *candy* aesthetics, but I'm sure I could learn
to live with it if I used it more. In general, though, I really don't much
care what the menus, scrollbars, and buttons look like as long as they work
as expected. Anti-aliasing I have no problem with--text clarity is already
very good and any improvements seem more dependent on display changes than
OS changes.
> As for where XP falls down for me compared to 10.3, we have:
> · Need for polling in the user view of the filesystem.
> · Really brain-dead structure and navigation of the directory tree.
> Rooting my home folder in \Documents and Settings\ is lame.
Ah, maybe, but most users never navigate to their home directory that
way--they just use 'My Documents'. They wouldn't notice one way or the
other if 'My Documents' linked to some other home directory.
> · Inflexibility about boot partitions, too much junk in the boot
> partition to easily boot between multiple OS's (eg. Win2k and XP)
> without having to reinstall every needed app on each OS.
Another developer-only issue. But it's not mainly not about the boot
partition, though. Having to re-install most apps on each OS instance is a
property of the install process and registry--it's not going to be changed.
Developers will have to continue to live with it.
> The OS user experience itself is pure crap compared to 10.3 and I am
> indeed happy that Micorsoft is shitcanning it completely and reworking
> it from scratch for NT6...
Backward compatibility and user expectations will limit how radical the
changes will turn out to be when Longhorn is actually released. At the
beginning of the cycle, the next version is always going to be completely
different. But at release time, it never is.
> Same thing with Dell, too. The Xbox 2's specs kick total ass on today's
> top of the line hardware, really. Microsoft would own the world if it
> released NT6 on that PPC box.
Xbox2 and PS3 are going to have to stick with something close to the current
pricepoint. Which means selling machines at a loss and recouping with
licensed game sales. Which means actively preventing buyers from using the
machine as a 'kick ass' general purpose machine.
imout...@mac.com
> > My point was the design energy that went into that damn dog could
and
> > should have gone into making the search functionality itself more
> > usable (like it is on 10.3 at least).
>
> Energy isn't in short supply--MS can hire an essentially infinite
quantity
> of software dev energy.
They can hire an infinite amount, but as this case shows at some point
the labor rubber must meet the GM (golden master) road.
My original point wrt the yellow dog was that usability/productivity
seems to correlate with visual
'aesthetics' of a GUI. Whether their is a causal link is debatable, but
at any rate without having /used/ longhorn yet I can only take it that
its supposed improvements in aesthetics will also result in improved
usability, since this is the general pattern of my computer experience
dating from the beginnings of time (~1979).
> The search changes from 2K to XP were just a bad
> call. But better search won't require Longhorn--Google desktop
search is
> giving them the sense of urgency, and the 'Lookout' search
acquisition is
> supplying the technology. Won't be long for that--a beta's already
out.
That's good. What with the announcement that "Avalon" will ship for XP,
perhaps I won't need to pay the upgrade cost for Longhorn at all :)
> > I've run enough cat-5 to know that wire drops suck. Wireless is the
> > future and opens up a lot of functionality possibilies... eg.
imagine
> > an xbox with a wireless modem instead of having to run a phoneline
to
> > the TV center... I used to joke that what we really needed is the
> > wireless wire. My joke is coming true with UWB and even wireless
> > firewire.
>
> You don't have to imagine a wireless xbox -- there's already a wifi
adapter.
> For networking, I have a mix of wired and wireless -- Cat 5 where
it's not
> too much trouble and Wifi for notebooks and rooms where running wire
would
> have been a problem. But I'd rather have wire for the non-portable
machines
> if it weren't hard to run it. Wires are simpler, more reliable, have
no
> interference issues, no security issues, etc.
Sure. I have a 4-port hub to interconnect my core stuff; wireless does
also have annoying bandwidth limitations.
I'm just talking about the future, really. The future is wireless, and
it's going to be cool. And Apple's on it.
> > Less eyestrain, better WYSIWYG, perhaps more usable UI design. I'm
just
> > talking qualitative jumps here, btw.
>
> All that means is 'somewhat sharper'.
No, 'somewhat sharper' is a quantitative difference. My PBG4 display is
"somewhat sharper" than my 17" studio display (the pbg4 has ~100 DPI,
while the LCD has an 85 DPI.
The argument here arose from what I consider to be the differences
between innovation in Apple and x86 land. In the 80s & 90s Apple had
the balls to deliver innovation that went beyond minor incremental
improvements like the piddling jump from 85DPI to 100DPI in the past. I
assert that there *is*, indeed, a *qualitative* threshold where display
technology looks as perfect as printed paper, and that is a meaningful
innovation outside of the boring *quantitative* incrementalism that we
see with uncoordinated, normal PC evolution.
I've talked with Peter Graffagnino, division director of Apple's
graphics group, firsthand about this several times, and have seen the
work they are doing with this. Moving to a paper-like interface
requires the OS to paradigm-shift from pixel measurements to
dimensional measurements. Indeed, Apple probably dropped the ball by
not changing their API when they had the chance with 10.0, though in
the interim this can be faked somewhat by translating (behind the app's
back) notional pixels into device pixels.
In the PalmOS space I saw technology run into this problem head-on, and
the OS solutions were not so good.
Linux is totally useless for these kinds of interesting innovations
(that I call I*), since they are reactive, copycats, and get more
mileage out of philosophical debates and whacking off to slashdot than
getting work on the screen.
Micorsoft development has historically been in a similar category of
gating rather than pushing I* innovation, though with the media center
and tablet efforts, not to mention the immense amount of money they
burn in the pure R&D division, they seem to changing their spots.
> Given that current displays are
> already more than capable of photo realism, I don't see them limiting
what
> is possible in UI design. If what everybody wanted were scrollbars
that
> appeared to be carved out of quarter-sawn oak, it could be done now.
I sorta agree, which is why I equivocated with "perhaps more usable UI
design". Still, I suspect once DPI moves past 180 or so we'll never go
back to 85 or even 150. There *is* a difference, I've seen it.
> > (while eg. Dell and I believe IBM have been offering higher-res
> > displays for years now, and there is some form of support for this
in
> > Windows, but not sure how well it is).
>
> You can pick a nominal DPI you like and Windows scales the fonts,
titlebars,
> etc. (96 and 120 dpi are the standard settings or you can specify a
custom
> setting). What I find more useful than that, though, is the CTRL
> scroll-wheel font scaling for web pages (Firefox does this especially
> nicely).
That'd be cool.
> > Actually going resolution-independent and enabling qualitatively
better
> > WYSIWYG would be an immense boost for wanting to upgrade, or even
> > switch from Windows, to a Mac.
>
> Well, Windows already has the ability to be adapted to different
DPIs.
> Maybe that's why PC manufacturers offer 1900x1200 and 1680x1050 15.4"
> widescreen notebooks and Apple doesn't?
Most certainly. Apple is behind Wintel here, full stop.
> > · No fully capable commandline without having to install eg.
mingw.
> > Microsoft's working on this.
>
> A developer-only issue. If you want a better command shell, install
one.
> And surely you're not advocating MS using it's market power to
'bundle' an
> advanced shell and put the ISVs out of business? ;)
As a developer, it's my issue. Tho of course it's ironic that a Mac
currently has a better CLI than a Micorsoft box. Whodathunk? Like its
abandonment of IE, Micorsoft sure has an odd habit of limping along
with crappy tech for far too long...
> > · Search really really sucks. Ooops, WinFS was moved past
Longhorn.
> > At least 10.4 looks to be shipping with an improved search (10.3's
is
> > already pretty good compared to XP).
>
> Improved search won't require Longhorn--it'll be available from MS,
Google,
> Copernic, and Yahoo. Competition is good, and I don't see any
particular
> reason search has to be tightly integrated into the OS.
You haven't seen Spotlight. Apple's doing something quite special with
this. I still wonder if they can pull it off, really.
> > · Fully compositing window server with backing store (Avalon). A
> > small aesthetic thing.
>
> I'm already dreading the pointless swooping, twisting, and spinning
effects
> ;)
I turn that crap off. The window backing store is nice though,
especially for programming. MacOS is pretty good about only doing
damage repair on resize events.
> > · Better system-wide antialiasing of text and UI decorations I
assume
> > (Aero Glass). XP's default theme is butt ugly in this department,
but
> > Win32 itself is somewhat better (eg. VisualStudio 7 is more refined
> > than the system UI proper).
>
> I dislike OS Xs literally eye *candy* aesthetics,
me too. Can't stand "aqua". Well I can, but I'd much prefer a more
muted, mature, thought-out, design.
> but I'm sure I could learn
> to live with it if I used it more. In general, though, I really
don't much
> care what the menus, scrollbars, and buttons look like as long as
they work
> as expected. Anti-aliasing I have no problem with--text clarity is
already
> very good and any improvements seem more dependent on display changes
than
> OS changes.
Text clarity isn't so good for me on XP. Mebbe I need that 150DPI
monitor :)
> > As for where XP falls down for me compared to 10.3, we have:
>
> > · Need for polling in the user view of the filesystem.
>
>
http://msdn.microsoft.com/library/default.asp?url=/library/en-us/fileio/base/obtaining_directory_change_notifications.asp
Somebody better give the explorer team that page...
> > · Really brain-dead structure and navigation of the directory
tree.
> > Rooting my home folder in \Documents and Settings\ is lame.
>
> Ah, maybe, but most users never navigate to their home directory that
> way--they just use 'My Documents'. They wouldn't notice one way or
the
> other if 'My Documents' linked to some other home directory.
Still, it's something that gets in my way and can be improved.
> > · Inflexibility about boot partitions, too much junk in the boot
> > partition to easily boot between multiple OS's (eg. Win2k and XP)
> > without having to reinstall every needed app on each OS.
>
> Another developer-only issue. But it's not mainly not about the boot
> partition, though. Having to re-install most apps on each OS
instance is a
> property of the install process and registry--it's not going to be
changed.
> Developers will have to continue to live with it.
yeah. I disagree about it being a developer-only issue. On the mac it's
still possible to separate the OS install from the apps, especially
shrink-wrap ISV app installs. This is great for archiving,
trouble-shooting, and safe computing in general, and the ideal OS
wouldn't have these silly interlinkages at all.
> > The OS user experience itself is pure crap compared to 10.3 and I
am
> > indeed happy that Micorsoft is shitcanning it completely and
reworking
> > it from scratch for NT6...
>
> Backward compatibility and user expectations will limit how radical
the
> changes will turn out to be when Longhorn is actually released. At
the
> beginning of the cycle, the next version is always going to be
completely
> different. But at release time, it never is.
True enough.
> > Same thing with Dell, too. The Xbox 2's specs kick total ass on
today's
> > top of the line hardware, really. Microsoft would own the world if
it
> > released NT6 on that PPC box.
>
> Xbox2 and PS3 are going to have to stick with something close to the
current
> pricepoint. Which means selling machines at a loss and recouping
with
> licensed game sales. Which means actively preventing buyers from
using the
> machine as a 'kick ass' general purpose machine.
Depends if it's a zero-sum game on the software side, IOW mebbe
enabling the hobbyists won't cut sales so much, or sales like compilers
etc could fill the gap (Sony charged $300+ for its SDK).
It's an interesting idea, one of the few tech ideas I'd like to be
employed on, really.
Mark Weaver
<imout...@mac.com> wrote in message
news:1104166435.7...@f14g2000cwb.googlegroups.com...
> They can hire an infinite amount, but as this case shows at some point
> the labor rubber must meet the GM (golden master) road.
Of course, but what I meant was that they didn't have to decide between the
yellow search dog and other search enhancements because they lacked the
capacity to do both--what they ended up with was what they (unfortunately)
decided was best.
> I can only take it that
> its supposed improvements in aesthetics will also result in improved
> usability, since this is the general pattern of my computer experience
> dating from the beginnings of time (~1979).
Yes, but after a couple of decades, things tend to converge with a new
technology. I don't expect PC UIs to change as much in the next 30 years as
they did in a few years in the 1980's--just as I don't expect the basic car
interface (wheel, pedals, turn signal stalk, power locks and windows) to
change--regardless of what radical changes may occur under the hood. Some
early cars had tiller steering--but now we're stuck with the wheel until, I
suspect, man no longer moves around the surface of any planet in wheeled
vehicles.
> No, 'somewhat sharper' is a quantitative difference. My PBG4 display is
> "somewhat sharper" than my 17" studio display (the pbg4 has ~100 DPI,
> while the LCD has an 85 DPI.
Well, it *is* a noticeable difference. I have compared my WXGA+ screen
side-by-side to a PBG4 and the extra resolution is quite apparent--1900x1200
in the same form factor would be even more so.
> I've talked with Peter Graffagnino, division director of Apple's
> graphics group, firsthand about this several times, and have seen the
> work they are doing with this. Moving to a paper-like interface
> requires the OS to paradigm-shift from pixel measurements to
> dimensional measurements.
Well, it doesn't really require that--it's possible to remain pixel-oriented
and still support variable DPIs. Text and vector graphics could be
'de-pixelized', but digital cameras are going to be generating pixel-based
images indefinitely. Photoshop is not going to stop dealign with pixels.
> In the PalmOS space I saw technology run into this problem head-on, and
> the OS solutions were not so good.
Yes, the fixed-res Palm screen was a serious mistake.
> I sorta agree, which is why I equivocated with "perhaps more usable UI
> design". Still, I suspect once DPI moves past 180 or so we'll never go
> back to 85 or even 150. There *is* a difference, I've seen it.
There has been a visible difference every time I've upgraded my display in
size and resolution, and I wouldn't want to go back. But it really hasn't
changed how I use the computer.
> > What I find more useful than that, though, is the CTRL
> > scroll-wheel font scaling for web pages (Firefox does this especially
> > nicely).
>
> That'd be cool.
Doesn't the Mac version of Firefox do that? It'd suck not having that.
> As a developer, it's my issue. Tho of course it's ironic that a Mac
> currently has a better CLI than a Micorsoft box. Whodathunk? Like its
> abandonment of IE, Micorsoft sure has an odd habit of limping along
> with crappy tech for far too long...
I don't know what MS is doing with IE. The idea that they're only going to
provide upgrades with new versions of Windows is nutso. Firefox is going to
hit a tipping point and they're going to lose the whole browser enchilada if
they don't get busy.
> Text clarity isn't so good for me on XP. Mebbe I need that 150DPI
> monitor :)
It'd be worth taking a look. Funny though that very few retail machines
include the high res displays even though cost-wise it's not that much of an
increase.
> yeah. I disagree about it being a developer-only issue. On the mac it's
> still possible to separate the OS install from the apps, especially
> shrink-wrap ISV app installs. This is great for archiving,
> trouble-shooting, and safe computing in general, and the ideal OS
> wouldn't have these silly interlinkages at all.
I'm not a fan of the centralized registry at all, but given all the apps in
the field that rely on it, it'd be pretty tough to get rid of at this point.
> Depends if it's a zero-sum game on the software side, IOW mebbe
> enabling the hobbyists won't cut sales so much, or sales like compilers
> etc could fill the gap (Sony charged $300+ for its SDK).
Or ISVs could be required to pay license fees when selling general-purpose
software for the platform just as game vendors do now. But I'd hate to see
a closed, proprietary model like that become the platform of the future.
imout...@mac.com
> <imout...@mac.com> wrote in message
> news:1104166435.7...@f14g2000cwb.googlegroups.com...
> > They can hire an infinite amount, but as this case shows at some
point
> > the labor rubber must meet the GM (golden master) road.
>
> Of course, but what I meant was that they didn't have to decide
between the
> yellow search dog and other search enhancements because they lacked
the
> capacity to do both--what they ended up with was what they
(unfortunately)
> decided was best.
yeah, I wasn't criticising Micorsoft for not spending enough on XP R&D
(hahah), I was criticising them for making the usual bodged effort.
> > I can only take it that
> > its supposed improvements in aesthetics will also result in
improved
> > usability, since this is the general pattern of my computer
experience
> > dating from the beginnings of time (~1979).
>
> Yes, but after a couple of decades, things tend to converge with a
new
> technology. I don't expect PC UIs to change as much in the next 30
years as
> they did in a few years in the 1980's--
wow. I do. I think we'll have retinal tracking in place of the mouse.
subvocal or thought pickup instead of typing and mouse buttons.
Infinite, dynamic knowledge spaces instead of fixed windows and
desktops.
I do agree though that present UI hasn't changed much in the past 25
years from its start in Xerox's labs. Not much at all.
>just as I don't expect the basic car
> interface (wheel, pedals, turn signal stalk, power locks and windows)
to
> change--regardless of what radical changes may occur under the hood.
I think in the next 30 we won't be driving cars at all (Minority
Report-style automatic driving). Unless of course we're stuck with the
present government leadership more interested in bombing camel herders
than gently directing civilian research efforts.
> Some
> early cars had tiller steering--but now we're stuck with the wheel
until, I
> suspect, man no longer moves around the surface of any planet in
wheeled
> vehicles.
I prefer the Spy Hunter driving interface myself :)
> > No, 'somewhat sharper' is a quantitative difference. My PBG4
display is
> > "somewhat sharper" than my 17" studio display (the pbg4 has ~100
DPI,
> > while the LCD has an 85 DPI.
>
> Well, it *is* a noticeable difference. I have compared my WXGA+
screen
> side-by-side to a PBG4 and the extra resolution is quite
apparent--1900x1200
> in the same form factor would be even more so.
yeah, incremental improvement is usually noticeable. The 25Mhz IIci was
noticeably faster than the IIcx, just as the Quadras were noticeably
faster than the IIci. But there was really nothing you could do on a
40Mhz 68040 Quadra that you couldn't also do on a 16Mhz IIcx with
approximately the same level of productivity.
> > I've talked with Peter Graffagnino, division director of Apple's
> > graphics group, firsthand about this several times, and have seen
the
> > work they are doing with this. Moving to a paper-like interface
> > requires the OS to paradigm-shift from pixel measurements to
> > dimensional measurements.
>
> Well, it doesn't really require that--it's possible to remain
pixel-oriented
> and still support variable DPIs.
Yes, but I think artifacting will still result, and/or the app-level UI
layout will not function optimally. Apple's transition plan is to
maintain backwards compatibility with the present nominal 72DPI
standard I guess. We'll see how well that works in practice. I haven't
seen the Windows UI function at extremely high res but I've heard that
it doesn't quite work seamlessly.
> Text and vector graphics could be
> 'de-pixelized', but digital cameras are going to be generating
pixel-based
> images indefinitely. Photoshop is not going to stop dealign with
pixels.
Pixels are artifacts of low-res displays. Text and pictures in eg.
magazines (generally) do not have recognizable pixels.
> > In the PalmOS space I saw technology run into this problem head-on,
and
> > the OS solutions were not so good.
>
> Yes, the fixed-res Palm screen was a serious mistake.
Sony was the first to go 320x320 in the same form factor and the API
for that was something of a kludge.
> > I sorta agree, which is why I equivocated with "perhaps more usable
UI
> > design". Still, I suspect once DPI moves past 180 or so we'll never
go
> > back to 85 or even 150. There *is* a difference, I've seen it.
>
> There has been a visible difference every time I've upgraded my
display in
> size and resolution, and I wouldn't want to go back. But it really
hasn't
> changed how I use the computer.
I think it can change how UI is displayed. I write apps for myself that
work both on a Clie and the Mac/PC. I can present a different UI on the
very high res Clie than I can on the Mac/PC (which is displayed on the
85DPI 17" LCD). Whether the high-res UI is significantly different is
open to interpretation, but in my experiments it certainly 'feels' a
lot different, like I'm operating a book not a software program.
> > > What I find more useful than that, though, is the CTRL
> > > scroll-wheel font scaling for web pages (Firefox does this
especially
> > > nicely).
> >
> > That'd be cool.
>
> Doesn't the Mac version of Firefox do that? It'd suck not having
that.
Dunno. I lounge with my powerbook on a sofa normally so for the most
part text is text.
> > As a developer, it's my issue. Tho of course it's ironic that a Mac
> > currently has a better CLI than a Micorsoft box. Whodathunk? Like
its
> > abandonment of IE, Micorsoft sure has an odd habit of limping along
> > with crappy tech for far too long...
>
> I don't know what MS is doing with IE. The idea that they're only
going to
> provide upgrades with new versions of Windows is nutso. Firefox is
going to
> hit a tipping point and they're going to lose the whole browser
enchilada if
> they don't get busy.
I think they assumed they had successfully killed Netscape/Mozilla and
so could move on to other threatening areas of compatibility the
non-Windows world was creating. Actually fixing IE to more accurately
conform to eg. CSS standards would be counterproductive to maintaining
market domination and hence was not funded (Micorsoft of course
/desires/ people designing websites to its broken CSS model, design
effort toward that reduces the ability to develop platform-agnositic
solutions).
(I consider this an example of general market failure and not per se
'evil' on the part of Micorsoft. They're trapped by their own success,
and our take-no-prisoners laissez-faire economic system requires
Micorsoft to defends its market position both actively and passively
however it can.)
> > Text clarity isn't so good for me on XP. Mebbe I need that 150DPI
> > monitor :)
>
> It'd be worth taking a look. Funny though that very few retail
machines
> include the high res displays even though cost-wise it's not that
much of an
> increase.
Too bad newegg doesn't have a DPI category greater than ~100. Odd
actually.
wonder what happened to this:
http://www.extremetech.com/article2/0,3973,525370,00.asp
I don't see anything like it at newegg. Perhaps the difficulty was
these high-DPI displays exceeded the capacity of normal DVI to drive
them. Apple's high-res stuff was delayed 4 months not by manufacturing
but by the lateness of the 6800 Ultra.
> > yeah. I disagree about it being a developer-only issue. On the mac
it's
> > still possible to separate the OS install from the apps, e
specially
> > shrink-wrap ISV app installs. This is great for archiving,
> > trouble-shooting, and safe computing in general, and the ideal OS
> > wouldn't have these silly interlinkages at all.
>
> I'm not a fan of the centralized registry at all, but given all the
apps in
> the field that rely on it, it'd be pretty tough to get rid of at this
point.
Nah, you can virtualize it and or otherwise rearchitect it such that
Longhorn apps do it one way and Win32 does it the old way.
> > Depends if it's a zero-sum game on the software side, IOW mebbe
> > enabling the hobbyists won't cut sales so much, or sales like
compilers
> > etc could fill the gap (Sony charged $300+ for its SDK).
>
> Or ISVs could be required to pay license fees when selling
general-purpose
> software for the platform just as game vendors do now. But I'd hate
to see
> a closed, proprietary model like that become the platform of the
future.
yeah, the economics of competing against Windows is difficult. Amiga,
Atari, Sharp, Fujitsu were in a more fractured market and hence going
proprietary didn't appear to be the ultimately quixotic move that it
turned out to be.
But I think the web has the potential to change things. IMV it has
certainly saved the Mac's bacon, like a government pension providing a
fixed but luckily significant amount of life support in an otherwise
Windows world. With web standards compatibility you've got about 30% of
what you need a home machine for. The Java API could provide the other
30%, and perhaps OpenGL and OpenAL would be another 30%, leaving 10%
for me to think about as I hit the post button :)
Mark Weaver
<imout...@mac.com> wrote in message
> >
> > Yes, but after a couple of decades, things tend to converge with a
> > new technology. I don't expect PC UIs to change as much in the next 30
> > years as they did in a few years in the 1980's--
>
> wow. I do. I think we'll have retinal tracking in place of the mouse.
> subvocal or thought pickup instead of typing and mouse buttons.
> Infinite, dynamic knowledge spaces instead of fixed windows and
> desktops.
>
I'd take that bet. Human brains were built to manipulate things through the
use of hands. Maybe direct manipulation (stylus on screen) will replace
mice, but I doubt even that. As for 'infinite, dynamic knowledge spaces'
instead of fixed windows and desktops--the web is already an 'infinite,
dynamic knowledge space' but users access it using virtual rectangular
pages of text and images (or virtual rectangular view of scrolls, if you
prefer). That really hasn't changed since the Gutenberg Bible (or, hell,
since the Dead Sea Scrolls), so I really don't expect that fundamental
aspect of computer UIs to change.
>
> I think in the next 30 we won't be driving cars at all (Minority
> Report-style automatic driving). Unless of course we're stuck with the
> present government leadership more interested in bombing camel herders
> than gently directing civilian research efforts.
>
Such a vision is not possible without an incredibly centralized, pervasive,
big-brotherish system. Ugh. And do you envision automatic driving
everywhere there are roads? Even dirt 2-tracks? And what about areas where
there are cyclists, pedestrians, etc?
Maybe, MAYBE, well eventually have auto driving on limited-access 'smart'
highways (that vision has been around for 50 years at least, BTW) but beyond
outside that, we'll be driving a vehicle with wheels controlled by steering
wheel, brake, and gas pedals (and it'll be called a 'gas' pedal long after
there's no more gas in cars). There will be better safety equipment and
individual vehicles will have smart collision avoidance equipment
(ultimately probably mandated), but we'll still be driving.
>
> > Text and vector graphics could be
> > 'de-pixelized', but digital cameras are going to be generating
> pixel-based
> > images indefinitely. Photoshop is not going to stop dealign with
> pixels.
>
> Pixels are artifacts of low-res displays. Text and pictures in eg.
> magazines (generally) do not have recognizable pixels.
>
Even when pixels are small enough not to be individually visible (which, is
pretty much the case now), digital photos are still going to *consist* of
pixels, and editing is going to involve changing the color value of large
numbers of individual pixels.
>
> > > > What I find more useful than that, though, is the CTRL
> > > > scroll-wheel font scaling for web pages (Firefox does this
> especially
> > > > nicely).
> > >
> > > That'd be cool.
> >
> > Doesn't the Mac version of Firefox do that? It'd suck not having
> that.
>
> Dunno. I lounge with my powerbook on a sofa normally so for the most
> part text is text.
>
Well, right--but many web pages specify specific text sizes which may be
larger or smaller than you'd like. Firefox lets you scale and reflow the
text on the fly with the scroll wheel.
>
> I think they assumed they had successfully killed Netscape/Mozilla and
> so could move on to other threatening areas of compatibility the
> non-Windows world was creating. Actually fixing IE to more accurately
> conform to eg. CSS standards would be counterproductive to maintaining
> market domination and hence was not funded (Micorsoft of course
> /desires/ people designing websites to its broken CSS model, design
> effort toward that reduces the ability to develop platform-agnositic
> solutions).
>
Conformance isn't really the issue -- users don't care about that because
web designers have to make sure their sites work with IE. The issue is
features and, more importantly, security. MS has essentially stopped
developing standalone versions of IE. So at this point, if you like tabs in
your browser or the scrolling I talked about, there's no point in waiting
for another IE release to add them, because MS has said there isn't going to
be one (until you upgrade to a new version of Windows). I do expect them to
realize this is a mistake and backtrack, but I wonder how much momentum
they'll let Firefox get before making that call.
> (I consider this an example of general market failure and not per se
> 'evil' on the part of Micorsoft. They're trapped by their own success,
> and our take-no-prisoners laissez-faire economic system requires
> Micorsoft to defends its market position both actively and passively
> however it can.)
>
Well, but thing is that they're NOT defending their browser market position
effectively at this point.
And, BTW, if our economic system were centrally managed and less open, the
revolutionary changes in technology that have emerged in the last couple of
decades probably would have been delayed or even prevented (on the grounds
that they might be highly disruptive of established interests--which in
fact, they have been).
>
> Too bad newegg doesn't have a DPI category greater than ~100. Odd
> actually.
>
AFAIK, nobody sells the high-res 15 or 17 inch displays as standalone
monitors, only as laptop displays. I don't know why that is. Unfortunate,
too. I have a 20" LCD and I'm not sure I ever want a screen larger than
that on my desk, but higher-res would be nice.
>
> Nah, you can virtualize it and or otherwise rearchitect it such that
> Longhorn apps do it one way and Win32 does it the old way.
>
Right, but that means Longhorn has to retain the registry for existing apps.
Any any new apps developers continue to write to the Win32 standard (which,
face it, for years most ISVs will do to get automatic compatibility for
people still running XP). Ergo, registry lives on.
> But I think the web has the potential to change things. IMV it has
> certainly saved the Mac's bacon, like a government pension providing a
> fixed but luckily significant amount of life support in an otherwise
> Windows world. With web standards compatibility you've got about 30% of
> what you need a home machine for. The Java API could provide the other
> 30%, and perhaps OpenGL and OpenAL would be another 30%, leaving 10%
> for me to think about as I hit the post button :)
>
A lot of what computers are used for is still supported by native code
calling native APIs. It's certainly not in the interest of MS to change
that. And neither Linux nor OS X shows signs of going away from it either.
imout...@mac.com
> <imout...@mac.com> wrote in message
> > >
> > > Yes, but after a couple of decades, things tend to converge with
a
> > > new technology. I don't expect PC UIs to change as much in the
next 30
> > > years as they did in a few years in the 1980's--
> >
> > wow. I do. I think we'll have retinal tracking in place of the
mouse.
> > subvocal or thought pickup instead of typing and mouse buttons.
> > Infinite, dynamic knowledge spaces instead of fixed windows and
> > desktops.
> >
>
> I'd take that bet. Human brains were built to manipulate things
through the
> use of hands.
oh, we'd still use our hands no doubt, perhaps like either the VR
scenes in Johnny Mnemonic or the big displays in Minority Report
(probably more efficiently though). I just question the need to
t-y-p-e- -o-u-t- -e-v-e-r-y- -d-a-m-n- -c-h-a-r-a-c-t-e-r by hand when
better voice recognition is possible with further engineering.
Part of the advances of the next 30 years will be increasing the input
bandwidth from the user.
> Maybe direct manipulation (stylus on screen) will replace
> mice, but I doubt even that.
Mice suck. Especially with keyboards.
> As for 'infinite, dynamic knowledge spaces'
> instead of fixed windows and desktops--the web is already an
'infinite,
> dynamic knowledge space' but users access it using virtual
rectangular
> pages of text and images (or virtual rectangular view of scrolls, if
you
> prefer). That really hasn't changed since the Gutenberg Bible (or,
hell,
> since the Dead Sea Scrolls), so I really don't expect that
fundamental
> aspect of computer UIs to change.
old physical paradigms will pass. There's no reason eg. the OS desktop
must exactly match the dimensions of the display -- 10.3's Expose is an
interesting yet tentative first step into breaking the 1:1 binding
between information content and display. If Moore's Law at all holds
we'll have the processing power of a roomfull of x-serves in the palm
of our hand in 30 years.
I've seen experiments from MIT's Media Lab about 'fractal' or otherwise
associative display of information. While not necessarily usable they
were certainly thought provoking. I can identify very interesting and
usable improvements that can be made to the newsreader I'm using now.
Better threading, better push/pop stacking of items of interest
("tabbing" of browser window content was one such interesting advance,
something obvious but not implemented until recently).
> > I think in the next 30 we won't be driving cars at all (Minority
> > Report-style automatic driving). Unless of course we're stuck with
the
> > present government leadership more interested in bombing camel
herders
> > than gently directing civilian research efforts.
> >
>
> Such a vision is not possible without an incredibly centralized,
pervasive,
> big-brotherish system. Ugh. And do you envision automatic driving
> everywhere there are roads? Even dirt 2-tracks? And what about
areas where
> there are cyclists, pedestrians, etc?
Dunno. That's what the R&D is for.
> Maybe, MAYBE, well eventually have auto driving on limited-access
'smart'
> highways (that vision has been around for 50 years at least, BTW) but
beyond
> outside that, we'll be driving a vehicle with wheels controlled by
steering
> wheel, brake, and gas pedals (and it'll be called a 'gas' pedal long
after
> there's no more gas in cars). There will be better safety equipment
and
> individual vehicles will have smart collision avoidance equipment
> (ultimately probably mandated), but we'll still be driving.
No need for 'driving' when you are entrained in traffic. The fuel cost
savings of entrained traffic must be significant, and will ultimately
drive out the oldschool independent driver.
This is indeed a remunerative area of investment for our society to
pursue. Perhaps with the cost and time savings we'll have more track
days at race tracks and drag strips.
But who am I kidding. This requires an actual enlightened directed
research government to overcome the local maxima of existing profit
centers. You may be right, we'll be probably be driving gas-powered
cars virtually unchanged from now in 30 years, and wondering why the
Chinese and Indian economies are so far ahead of us in R&D of these
technological areas.
> > > Text and vector graphics could be
> > > 'de-pixelized', but digital cameras are going to be generating
> > pixel-based
> > > images indefinitely. Photoshop is not going to stop dealign with
> > pixels.
> >
> > Pixels are artifacts of low-res displays. Text and pictures in eg.
> > magazines (generally) do not have recognizable pixels.
> >
>
> Even when pixels are small enough not to be individually visible
(which, is
> pretty much the case now), digital photos are still going to
*consist* of
> pixels, and editing is going to involve changing the color value of
large
> numbers of individual pixels.
I can still see pixels at 100 DPI. I've used Apple's new displays and
weren't that impressed with the improvement in readability and WYSIWYG.
With a truly high-density display that there are still picture elements
present is as immaterial as the fact there are carbon atoms laid down
on paper when I use a pencil; just an implementation detail that does
not materially affect the act of creation, unlike the current case
where pixel blockiness at native resolution makes the user experience
of existing mass-market displays substantially inferior to the ease of
use of actual paper-based imaging.
> > I think they assumed they had successfully killed Netscape/Mozilla
and
> > so could move on to other threatening areas of compatibility the
> > non-Windows world was creating. Actually fixing IE to more
accurately
> > conform to eg. CSS standards would be counterproductive to
maintaining
> > market domination and hence was not funded (Micorsoft of course
> > /desires/ people designing websites to its broken CSS model, design
> > effort toward that reduces the ability to develop
platform-agnositic
> > solutions).
> >
>
> Conformance isn't really the issue -- users don't care about that
because
> web designers have to make sure their sites work with IE. The issue
is
> features and, more importantly, security. MS has essentially stopped
> developing standalone versions of IE.
Not only that, they (had) stopped the development of IE en toto, AFAIK.
> So at this point, if you like tabs in
> your browser or the scrolling I talked about, there's no point in
waiting
> for another IE release to add them, because MS has said there isn't
going to
> be one (until you upgrade to a new version of Windows). I do expect
them to
> realize this is a mistake and backtrack, but I wonder how much
momentum
> they'll let Firefox get before making that call.
>From what I've read of Microsoft's blogging (not much) they are
apparently ramping up the browser development again.
> > (I consider this an example of general market failure and not per
se
> > 'evil' on the part of Micorsoft. They're trapped by their own
success,
> > and our take-no-prisoners laissez-faire economic system requires
> > Micorsoft to defends its market position both actively and
passively
> > however it can.)
> >
>
> Well, but thing is that they're NOT defending their browser market
position
> effectively at this point.
Yes. Odd, that. Perhaps there is gain from them freezing the user
experience of browsing until XAML obviates the browser-hosted app
threat.
> And, BTW, if our economic system were centrally managed and less
open, the
> revolutionary changes in technology that have emerged in the last
couple of
> decades probably would have been delayed or even prevented (on the
grounds
> that they might be highly disruptive of established interests--which
in
> fact, they have been).
I'm no socialist, I'm just no free marketeer either. Microsoft is an
interesting case.
> > Too bad newegg doesn't have a DPI category greater than ~100. Odd
> > actually.
> >
>
> AFAIK, nobody sells the high-res 15 or 17 inch displays as standalone
> monitors, only as laptop displays. I don't know why that is.
Unfortunate,
> too. I have a 20" LCD and I'm not sure I ever want a screen larger
than
> that on my desk, but higher-res would be nice.
Yeah, I'd rather have a 17" at 200DPI than a 30" at 100DPI.
> > Nah, you can virtualize it and or otherwise rearchitect it such
that
> > Longhorn apps do it one way and Win32 does it the old way.
> >
>
> Right, but that means Longhorn has to retain the registry for
existing apps.
> Any any new apps developers continue to write to the Win32 standard
(which,
> face it, for years most ISVs will do to get automatic compatibility
for
> people still running XP). Ergo, registry lives on.
Dunno about that. There's ways to advance without breaking stuff.
> > But I think the web has the potential to change things. IMV it has
> > certainly saved the Mac's bacon, like a government pension
providing a
> > fixed but luckily significant amount of life support in an
otherwise
> > Windows world. With web standards compatibility you've got about
30% of
> > what you need a home machine for. The Java API could provide the
other
> > 30%, and perhaps OpenGL and OpenAL would be another 30%, leaving
10%
> > for me to think about as I hit the post button :)
> >
>
> A lot of what computers are used for is still supported by native
code
> calling native APIs. It's certainly not in the interest of MS to
change
> that. And neither Linux nor OS X shows signs of going away from it
either.
Hmm, that's a good point. What would a Java 1.5 install actually call
into on a PS3 ??? Dunno really.
I suppose you'd need Linux underneath, then you'd be set. Ugh. We're
never getting away from the tyranny of POSIX.
Mark Weaver
<imout...@mac.com> wrote in message
> > I'd take that bet. Human brains were built to manipulate things
> > through the use of hands.
>
> oh, we'd still use our hands no doubt, perhaps like either the VR
> scenes in Johnny Mnemonic or the big displays in Minority Report
> (probably more efficiently though). I just question the need to
> t-y-p-e- -o-u-t- -e-v-e-r-y- -d-a-m-n- -c-h-a-r-a-c-t-e-r by hand when
> better voice recognition is possible with further engineering.
>
> Part of the advances of the next 30 years will be increasing the input
> bandwidth from the user.
>
To do that would require genetic engineering, not software engineering ;)
As for voice recognition -- do you realize how ungodly fcking annoying it
would be to have people around you talking to their computers? Can you
imagine sitting next to somebody on an airplane working that way? Or at the
next table in a coffee shop? You'd really have no choice but to strangle
them. No, I think QWERTY keyboards are here to stay.
> > Maybe direct manipulation (stylus on screen) will replace
> > mice, but I doubt even that.
>
> Mice suck. Especially with keyboards.
>
I'll be convinced they suck when somebody offers me something that works
better.
>
> old physical paradigms will pass. There's no reason eg. the OS desktop
> must exactly match the dimensions of the display
No, of course not -- but large virtual desktops and 'multiple desktop
instances' have been available probably 15 years. Hardly anybody uses them.
There's just not much point in figuring out how to lay out way more stuff on
virtual or multiple desktops when it greatly exceeds the human short-term
memory span.
>
> No need for 'driving' when you are entrained in traffic. The fuel cost
> savings of entrained traffic must be significant, and will ultimately
> drive out the oldschool independent driver.
>
A Prius already gets better mileage in city driving than expressway driving.
Think of 'oldschool independent driving' as an incredible distributed
parallel processing system -- each car equipped with an independent image
processing / object recognition supercomputer (human brain).
> This is indeed a remunerative area of investment for our society to
> pursue. Perhaps with the cost and time savings we'll have more track
> days at race tracks and drag strips.
>
> But who am I kidding. This requires an actual enlightened directed
> research government to overcome the local maxima of existing profit
> centers. You may be right, we'll be probably be driving gas-powered
> cars virtually unchanged from now in 30 years, and wondering why the
> Chinese and Indian economies are so far ahead of us in R&D of these
> technological areas.
>
I'm not saying they'll be gas-powered or unchanged. I expect they won't be
gas powered. But they'll still be operated by a driver with a steering
wheel. And you'll still get in them to go do the same sort of things you do
now.
I'm not sure why you're enamored, though, of a vision where the government
designs and imposes a system on everybody. If there's any lesson from the
tech revolution(s) of the last quarter century, it's that decentralized is
the way to go.
Go get yourself a copy of 'Seeing Like A State' and ponder if you *really*
would like a centralized, government-run traffic computerized traffic system
to control all of your travel.
ZnU
"Mark Weaver" <wea...@nospam-corvusdev.com> wrote:
> <imout...@mac.com> wrote in message
[snip]
> > I think in the next 30 we won't be driving cars at all (Minority
> > Report-style automatic driving). Unless of course we're stuck with the
> > present government leadership more interested in bombing camel herders
> > than gently directing civilian research efforts.
> >
>
> Such a vision is not possible without an incredibly centralized, pervasive,
> big-brotherish system. Ugh.
No, that's what it would require if we were to implement it *now*. With
30 years of improvements in e.g. machine vision systems, it might not
require any kind of centralized system. Cars could be sufficiently aware
of their own environments just based on what they could see and possibly
on short-range communication with surrounding cars.
> And do you envision automatic driving everywhere there are roads?
> Even dirt 2-tracks? And what about areas where there are cyclists,
> pedestrians, etc?
If you're imagining blind vehicles hurtling around on pre-specified
paths, yeah, all of those things will present problems. But there's no
reason to think we won't be capable of more in 30 years. In theory,
anyway, a computer can drive a car much more safely than a human -- it
can watch in all directions at once, it can take into account things
like the slipperiness of the road surface or the pressure in the tires,
it's got a much faster reaction time. And communication with surrounding
vehicles could be extremely useful. Imagine, for instance, a boulder
falling into the middle of a road. A computer can talk to the
surrounding vehicles so they can plot avoidance courses that don't
intersect with each other. If a bunch of human-driven cars start
swerving to avoid an obstacle, it's going to require a lot of luck to to
avoid a pileup.
And machine-driven cars would be miraculous for traffic flow patters.
Right of the bad, you'd totally eliminate stop-and-go traffic. I've seen
simulations suggesting you could even get rid of traffic lights at
intersections; if cars can communicate with each other and can adjust
their speeds very precisely, you can simply have perpendicular traffic
streams flow right through each other, with only a small decrease in
average speed.
Now, it's possible none of this will happen, but if it doesn't, I'll bet
it's for social reasons, not technological reasons.
> Maybe, MAYBE, well eventually have auto driving on limited-access 'smart'
> highways (that vision has been around for 50 years at least, BTW) but beyond
> outside that, we'll be driving a vehicle with wheels controlled by steering
> wheel, brake, and gas pedals (and it'll be called a 'gas' pedal long after
> there's no more gas in cars). There will be better safety equipment and
> individual vehicles will have smart collision avoidance equipment
> (ultimately probably mandated), but we'll still be driving.
>
> >
> > > Text and vector graphics could be
> > > 'de-pixelized', but digital cameras are going to be generating
> > pixel-based
> > > images indefinitely. Photoshop is not going to stop dealign with
> > pixels.
> >
> > Pixels are artifacts of low-res displays. Text and pictures in eg.
> > magazines (generally) do not have recognizable pixels.
> >
>
> Even when pixels are small enough not to be individually visible (which, is
> pretty much the case now), digital photos are still going to *consist* of
> pixels, and editing is going to involve changing the color value of large
> numbers of individual pixels.
Pixels right now aren't very visible individually when you aren't
looking for them, but the effects of the relatively low resolution of
present-day displays *are* visible, all the time. Fonts are not nearly
as crisp as they are on paper, for instance. Lines at certain weights
(after being antialiased to display as well as possible on
low-resolution devices) look a bit blurred.
[snip]
> AFAIK, nobody sells the high-res 15 or 17 inch displays as standalone
> monitors, only as laptop displays. I don't know why that is. Unfortunate,
> too. I have a 20" LCD and I'm not sure I ever want a screen larger than
> that on my desk, but higher-res would be nice.
There are companies selling ~200dpi desktop displays, mostly for
specialized industry applications. They tend to cost about four or five
times as much as their lower-resolution counterparts. (Which makes
sense, since they have four or five times as many pixels.)
[snip]
--
"Free societies are hopeful societies. And free societies will be allies against
these hateful few who have no conscience, who kill at the whim of a hat."
--George W. Bush in Washington, D.C., Sept. 17, 2004
ZnU
"Mark Weaver" <wea...@nospam-corvusdev.com> wrote:
> <imout...@mac.com> wrote in message
> > > I'd take that bet. Human brains were built to manipulate things
> > > through the use of hands.
> >
> > oh, we'd still use our hands no doubt, perhaps like either the VR
> > scenes in Johnny Mnemonic or the big displays in Minority Report
> > (probably more efficiently though). I just question the need to
> > t-y-p-e- -o-u-t- -e-v-e-r-y- -d-a-m-n- -c-h-a-r-a-c-t-e-r by hand when
> > better voice recognition is possible with further engineering.
> >
> > Part of the advances of the next 30 years will be increasing the input
> > bandwidth from the user.
> >
>
> To do that would require genetic engineering, not software engineering ;)
>
> As for voice recognition -- do you realize how ungodly fcking annoying it
> would be to have people around you talking to their computers? Can you
> imagine sitting next to somebody on an airplane working that way? Or at the
> next table in a coffee shop? You'd really have no choice but to strangle
> them. No, I think QWERTY keyboards are here to stay.
http://www.findbiometrics.com/viewnews.php?id=925
'NASA Develops System To Computerize Silent, "Subvocal Speech"'
Not nearly good enough for full-language speech recognition, but one day
it might be.
Mark Weaver
"ZnU" <z...@acedsl.com> wrote in message
news:znu-BCAAD8.1...@individual.net...
> In article <psednQ8p6a-...@comcast.com>,
> "Mark Weaver" <wea...@nospam-corvusdev.com> wrote:
>
> > <imout...@mac.com> wrote in message
>
> [snip]
>
> > > I think in the next 30 we won't be driving cars at all (Minority
> > > Report-style automatic driving). Unless of course we're stuck with the
> > > present government leadership more interested in bombing camel herders
> > > than gently directing civilian research efforts.
> > >
> >
> > Such a vision is not possible without an incredibly centralized,
pervasive,
> > big-brotherish system. Ugh.
>
> No, that's what it would require if we were to implement it *now*. With
> 30 years of improvements in e.g. machine vision systems, it might not
> require any kind of centralized system. Cars could be sufficiently aware
> of their own environments just based on what they could see and possibly
> on short-range communication with surrounding cars.
>
> In theory,
> anyway, a computer can drive a car much more safely than a human -- it
> can watch in all directions at once, it can take into account things
> like the slipperiness of the road surface or the pressure in the tires,
> it's got a much faster reaction time.
We are very far from machine vision systems that can process images and
recognize and track objects even as well as a single pair of human eyes.
What 40+ years of AI research has shown (through failure to achieve anything
close) is just how *hard* a problem general-purpose machine vision really
is. It can be done in highly constrained situations (factory automation,
OCR) but we're nowhere near being able to do that out in the world.
>
> And machine-driven cars would be miraculous for traffic flow patters.
> Right of the bad, you'd totally eliminate stop-and-go traffic. I've seen
> simulations suggesting you could even get rid of traffic lights at
> intersections; if cars can communicate with each other and can adjust
> their speeds very precisely, you can simply have perpendicular traffic
> streams flow right through each other, with only a small decrease in
> average speed.
>
As long as there are no 'inconvenient objects' that aren't part of the
communication stream--say pedestrians, bicyclists, stray dogs or children,
deer (not to mention cars whose electronics aren't working properly).
>
> > AFAIK, nobody sells the high-res 15 or 17 inch displays as standalone
> > monitors, only as laptop displays. I don't know why that is.
Unfortunate,
> > too. I have a 20" LCD and I'm not sure I ever want a screen larger than
> > that on my desk, but higher-res would be nice.
>
> There are companies selling ~200dpi desktop displays, mostly for
> specialized industry applications. They tend to cost about four or five
> times as much as their lower-resolution counterparts. (Which makes
> sense, since they have four or five times as many pixels.)
>
Right, but I'm talking about displays that have resolutions that are 33 to
50 percent higher res-- 15 inch and 17 inch displays with SXGA+, UXGA,
WSXGA+, or WUXGA res which *are* readily available on notebooks (and for a
not a lot more money--$50 or $100 extra) but for some reason nobody puts
these panels in a desktop monitor.
imout...@mac.com
> <imout...@mac.com> wrote in message
> > > I'd take that bet. Human brains were built to manipulate things
> > > through the use of hands.
> >
> > oh, we'd still use our hands no doubt, perhaps like either the VR
> > scenes in Johnny Mnemonic or the big displays in Minority Report
> > (probably more efficiently though). I just question the need to
> > t-y-p-e- -o-u-t- -e-v-e-r-y- -d-a-m-n- -c-h-a-r-a-c-t-e-r by hand
when
> > better voice recognition is possible with further engineering.
> >
> > Part of the advances of the next 30 years will be increasing the
input
> > bandwidth from the user.
> >
>
> To do that would require genetic engineering, not software
engineering ;)
Nah, computers will be integrated into us sooner or later.
> As for voice recognition -- do you realize how ungodly fcking
annoying it
> would be to have people around you talking to their computers? Can
you
> imagine sitting next to somebody on an airplane working that way? Or
at the
> next table in a coffee shop? You'd really have no choice but to
strangle
> them. No, I think QWERTY keyboards are here to stay.
:)
compare and contrast:
http://groups-beta.google.com/group/comp.sys.next/msg/88d3ecfc2110c44a
vs.
http://groups-beta.google.com/group/comp.sys.next/msg/70795b211b671e04
not sayin' I'm as prescient as Salinas there, but methinks you are too
stuck in the past & present.
As for the annoyance of voice recognition, think about the cues a
computer needs from a user to pick up natural language input.... sorta
mumble to yourself what you want to type, really really softly, almost
inaudibly to yourself. That is information the computer can
theoretically attend to and process into text.
Today we get a really crappy hit rate when computers try this, but we
really haven't even begun to tackle Natural Language Processing. NLP is
going to be the next www, we just can't see it yet.
> > > Maybe direct manipulation (stylus on screen) will replace
> > > mice, but I doubt even that.
> >
> > Mice suck. Especially with keyboards.
> >
>
> I'll be convinced they suck when somebody offers me something that
works
> better.
In 5 years? No way. 50? I wouldn't bet against it. 30? Probably pushing
it (extrapolating from present trends), but anything more than 10 years
out is really the Great Unknown as far as information processing goes.
Clearly the first step is getting computers able to parse static text
into semantic meanings. I just read something on the web about a news
summarizing robot. That's progress apparently, and without IT
understanding the (abstact) domain of natural language we're not going
to have any progress in real-world speech recognition.
> > old physical paradigms will pass. There's no reason eg. the OS
desktop
> > must exactly match the dimensions of the display
>
> No, of course not -- but large virtual desktops and 'multiple desktop
> instances' have been available probably 15 years. Hardly anybody
uses them.
> There's just not much point in figuring out how to lay out way more
stuff on
> virtual or multiple desktops when it greatly exceeds the human
short-term
> memory span.
People have excellent spatial memories. I don't see why this won't
transfer into the IT domain. The issue is moving the viewport of the
display around the world. I'm not particularly enamored with the
overlapping window ratsnest and think we can do better.
> > No need for 'driving' when you are entrained in traffic. The fuel
cost
> > savings of entrained traffic must be significant, and will
ultimately
> > drive out the oldschool independent driver.
> >
>
> A Prius already gets better mileage in city driving than expressway
driving.
As anyone familar with basic physics (and the Prius) will attest, the
EPA city test is bogus (it "tests" unrecharged battery power, not gas
consumption per se, for the 11 mile city loop).
> Think of 'oldschool independent driving' as an incredible distributed
> parallel processing system -- each car equipped with an independent
image
> processing / object recognition supercomputer (human brain).
That has to stop, idle and restart at *least* once or twice a mile in
the city. Regenerative braking is not 100% efficient, plus there's
capital cost and weight involved in carrying it... how much better just
not have to stop at all! And to be honest, instead of spending $250B+
trying to control Iraq's oil we could have invested this in actual IP
and infrastructure, not to mention high-tech jobs of solving the
pressing societal issue of traffic safety and energy efficiency.
Man, $250B was 25,000 engineers at $500k/head working for 20 years.
Sigh.
> > This is indeed a remunerative area of investment for our society to
> > pursue. Perhaps with the cost and time savings we'll have more
track
> > days at race tracks and drag strips.
> >
> > But who am I kidding. This requires an actual enlightened directed
> > research government to overcome the local maxima of existing profit
> > centers. You may be right, we'll be probably be driving gas-powered
> > cars virtually unchanged from now in 30 years, and wondering why
the
> > Chinese and Indian economies are so far ahead of us in R&D of these
> > technological areas.
> >
>
> I'm not saying they'll be gas-powered or unchanged. I expect they
won't be
> gas powered. But they'll still be operated by a driver with a
steering
> wheel. And you'll still get in them to go do the same sort of things
you do
> now.
>
> I'm not sure why you're enamored, though, of a vision where the
government
> designs and imposes a system on everybody.
http://www.fhwa.dot.gov/hep10/images/nhsjpg.jpg
http://science.ksc.nasa.gov/history/apollo/images/apollo.gif
http://tokyoyakei.halfmoon.jp/tokyo/railway/shinkansen/shinkansen-10.jpg
http://www.computerhistory.org/exhibits/internet_history/full_size_images/1983_topo.gif
> If there's any lesson from the
> tech revolution(s) of the last quarter century, it's that
decentralized is
> the way to go.
If there's any lesson from the past 100 years, it's free enterprise
needs a helpful collective push from above now and then to overcome
organizational inertia and accomplish anything interesting.
Don't get me wrong, command economies do indeed suck, and it takes a
government to really screw things up but good (like Brasilia). But the
fairy tales about daring entrepreneurs scaling new heights of
innovation is mostly poppycock since all these folks stood on the
shoulders of government spending in one way or another. Free enterprise
is good for getting cheap stuff at Walmart, and otherwise productizing
what already exists into cheaper and/or crappier offerings, but for the
good stuff you've got to pick up trickle-out tech that was sponsored by
government spending and/or policy, spending that supports nascent stuff
that may or may not pan out.
> Go get yourself a copy of 'Seeing Like A State' and ponder if you
*really*
> would like a centralized, government-run traffic computerized traffic
system
> to control all of your travel.
I'm a lefty-libertarian so I am sympathetic to the preference of
individual liberty and distrust of centralized decision-making. But the
lefty side of me is equally distrustful of centralized CORPORATE
decision-making and market failure. A balance must be struck between
democratic socialism and free enterprise; both are necessary but
neither is sufficient alone to maximize social gain in the real world.
IOW the market is guaranteed to latch onto local maxima, but it takes
social policy to shift to greater maxima (and/or fall into minima if it
is stupid).
Sandman
<wea...@nospam-corvusdev.com> wrote:
>> oh, we'd still use our hands no doubt, perhaps like either the VR
>> scenes in Johnny Mnemonic or the big displays in Minority Report
>> (probably more efficiently though). I just question the need to
>> t-y-p-e- -o-u-t- -e-v-e-r-y- -d-a-m-n- -c-h-a-r-a-c-t-e-r by hand
>> when better voice recognition is possible with further engineering.
>>
>> Part of the advances of the next 30 years will be increasing the
>> input bandwidth from the user.
>
> To do that would require genetic engineering, not software engineering
> ;)
>
> As for voice recognition -- do you realize how ungodly fcking annoying
> it would be to have people around you talking to their computers? Can
> you imagine sitting next to somebody on an airplane working that way?
> Or at the next table in a coffee shop? You'd really have no choice
> but to strangle them. No, I think QWERTY keyboards are here to stay.
This is something that's often brought up when we are discussing voice
recognition. I don't really understand it.
It would seem that this kind off logic is based on the idea that voice
recognition would in the end result in a situation where you loudly give your
computer specific commands, such as:
"Computer, open Word"
"Computer, open the document 'Budget' in the folder 'Annual Report' on
my system drive"
"Computer, send mail to bob at apple dot com - Hello Bob - comma - how
are you - question mark - I wanted to ask you about the low oustanding
amount - period"
That would truly suck! That's like spoken AppleScript. :)
No, voice recognition wouldn't be about recognizing literal voice commands one
by one. Sitting next to a person on the plane doing the above would be annoying
indeed. But imagine the same person on his cell phone, talking to his secretary:
"John, could you check the numbers for the last budget? Yeah, check the
outstanding amount... Ok... Could you send a mail to Bob and ask him why
it's so low? Thanks".
Now THAT'S voice recognition! John understood everything without having to have
it spelt out command by command, and that's how computers should evolve. The
guy on the plane should be saying something like this to his computer:
"Open last budget" - *checks the budget on screen* - "Mail Bob and ask
why the outstanding amount is so low"
And the computer would understand that.
--
Sandman[.net]
ZnU
"Mark Weaver" <wea...@nospam-corvusdev.com> wrote:
You're not quite being fair. We haven't, over the last 40 years, had the
kind of processing power we've got now. With generalized AI, of course,
it seems pretty clear that the problem is not just processing power, but
that we don't have much of a clue where to begin conceptually either. In
the specific case of machine vision, however, that doesn't seem to be
quite as true.
And keep in mind that there's no reason self-driving cars can't "cheat".
There's no need for them to be able to build up a conception of a 3D
environment from 2D images -- or even stereoscopic pairs of 2D images.
You can just stick radar on them.
I wouldn't be surprised if the current state-of-the art was more than
good enough to follow roads and avoid obstacles on them. Sharing the
road with non-robotic traffic and understanding traffic signals are
probably not practical yet.
> > And machine-driven cars would be miraculous for traffic flow patters.
> > Right of the bad, you'd totally eliminate stop-and-go traffic. I've seen
> > simulations suggesting you could even get rid of traffic lights at
> > intersections; if cars can communicate with each other and can adjust
> > their speeds very precisely, you can simply have perpendicular traffic
> > streams flow right through each other, with only a small decrease in
> > average speed.
> >
>
> As long as there are no 'inconvenient objects' that aren't part of the
> communication stream--say pedestrians, bicyclists, stray dogs or children,
> deer (not to mention cars whose electronics aren't working properly).
Simply avoiding things in the road is a fairly trivial problem. Even a
rather dim computer program can figure out if there's something in the
way. It might not be able to figure out whether it's a plastic bag or a
dog, however.
> > > AFAIK, nobody sells the high-res 15 or 17 inch displays as standalone
> > > monitors, only as laptop displays. I don't know why that is.
> Unfortunate,
> > > too. I have a 20" LCD and I'm not sure I ever want a screen larger than
> > > that on my desk, but higher-res would be nice.
> >
> > There are companies selling ~200dpi desktop displays, mostly for
> > specialized industry applications. They tend to cost about four or five
> > times as much as their lower-resolution counterparts. (Which makes
> > sense, since they have four or five times as many pixels.)
> >
>
> Right, but I'm talking about displays that have resolutions that are 33 to
> 50 percent higher res-- 15 inch and 17 inch displays with SXGA+, UXGA,
> WSXGA+, or WUXGA res which *are* readily available on notebooks (and for a
> not a lot more money--$50 or $100 extra) but for some reason nobody puts
> these panels in a desktop monitor.
--
ed
Sandman <m...@sandman.net> typed:<snip>
>> Can
>> you imagine sitting next to somebody on an airplane working that way?
>> Or at the next table in a coffee shop? You'd really have no choice
>> but to strangle them. No, I think QWERTY keyboards are here to stay.
>
> This is something that's often brought up when we are discussing voice
> recognition. I don't really understand it.
>
> It would seem that this kind off logic is based on the idea that voice
> recognition would in the end result in a situation where you loudly
> give your computer specific commands, such as:
>
> "Computer, open Word"
> "Computer, open the document 'Budget' in the folder 'Annual
> Report' on my system drive"
> "Computer, send mail to bob at apple dot com - Hello Bob - comma
> - how are you - question mark - I wanted to ask you about the
> low oustanding amount - period"
>
> That would truly suck! That's like spoken AppleScript. :)
>
> No, voice recognition wouldn't be about recognizing literal voice
> commands one by one. Sitting next to a person on the plane doing the
> above would be annoying indeed. But imagine the same person on his
> cell phone, talking to his secretary:
>
> "John, could you check the numbers for the last budget? Yeah,
> check the outstanding amount... Ok... Could you send a mail to
> Bob and ask him why it's so low? Thanks".
>
> Now THAT'S voice recognition!
THAT'S annoying too! especially when there's a whole plane full of folks
doing it!
<snip>
Sandman
"ed" <ne...@no-atwistedweb-spam.com> wrote:
There -won't- be a plane full of people doing this. There aren't even planes
full of people with laptops today. Not everyone has, and probably won't have
the need to check their annual budget for a full eight hour plane trip.
--
Sandman[.net]
Mark Weaver
"Sandman" <m...@sandman.net> wrote in message news:mr-
> > > Now THAT'S voice recognition!
> >
> > THAT'S annoying too! especially when there's a whole plane full of
folks
> > doing it!
>
> There -won't- be a plane full of people doing this. There aren't even
planes
> full of people with laptops today. Not everyone has, and probably won't
have
> the need to check their annual budget for a full eight hour plane trip.
>
Well, where I live there are coffee shops full of students (quietly) doing
their homework--if they all had headsets on and were talking to their
laptops, it'd be pretty bad. It'd also be pretty bad in a cubicle office if
everyone was sitting jabbering away. If voice recognition is every
perfected, it'll certainly be banned in offices, restaurants, and on planes
shortly thereafter ;)
Mark Weaver
> kind of processing power we've got now. With generalized AI, of course,
> it seems pretty clear that the problem is not just processing power, but
> that we don't have much of a clue where to begin conceptually either. In
> the specific case of machine vision, however, that doesn't seem to be
> quite as true.
>
> And keep in mind that there's no reason self-driving cars can't "cheat".
> There's no need for them to be able to build up a conception of a 3D
> environment from 2D images -- or even stereoscopic pairs of 2D images.
> You can just stick radar on them.
>
> I wouldn't be surprised if the current state-of-the art was more than
> good enough to follow roads and avoid obstacles on them. Sharing the
> road with non-robotic traffic and understanding traffic signals are
> probably not practical yet.
>
No, to be as good as a human, a robotic car has to notice the kids playing
near the road and quickly get a sense of how old they are and determine if
they're paying attention and have noticed the approaching car. It has be be
able to recognize a person with a white cane and realize what that means.
And 'pretty close most of the time' isn't good enough. Even if the failure
rate was no higher than that of human drivers, it'd still be a liability
nightmare -- the driver would no longer be responsible for the accident, the
car manufacturer / AI driving system vendor would be.
>
> Simply avoiding things in the road is a fairly trivial problem. Even a
> rather dim computer program can figure out if there's something in the
> way. It might not be able to figure out whether it's a plastic bag or a
> dog, however.
>
And if there are objects near the road (moving or not), it has to figure out
what they are and determine whether or not there's a chance of that object
moving into the car's path. A stationary deer near the road is a *potential
ly fatal* hazard. A stationary bush near the road is not.
Just following roads and even avoiding stationary obstacles is nowhere close
to good enough. This is not one of those situations where the 80/20 rule
applies ;)
Mark Weaver
>
> compare and contrast:
>
> http://groups-beta.google.com/group/comp.sys.next/msg/88d3ecfc2110c44a
>
> vs.
>
> http://groups-beta.google.com/group/comp.sys.next/msg/70795b211b671e04
>
> not sayin' I'm as prescient as Salinas there, but methinks you are too
> stuck in the past & present.
>
Except that what Salinas was talking about did not require any revolution in
PC software--just HTML and a browser (both pretty damn simple software
technologies--especially in their original incarnations). By 1989, the
network part of the puzzle had already been in place for years.
The fact that big changes occurred early in the evolution of personal
computers doesn't mean that we should expect them to continue at the same
pace. Technologies mature, there get to be a lot of inertia and investment
in the status quo, and the pace of change slows.
> As for the annoyance of voice recognition, think about the cues a
> computer needs from a user to pick up natural language input.... sorta
> mumble to yourself what you want to type, really really softly, almost
> inaudibly to yourself. That is information the computer can
> theoretically attend to and process into text.
>
> Today we get a really crappy hit rate when computers try this, but we
> really haven't even begun to tackle Natural Language Processing. NLP is
> going to be the next www, we just can't see it yet.
>
It's not that I doubt that sub-vocal speech recognition is possible (though
I am skeptical), it's that I wouldn't *want* it. Think about how you
interact with a text editor. You don't generate a continuous stream of text
(or at least I don't). I generate a bit, then move back and erase something
or insert something or move it around--and a keyboard and mouse are much
better for that than speech.
>
> In 5 years? No way. 50? I wouldn't bet against it. 30? Probably pushing
> it (extrapolating from present trends), but anything more than 10 years
> out is really the Great Unknown as far as information processing goes.
It really isn't though. In a couple of days it'll be 2005. OK, go back to
1995. How different, *really* are things now than in 1995? And I think
it's a
pretty fair guess that things will change less dramatically in the next 10
years than in the last 10.
> Clearly the first step is getting computers able to parse static text
> into semantic meanings. I just read something on the web about a news
> summarizing robot. That's progress apparently, and without IT
> understanding the (abstract) domain of natural language we're not going
> to have any progress in real-world speech recognition.
>
You really can't do full-blown natural language processing or speech
recognition without full-blown human-level artificial intelligence.
Unfortunately, there's no syntactic-oriented shortcut possible -- to process
text or speech in a sophisticated way requires actually *understanding* it.
>
> People have excellent spatial memories. I don't see why this won't
> transfer into the IT domain.
>
People have excellent *long term* spatial memories, but windows on a desktop
deals with short-term configurations of short-term instances. And, in any
case, I'd rather let the computer help me in pulling back up what I want to
work on rather than relying on my spatial memory of where it resides. So
fast, smart searches to find documents rather than virtual 'places' to put
them and a task bar to select active processes from rather than virtual
places to put them.
> The issue is moving the viewport of the
> display around the world. I'm not particularly enamored with the
> overlapping window ratsnest and think we can do better.
>
But as it is, there's no reason to leave all your active windows all piled
up on the
screen. Minimize the ones you don't need at the moment. Simple, not sexy,
but effective.
>
> As anyone familar with basic physics (and the Prius) will attest, the
> EPA city test is bogus (it "tests" unrecharged battery power, not gas
> consumption per se, for the 11 mile city loop).
>
The point is that regenerative braking and shutting down the engine at idle
have reduced the penalty of stop-and-go driving. And high-speed highway
driving has its own inefficiencies (increased drag).
> And to be honest, instead of spending $250B+
> trying to control Iraq's oil we could have invested this in actual IP
> and infrastructure, not to mention high-tech jobs of solving the
> pressing societal issue of traffic safety and energy efficiency.
>
> Man, $250B was 25,000 engineers at $500k/head working for 20 years.
> Sigh.
>
Oh, man, the government investing hundreds of billions to design a road
control system to guide all our travel -- you really are serious, aren't
you? As I said, get yourself a copy of 'Seeing Like A State'.
>
> http://www.fhwa.dot.gov/hep10/images/nhsjpg.jpg
> http://science.ksc.nasa.gov/history/apollo/images/apollo.gif
> http://tokyoyakei.halfmoon.jp/tokyo/railway/shinkansen/shinkansen-10.jpg
>
http://www.computerhistory.org/exhibits/internet_history/full_size_images/1983_topo.gif
>
> > If there's any lesson from the
> > tech revolution(s) of the last quarter century, it's that
> decentralized is
> > the way to go.
>
> If there's any lesson from the past 100 years, it's free enterprise
> needs a helpful collective push from above now and then to overcome
> organizational inertia and accomplish anything interesting.
>
The highway system and DARPA net are good examples of decentralization. The
government builds some basic infrastructure, but it doesn't design the
vehicles that run on the highway any more than it designed the applications
that run on the internet. The government did not plan (and actually had no
clue) what the DARPA net was going to become when it morphed into the
Internet--and that's a good thing, because if it had set out to do it
intentionally what we'd have ended up with was something like France's
Minitel.
I'm not saying the government has no role to play--interstate highways
obviously could not have been built without eminent domain (though they
*could* have been constructed without tax dollars -- as toll sections of the
system were). Government launch GPS satellites, good. Government design an
auto-driver system to be installed in every car, bad.
> Free enterprise
> is good for getting cheap stuff at Walmart, and otherwise productizing
> what already exists into cheaper and/or crappier offerings, but for the
> good stuff you've got to pick up trickle-out tech that was sponsored by
> government spending and/or policy, spending that supports nascent stuff
> that may or may not pan out.
>
Off the top of my head, here's some good stuff:
Powered flight -- the Wrights, not a government-initiated invention.
Television -- Philo T Farnsworth, not a government-initiated invention.
Automobiles -- ditto.
Photography, sound recordings (analog and digital), motion pictures,
electric lights, telephone -- ditto.
Video and digital cameras -- ditto.
Velcro, polartec, fiberglass composites -- ditto.
In many cases, the government did see a use for these emerging technologies
and became a major customer, and contributed to development in that way, but
it didn't take the government to get these things started. Nor was the
government purchasing crucial. Development may have been accelerated
because of government spending (usually defense spending) but passenger
aircraft, for example, were being steadily developed even before the big
WWII defense buildup.
>
> IOW the market is guaranteed to latch onto local maxima, but it takes
> social policy to shift to greater maxima (and/or fall into minima if it
> is stupid).
>
No, it clearly does not take social policy to take risks. There are huge
rewards to be had for those who invent and produce new, advanced, disruptive
technologies. Which is why corporations invest large amounts in R&D.
Probably Kodak and Fuji would've loved to milk their cash- cow film, paper &
chemical business indefinitely, but the development of digital cameras made
that impossible. And ironically, Kodak itself made a number of the main
early advances in digital imagery. But they couldn't just stick with their
happy local minimum, because if they hadn't done it, somebody else would
have.
Sandman
<wea...@nospam-corvusdev.com> wrote:
>>>> Now THAT'S voice recognition!
>>>
>>> THAT'S annoying too! especially when there's a whole plane full of
>>> folks doing it!
>>
>> There -won't- be a plane full of people doing this. There aren't even
>> planes full of people with laptops today. Not everyone has, and
>> probably won't have the need to check their annual budget for a full
>> eight hour plane trip.
>
> Well, where I live there are coffee shops full of students (quietly)
> doing their homework--if they all had headsets on and were talking to
> their laptops, it'd be pretty bad.
Yeah, but what they are doing isn't something that could be replaced by a
secretary, or personal assistant, right? You need to think about voice
recognition in that way. Obviously writing a full report isn't a very good task
of voice recognition, since you usually type faster than you speak - or rather,
you usually type at the speed your mind creates the sentences. I think the
keyboard will still be here for those moments.
> It'd also be pretty bad in a
> cubicle office if everyone was sitting jabbering away.
They are already talking on the phone in cubicle offices without that being a
huge problem.
> If voice
> recognition is every perfected, it'll certainly be banned in offices,
> restaurants, and on planes shortly thereafter ;)
I highly doubt it, based on my premise on how it should work.
--
Sandman[.net]
ZnU
"Mark Weaver" <wea...@nospam-corvusdev.com> wrote:
> "ZnU" <z...@acedsl.com> wrote in message news:znu-
> >
> > You're not quite being fair. We haven't, over the last 40 years, had the
> > kind of processing power we've got now. With generalized AI, of course,
> > it seems pretty clear that the problem is not just processing power, but
> > that we don't have much of a clue where to begin conceptually either. In
> > the specific case of machine vision, however, that doesn't seem to be
> > quite as true.
> >
> > And keep in mind that there's no reason self-driving cars can't "cheat".
> > There's no need for them to be able to build up a conception of a 3D
> > environment from 2D images -- or even stereoscopic pairs of 2D images.
> > You can just stick radar on them.
> >
> > I wouldn't be surprised if the current state-of-the art was more than
> > good enough to follow roads and avoid obstacles on them. Sharing the
> > road with non-robotic traffic and understanding traffic signals are
> > probably not practical yet.
> >
>
> No, to be as good as a human, a robotic car has to notice the kids playing
> near the road and quickly get a sense of how old they are and determine if
> they're paying attention and have noticed the approaching car. It has be be
> able to recognize a person with a white cane and realize what that means.
A robotic system might be able to make up for its lack of understanding
of contextual cues with faster and safer reactions. If a human on a busy
road has to swerve at the last instant, that's something very dangerous.
If a robotic car has to do it, it's probably pretty safe. The software
can know exactly how far it can push the car before it rolls over,
exactly how much traction it can get on a given road surface, and it can
communicate with surrounding robotic cars to avoid collisions.
> And 'pretty close most of the time' isn't good enough. Even if the failure
> rate was no higher than that of human drivers, it'd still be a liability
> nightmare -- the driver would no longer be responsible for the accident, the
> car manufacturer / AI driving system vendor would be.
If you're going to absolute demand perfection, you're never going to get
it. I acknowledged previously there could be social and political issues.
I'd like to think people would be a bit more reasonable than that.
Humans are not particularly safe drivers.
> > Simply avoiding things in the road is a fairly trivial problem. Even a
> > rather dim computer program can figure out if there's something in the
> > way. It might not be able to figure out whether it's a plastic bag or a
> > dog, however.
> >
> And if there are objects near the road (moving or not), it has to figure out
> what they are and determine whether or not there's a chance of that object
> moving into the car's path. A stationary deer near the road is a *potential
> ly fatal* hazard. A stationary bush near the road is not.
Recognizing immobile objects by the road which could become mobile is
probably one of the trickier tasks -- but again, you can cheat. Infrared
should be able to spot roadside animals fairly easily, and a bit of
image processing could probably tell animals from humans.
> Just following roads and even avoiding stationary obstacles is nowhere close
> to good enough. This is not one of those situations where the 80/20 rule
> applies ;)
--
Is Bush wearing a LifeVest defibrillator?
http://nyc.indymedia.org/newswire/display/136872/