56K analog modem status 10-Dec-96 - LONG
Frank Durda IV
Here is the latest information on the various schemes for 56K
transmission over analog lines. These descriptions are in no
particular order.
------------------
US Robotics X2(TM)
If you ignore AT&T discussing analog speeds at 56K nearly two years
ago, USR was the first to announce a 56K solution this year which
they call X2. This is a USR tradename and only describes the USR
56K solution. Other 56K solutions (or schemes) have their own
tradenames.
X2 requires one modem to be an analog modem with X2 code, and the
other modem must be a digital X2 modem connected to a T1 or PRI digital
line. In addition, the telephone link between the two modems must
not undergo more than one analog-to-digital conversion. If there
is more than one A-D conversion, 56K won't work.
So far, USR hasn't allowed any independent testing of their X2
scheme.
The reverse channel of an X2 modem runs at a maximum of 33.6Kbps
and can be lower depending on line quality. If you aren't getting
33.6 now, it certainly won't be the reverse channel speed you get
with X2 running.
USR currently claims to have new software for V.Courier modems
available at NO charge (if you register your interest by 31-Dec-96).
The code is supposed to be available in December for downloading.
Sportster owners who purchased modems after August 15th are supposed
to have modems that are upgradable. The actual modem may be identical
to one bought on the 14th, but USR has set an arbitrary date. Save
your sales receipt!
Then USR announced that people who purchased Sportsters after
November 25th would receive a free upgrade. No one has come out and
said exactly what the cost is to people with modems bought between the
August date and the November date. This has all changed several times,
so stay tuned.
USR says that when an international standards body adopts a 56K scheme,
USR will offer new code for their modems to support it. For
Sportster owners, this could mean another trip for the modem back
to the factory, or having to install another ROM at home. The latest
Sportster models (the 8th digit of the serial number on the box and
modem is a "3") have a socketed ROM, so you can change the part yourself.
USR X2 will appear in USR and TI modems. TI sells re-labeled
Sportster modems to Gateway 2000 and other PC makers. In addition,
Cardinal, who normally uses Rockwell parts has announced they will
support the X2 56K scheme. No other modem maker has announced that
they will support X2, and USR is currently the only provider of the
digital modems that ISPs and online service providers must own to
perform the X2 56K system. The digital USR modems are either the Total
Control Racks (TCRs) or Courier-I modems.
--------------------
Rockwell K56plus(TM)
Announced shortly after the USR proposal, K56plus also provides up
to 56K from a digital modem located at the ISP to an analog modem.
Unlike the USR proposal, K56plus supposedly can exceed 33.6Kbps
in the other direction. The precise return-channel speed is still
"not final" but won't be anywhere near the down-link speed.
As with the USR proposal, the Rockwell proposal has only been seen
in the confines of the trade show booth, and not under real life
conditions.
K56plus also requires a digital modem on the other end of the line
to do the 56K speed and there can only be one A-D conversion in the
telephone system between the two modems.
Because Rockwell has until now used hard-coded "datapumps" in their
modems, modems based on Rockwell chips will have to have both the
firmware replace and the datapump chip replaced to be able to
perform K56plus. Rockwell is at a time disadvantage since all modems
based on their design require hardware changes and consumer class
units are not expected on the market until March or April 1997.
Rockwell has seen the light on having the ability to change the
algorithms in their datapumps, and starting in 1997, Rockwell
will begin to make datapumps that can have their firmware changed
in the field. Currently USR and one or two high-end modems from
vendors like AT&T Paradyne and Sierra are the only modems that
have DSPs that can use downloadable code.
A lot of existing Rockwell modems will not upgrade to 56K, because
the datapump is soldered into the modem, and cannot economically
be replaced.
There is currently no upgrade program from Rockwell, but Motorola is
offering to even replace modems with non-Motorola brands if their modems
can't be made to do the eventual 56K standard.
K56plus is incompatible with the USR X2 proposal, and vice versa.
Rockwell says they will support any 56K that becomes the official
standard in addition to their K56plus scheme, when a standard appears.
Although Rockwell hasn't said this, but the OEMs that use their chips
have: if the X2 scheme by some chance becomes the overwhelming popular
56K system, Rockwell-based modems will support X2. Rockwell
publicly says they don't think X2 would become standard, but what
would you expect them to say?
Rockwell chips appear in Hayes, Microcom, Supra, Boca, GVC,
Practical Peripherals, Motorola, Cisco, Ascend, Shiva, Cascade
and numerous other brands of modems and digital modem systems.
Rockwell doesn't really make many finished modems, just the chipsets.
---------------------------------------------------
Lucent (Formally AT&T Microelectronics) V.flex2(R)
Although they announced later than USR and Rockwell, Lucent marketing
people apparently knew 56K speeds were coming far enough in advance
to have time to get a registered trademark (R) name for their 56K product
instead of using a simple trademark (TM) like the other vendors.
V.flex2 also offers 56K in one direction, and an unspecified speed
"in excess of 33.6" in the opposite direction under ideal conditions,
which is similar to what Rockwell is saying for their proposal.
V.flex2 also requires an analog and a digital modem in the conversation,
and limits the phone link to a single A-D conversion for 56K to work.
Despite the "V." on the Lucent product name, V.flex2 is not a
standard, just a marketing name.
Both the Rockwell and USR schemes are incompatible with the Lucent 56K
scheme.
However, shortly after the original Lucent announcement, Rockwell and
Lucent agreed that their 56K modems would eventually detect and talk to
the other manufacturers (Rockwell or Lucent) 56K scheme if the native 56K
scheme wasn't present in the other modem. This at least offers the
first hope of cross-compatibility between two of the many 56K proposals.
V.flex2 is incompatible with USR X2 proposal, and vice versa.
Lucent says chipsets supporting V.flex2/K56Plus schemes will be
available in the first quarter of 1997. Currently, Paradyne
is the only known vendor who has indicated that they will use these
chips. Since the Lucent proposal will eventually also support K56Plus,
it is likely that Boca and some of the other vendors following Rockwell
56K proposal might use the Lucent chips. This second source might
help drive prices down.
A final note. It appears that all of the 56K solutions, including
X2 and K56Plus, are based on patents AT&T recently received. This
makes them a stronger player in the game than you might have expected
and probably explains the buddy-buddy arrangement reached with
Rockwell.
---------------------------
Aetherworks v.Mach(TM) 43.2
This is the dark horse vendor of the vendors who have announced
56K solutions. The v.Mach solution limits transmissions to 43.2Kbit/sec,
but offers symmetrical transmissions, with 43.2Kbit in both directions.
This differs from the above 56K solutions that only offer 56K in
one direction and one of the V.34 speeds in the return direction.
Another interesting twist of v.Mach, is that it claims to eliminate
the need for a digital modem at one end of the conversation, and not
necessarily require there to be only one analog to digital conversion
in the telephone link. If both of these claims are true, v.Mach would
have a far better chance of working in a given location than any of
the 56K schemes announced so far.
I am not aware of any public showings of this system. It was announced
after Comdex ended.
As with the Lucent proposal, the "v." in the v.Mach name is not
a standard, just a marketing name.
----------------------------------------------
Brand "C", Brand "D" and Brand "N" 56K schemes
These are three vendors who have been privately discussing their
own implementations of a 56K transmission scheme, but so far have
not announced products based on their work, and to my knowledge, none
have actually shown working hardware outside of the lab.
As far as is known, all three of these schemes are incompatible
with one another, and with the other four schemes mentioned above.
-----------------------------------
What the Standards people are doing
These mystery firms may miss their chance if they don't speak up soon,
as USR, Rockwell and Lucent have all asked the EIA/TIA and the ITU
to consider their proposals to create a standard for 56K transmission.
As of this writing, the EIA has created a study group, and the ITU
has discussed setting one up during a recent lunch in England, but
hasn't formally created a group at this time.
An important note: It is extremely unlikely that the EIA/TIA or
the ITU would approve any one of the proposed transmission schemes
without making some changes to it, meaning the offical standard
won't be X2, K56Plus or V.flex2. It may be a hybrid of all three
or something completely out of left field.
Also, the EIA is far more likely to come out with a standard during
1997 than the ITU is. The ITU just doesn't move that fast for
many reasons.
All of the announced vendors have web sites that talk about their
plans. Rockwell even has technical papers on their implentation.
USR has a list of Courier models that they will have firmware
updates for, along with the forms for ordering Sportster 56K updates.
--------------------------------------
What does all of this mean to the ISP?
Well, it means a lot of vendors keep showing up trying to get you
to put your name on their list as endorsing or supporting a given scheme,
sight unseen. Getting put on a list is usually non-binding, unless you
have actually ordered hardware. But it can confuse (and anger) your
customers if an ISP says they will support X but eventually
purchase Y instead.
The ISP has enormous expense to consider for 56K. A digital modem unit
with 48 digital modems which connect to two T1 or PRI lines can
cost between $30,000 and $70,000 depending on which manufacturer and
whether you get any quantity price break. Assuming $40K for 48 lines,
that works out to $833 per modem, which would buy a lot of analog modems.
Plus, that T1 or PRI is going to cost several hundred dollars per
month EACH, with some prices as high as $50 per line. That's a
lot of analog phone lines.
Several of the ISPs and online service providers across the country
(including one smaller ISP in North Texas) have indicated that they
plan to charge extra for accounts to access the 56K analog service
when they finally offer it. The additional cost will help offset
the costs of the equipment required to offer it and the extra
bandwidth needed in their plant compared to 33.6 modems. (I don't
know the plans of any of the major providers in North Texas on this
point.)
Plus, what if the ISP guesses wrong, and the digital modems they
buy turn out to be the ones that don't become the official standard
(likely), and what if the hardware has to be replaced or exchanged
to be able to do the official standard? That $40,000 investment
could lose a lot of its value in just six or nine months and NOBODY
will want to buy a second-hand modem that won't speak the official
standard.
Another thing the ISP has to think about is how many of his customers
have which type of 56K modem, and how many of them can actually
make 56K connections from where they are located even if the ISP
does buy the compatible hardware. The ISPs don't know this answer.
The authors of the 56K proposals don't know, although Rockwell says
less than 6% of the phone lines in the USA can do 28,800bps and the
number that can do their 56K scheme will be less. As Mr. Eastwood
would say, "Do you feel lucky?"
With less than a 6% chance of success, your luck better be good.
Any ISP thinking about 56K analog should be paranoid about buying many
expensive boxes that potentially no one can actually talk to. That is
why most ISPs are waiting until they can actually touch and feel both
the analog and the digital 56K modems, and can try them from various
places in their service areas. If it won't even work across the street
from the ISP, then perhaps Rockwell is right about how few people will
be able to utilize it, and 56K would be a waste of money both for the
owners of the analog modems, and for the ISP.
Of course, this wait-and-see stuff is exactly what USR doesn't want
people to do, which is why they are offering X2 upgrades for
the Sportsters and Couriers with free offers that expire before the
end of the year. However, what they fail to mention is that the
X2 code for the digital modems your ISP must buy won't be available
until possibly February (according to USR), well after the point where
you can take the analog modem back to the store for a refund. For the
ISP, it means people calling them on Christmas day who find a "X2-ready"
modem under the tree but have nothing to talk to at 56K. The poor
technical support staff of every ISP and online service is going to hear
this question/gripe over and over as the Christmas approaches.
USR is trying to build installed base before Rockwell parts reach the
market, to give USR leverage in the bargaining rounds that frequently
occur in the standards making process. Rockwell has far more vendors
using their chips that USR does (and indirectly gets more votes in the
standards bodies), but USR probably sells more V.34 modems than all
the others. Despite the market share, 56K is a whole new ball game
and the installed base of V.34 modems doesn't mean a thing unless a
good chunk of those modems are already running a 56K system. Rockwell
did the same thing a few years ago with V.FC 28,800 modems, trying to
use a sizable installed base to force the ITU to come out with a
standard compatible with V.FC. Of course, the ITU shunned Rockwell
and came out with something incompatible. USR apparently is gambling
that this won't happen, but even if it does, a firmware update to the
USR modems should take care of things. Rockwell didn't have that
luxury when the ITU turned away from V.FC.
-----------------------------------------------------------------------
What does this mean to the owner of an analog modem (the ISP customer)?
At the least, you are getting marketing hype from USR. So far, everyone
else has been pretty quiet but the ads in the magazines for other
brands are starting to pick up "56K capable" feature bullets, even if
they have no 56K product in sight.
For now, if you absolutely have to be on the bleeding edge and have the
fastest possible connection, you might consider ISDN. It is a known
quantity, and you don't have to find out what brand of ISDN modem
your ISP owns so that you can go buy the same brand. Apart from fighting
the telephone company over installing ISDN service, it works, and the
value of your hardware investment won't be predicated on what happens in
some standards body in France in six months.
If you want more exotic, there are cable modems and satellite internet
access. If available to you, these are fast, at least until enough
people jump on with you. There is only so much bandwidth in that
satellite up there and in the cable and the owners of both will invariably
overbook the service. The speeds you see today on cable and satellite
will be the highest speeds you will ever see again. Just remember that.
If you want to go with one of the seven 56K proposals mentioned above,
you will have to know what brand of digital modems your ISP has purchased
so that you can buy one that match. Then be sure you don't decide
to also use a second ISP or switch to a different ISP who may have
decided to purchase modems based on one of the other 56K standards,
rendering your modems 56K capability useless when calling that second
provider.
Oh, you must also accept that 56K analog may not work from your house
or business to ANY ISP, no matter what brand of modem you and the ISP
have. You may have to move across town to get a phone line that is
connected to a modern switch AND hope the signal doesn't traverse
any of the many impediments the phone company can inadvertantly
place in the path of a successful 56K connection. And, (if that
wasn't enough), 56K may work between two locations today, and the
next time the phone company re-routes your calls via a different
trunk or switch, or decides to install SLCs to get more calls into
the same number of lines, you can kiss 56K goodbye. You have no
control or say over what happens inside the phone network, unless it
impacts voice calls.
If you are still serious about buying a new analog modem and want it
to be 56K capable, the only model I can suggest at this time is the
USR Courier, not because of the upcoming release of X2 code for this
modem, but because this modem can be completely reprogrammed via
software in the field when the time comes to change it to comply with
the official or "winning" standard, or to fix those bugs in X2
that will fall out in the first six weeks of use. Hello, Beta testers.
Think of the first cut of X2 like a 1.0 version of a Microsoft product.
The Sportster is also a reprogrammable modem, but right now you have
to order a part from USR and take the modem apart and install it.
This isn't hard, but USR can release software fixes on a FTP site a lot
faster than they can make up replacement parts and mail them.
No other vendor offering a modem for less than $450 offers the complete
reprogramability of the Courier at this time. When Rockwell gets
their reprogramable datapump on the market, this will no longer
be the case.
This game will be a lot easier to play once a 56K standard appears, but
I realize those that value speed won't wait that long. Just be
prepared for some hardship and grief in trying to get 56K analog before
the standard appears.
I do NOT recommend 56K analog as a Christmas gift, unless the
sales receipt is in the box, along with a map showing how to get
back to the store.
Disclaimer:
That's all my opinion, and I am cannot be held responsible for your
actions in following or ignoring this information, nor in any actions
the various vendors might take. It is your responsibility to
research your purchases, plus the risk of investing in untried
technologies. This information is provided "AS IS", without warranty
of any kind.
Frank Durda IV <uhc...@nemesis.lonestar.org>|"The Knights who say "LETNi"
or uhclem%nem...@rwsystr.nkn.net | demand... A SEGMENT REGISTER!!!"
|"A what?"
or ...letni!rwsys!nemesis!uhclem |"LETNi! LETNi! LETNi!" - 1983
(c) 1996, ask before reprinting.
Michael LaPorta
David Levin wrote:
>
> On Thu, 12 Dec 1996 06:04:39 GMT, uhc...@nemesis.lonestar.org (Frank
> Durda IV) wrote:
>
> >If you are still serious about buying a new analog modem and want it
> >to be 56K capable, the only model I can suggest at this time is the
> >USR Courier, not because of the upcoming release of X2 code for this
> >modem, but because this modem can be completely reprogrammed via
> >software in the field when the time comes to change it to comply with
> >the official or "winning" standard, or to fix those bugs in X2
> >that will fall out in the first six weeks of use. Hello, Beta testers.
> >Think of the first cut of X2 like a 1.0 version of a Microsoft product.
> >
> >The Sportster is also a reprogrammable modem, but right now you have
> >to order a part from USR and take the modem apart and install it.
> >This isn't hard, but USR can release software fixes on a FTP site a lot
> >faster than they can make up replacement parts and mail them.
> >
> >No other vendor offering a modem for less than $450 offers the complete
> >reprogramability of the Courier at this time. When Rockwell gets
> >their reprogramable datapump on the market, this will no longer
> >be the case.
> >
>
> driver. I use this modem and I can highly recommend it. USR will have
> a 56k driver for it in January. The retail price for this modem is just
> over $100.
>
> David Levin
Guess it's a shame that only 5% of the people in the US can handle the
33,6 connect rate ( 3429 symbol rate ). Don't hold your breath on the
57,6......
Frank Durda IV
David Levin (dle...@dfwmm.net) wrote:
: The USR's Winmodem is also a software reprogrammable modem through it's
: driver. I use this modem and I can highly recommend it. USR will have
: a 56k driver for it in January. The retail price for this modem is just
: over $100.
Sorry, but I don't count "soft modems" with detached brains as being modems.
If I can't plug a modem into a computer running any operating system I
choose (or am forced to choose, such as NT) and expect it to work, it
isn't a modem. It is now a rock.
I don't count RPI modems, AT&T HSM modems and any of the other "you
need to install some Windows '95 drivers first" modems. These modems
save money up front but have too many customer support hassles.
Yes, they are easy to upgrade providing (A) the vendor actually bothers
to offer upgrades, and (B) you can get them for the operating system
you need to use. Personally, I buy the operating system to fit the
application, not the operating system and applications to fit the modem.
Has USR promised to provide new Windows drivers next summer when
'97 comes out and your existing drivers won't work anymore? Or when
you decide to go to NT? Or when you decide to run some other operating
system? I thought not.
My advice: Spend the extra $30 for a modem with all of its brains.
Frank Durda IV <uhc...@nemesis.lonestar.org>|"You might even need a whole new
or uhclem%nem...@rwsystr.nkn.net | brain!" "Oh, I could never
| afford a whole new brain."
or ...letni!rwsys!nemesis!uhclem |"Well, let's look in the catalog."
Jack Beech
David Levin (dle...@dfwmm.net) wrote:
: On Thu, 12 Dec 1996 22:53:46 -0600, Michael LaPorta <mlap...@gte.net>
: wrote:
:
: >Guess it's a shame that only 5% of the people in the US can handle the
: >33,6 connect rate ( 3429 symbol rate ). Don't hold your breath on the
: >57,6......
: But most people get close to the maximum of the technology (24k, 26.4k,
: 28.8k, 31.2k). I see no reason why people won't get reasonable close to
: the 56K connect rate.
You really need to sit down and listen to the USR spill on their service.
If you have a hard time holding 28.8 or 33.6, you wont be even able to
acheieve 56k. Order the comdex tapes and listen to the USR conference..
Or call USR in dallas and ask for a demo. There are several reason why
people wont get 56k. Mostly the wiring in place is not upto standards to
support it. majority of the people have too many boosters in between them
and the co. Your only allowed 1. Anyone outside of a "0" mile radius has
a 75% chance of having atleast 2. You think bell is going to run cat5
cabeling to everyones houses?
Jack
David Reese
I think there is question here about how X2 works. My understanding, which
is admittedly limited, is that X2 kicks in if and only if the modems are
able to negotiate a full speed connection, either 28.8 or 33.6.
I believe that was the point of the X2 name. There are no intermediate
speeds. You will either connect at 26.4 or 56k.
Can someone clarify this point?
David Reese
dre...@onramp.net
David Levin <dle...@dfwmm.net> wrote in article
<32b1e291...@news.why.net>...
David Levin
On 13 Dec 1996 21:52:08 GMT, j...@dfw.dfw.net (Jack Beech) wrote:
>Kinda strange since USR really doesnt suggest buying the modem. They
>admit the modem is only on the market to attract the people wanting to buy
>a lease expensive modem. That is what it was designed for.
>
>Jack
Where did you get the WILD idea that USR doesn't suggest buying the
Winmodem ??? If you are running Windows 3.1 or Windows 95 and you have
no old DOS based telecommuncation programs, it is the BEST choice to
make.
Harold Stevens
In article <58u6v7$t...@library.airnews.net>, bo...@127.0.0.1 writes:
|> X-No-Archive:yes
|>
|> In <32b324a2...@news.airmail.net>, Adst...@toobad.bozo (baldeagl) writes:
[Snip...]
|> >This may come as a surprise to you, Harrison, but your way is not the
|> >only way, not necessarily the best way, and certainly not the popular
|> >way.
|> >
|> >Snobist, purist, Unix, text based browsing, techno-nerds are not the
|> >high priests of proper computing.
|> >
|>
|> But snobist, puritst unix text based browsing techno-nerds have much
|> more power, speed, reliability, scalability and much fewer [GPF-ed.].
[Snip...]
And more years. Well, speaking personally. Grey-bearded putz, me. But, many
fewer viruses. Anyone got a copy of IBMAV for x86 Linux handy? WHY NOT? :)
Anyway: who *ARE* these "high priests of proper computing" du jour? Hopper?
Wozniak? Gates? Negroponte? Mitnick? Andreesen? Canter & Siegel? My! Has it
been reduced to a highschool cheerleader selection process? Don't let on to
my Linux at home; it would be suicidal if it wasn't thought of as "popular"
and gawd forbid called "cheap" or such. :)
I think the operative phrase here is "anthropomorphism". :)
Ya know, everytime I hit the road with my 1982 Toshiba T1000 laptop running
Telix (DOS 2.11), I catch up on all my Unix sendmail in a text mode just as
easily as the mailtool GUI on my Xterm in the office. The major difference,
at least ergonomically, is not having to endure those shreiks of the fellow
engineers with NT jammed down their throats put to the K-mart music torture
on a helpline after GPF, server not serving, 32/16 bit app mismatch, etc. I
suppose it's not so much a function of expecting the OS to save you as much
as using what's at hand effectively. GUI's are nice and you'd have to fight
to take my *FREE* copy of Mosaic back, but hey: easy come, easy go. :)
Maybe I'm just simple minded and easily entertained, but me, DOS, text, and
Unix, well: we all get along just swimmingly. I am sure Windows and I could
romance as well, if it would get over its sojourn in the land of 640K. Like
I care it was married at one time to a 16-bit address space. :)
I mean, sendmail doesn't turn up its nose at my lowly DOS offerings over an
insouciant 1200 baud line... :)
Regards, Weird
wy...@ti.com
Standard Disclaimer: These are my opinions alone; TI is not responsible for
them, nor I for their stable of bloodthirsty patent lawyers. And that stuff
about torture on the helpline is just poetic license. I think. It's hard to
tell when someone's really going postal in the Dilbert Zone...
Harrison Bergeron
Harold Stevens ste...@adam.dseg.ti.com came up with the following:
>
>And more years. Well, speaking personally. Grey-bearded putz, me. But, many
>fewer viruses. Anyone got a copy of IBMAV for x86 Linux handy? WHY NOT? :)
>
I guess no one makes really nice Unix anti virus stuff.
Any clues?
>Anyway: who *ARE* these "high priests of proper computing" du jour? Hopper?
>Wozniak? Gates? Negroponte? Mitnick? Andreesen? Canter & Siegel? My!
>
I vote for Robert Morris Jr.
[...]
>
>I think the operative phrase here is "anthropomorphism". :)
>
Never forget - Computers HATE to be anthropomorphised.
--
dav...@davids.psyberlink.net |
Steinberger: | Please pass the grey poupon.
State of the Instrument |
baldeagl
On 14 Dec 1996 00:44:59 GMT, ro...@davids.psyberlink.net (Harrison
Bergeron) wrote:
[snip]
> Anyone trying to improve their
> computing speed is going to offload as much work as they can.
What a novel idea. Some would actually have the audacity to get a
better processor!
[snip]
>
> No, it's a whole new set of obscure problems.
Funny.....I've been using an RPI 14.4 modem for over two years. While
all the speedsters have been struggling with spiraling death syndrome
and a host of other problems, my old RPI modem has been chugging along
without so much as a hiccup. Meanwhile, my state of the art processor
(well....it was when I bought it) has been handling the "offloaded
chores" without so much as hiccup.
>
> Once again, I highly recommend this modem.
> >
> I'll take it under advisement next time I downgrade.
Some people drive Chevys; others Fords; others Mercedes.
Some people believe in big, fire-breathing, gas gulping monster
motors. Other's believe in small powerplants, nimble suspensions and
great gear ratios.
What the hell's the difference so long as you arrive at your
destination with a grin?
This may come as a surprise to you, Harrison, but your way is not the
only way, not necessarily the best way, and certainly not the popular
way.
Snobist, purist, Unix, text based browsing, techno-nerds are not the
high priests of proper computing.
Besides, UNIX SUXS. <G>
What do you want to fix today?
Go to http://www.isisnet.com/terrymo/
for help with Windows 95 problems
bald...@airmail.net (Paul Schmehl)
http://www.utdallas.edu/~pauls/
baldeagl
On 14 Dec 1996 17:20:25 GMT, ro...@davids.psyberlink.net (Harrison
Bergeron) wrote:
[snip]
Look, Harrison, I have no desire to start an O/S war. <Every> O/S has
it's problems, even your precious Unix. We have eight brand new 9GB
Micropolis hard drives that won't work at all on Unix, but do
perfectly fine in a Windows environment.
The point I've been trying to make through all the "I hate Bill Gates"
noise is; <most> people use a Windows or Mac based O/S. Only a few
use the more esoteric Unix/Linux/et.al O/Ss. Only the most
technically proficient and knowledgeable would even attempt to use
Unix anyway. (And yes, I do use it and can find my way around in it.)
But no one, who has a problem and wants an answer, wants to have to
endure your constant smug denigration of "other O/Ss".
They just want answers. OK???
And what's the point anyway? All the bitching in the world about
Gates' market share won't change a thing. So UNIX is supposedly
"immune" to viruses? So what? It sure likes to do core dumps when it
chokes on input. If it was so perfect, then tell me why IA struggles
constantly with problems with it on their system?
Here's a hint: NO system, no matter how much you love it, is perfect.
Now if someone were to come along and build a better company than
Gates (notice I didn't say product), then they just might knock him
off his pedestal.
I bet he would die laughing (all the way to the bank) if he were to
read the petty attempts to dissuade people from using his products
that are posted all over USENET. I doubt he has the time though.
He's too busy planning the next campaign, the next big rollout, the
next big moneymaker.
Instead of bitching about Gates, why don't you write an O/S the world
can use; that doesn't crash, can run anything the consumer want's to
run, is easy to use and works well on a Network? Oh...and don't
forget. It's needs to be "user friendly". Most of your customers
will not be nearly as intelligent, well schooled, or technically
knowledgeable as you are.
Can't do it? I didn't think so.
Frank Durda IV
>Snobist, purist, Unix, text based browsing, techno-nerds are not the
>high priests of proper computing.
True, but don't expect his one-liness Bill to come to your rescue when
his gang make all of your existing drivers non-functional and you
have to go and beat on the ATIs, Bocas, and USRs of the world,
trying to get new drivers for a model of device they don't
make any more. This isn't idle speculation, it will happen this
summer.
It is likely you will have some of these headaches with older video cards
and sound cards, so why on earth would you add modems to that list too?
Spend the extra $30 and avoid this grief.
Frank Durda IV
David Levin (dle...@dfwmm.net) wrote:
: Has your video card maker promised you a driver when 97 comes out ?.
Nope, but the Standard VGA drivers Microsoft supplies will work to
some extent. Yes, I may not be able to use every ounce of accelerated
stuff in the Matrox-Super-Diamondgrafix-pontoon-ATI-accelerator-chip, but
it will still do something and be usable. If I don't have a new driver
from the no-controller modem vendor, the modem does NOTHING. And
I can guarantee you that Microsoft is not going to get into the
modem driver business. Video, disk and Sound drivers they do, but not
modems.
: Has your printer maker promised you a driver when 97 comes out ?
Not an issue. Printer "drivers" won't be affected. They are really
just translators anyway and not real drivers. Besides, I can plug
that printer into a UNIX, DOS, or even a Cray-XMP and guess what?
It will print. Not so with a no-controller modem. It will just
sit there consuming electricity and its owners patience.
: Has your sound card maker promised you a driver when 97 comes out ?
See above. Microsoft provides stock SoundBlaster compatible drivers,
and I bet my sound card will emit some sound even if Creative Labs didn't
make it? Why? Because SoundBlaster port and command sequences are
standards and many board makers follow them explicitly.
Now if I have gone and bought some non-standard sound card, yeah I am
asking for it, but that is exactly what the person buying a
no-controller modem has done: bought a non-standard modem and the
risks involved.
: My advice to everybody is that the USR Winmodem is a perfect choice for
: someone who wants to easily update the modem with the latest software
: developments.
I must disagree. Although a "soft" or "controllerless" modem can adapt
to very select modem technology changes (most require physical changes
to the DAA and they are just lucking out on 56K), such a modem cannot
hope to be as portable as a modem that has its own controller and own
life. In the WinModem had existed in the days of 14,400 bps modems,
you would not have been able to load new drivers and get 28,800bps.
An entirely new DAA was required.
: This is the start of things to come. You will soon
: see more of these type modems come out in the future.
Do you recall Intels NSP? Did that go anywhere? Using 100% of a
Intel Pentium 100MHz to do something that could have been accomplished
with a $10 microprocessor. Even Intel finally realized the economics
were just stupid and the NSP strategy died. Motorola apparently get
did not the word, as they demoed at Comdex a modem that was NSP (completely
operated by the main CPU) and used most of a Pentium 133. Bet Quake
just crawls on that baby.
Personally, I would rather blow an extra $30 on a modem with all
of this brains than suck half (or more) of the processing power out
of that $2,000 Pentium system.
: BTW, this modem is not a RPI modem.
True. It is worse. A RPI modem can at least make a phone call
and connect without special drivers. A HSM (AT&T / Lucent) or USRs
Winmodem cannot even take the phone off hook without assist from these
drivers. You consider this an advantage?
: This modem has nowhere near the problems that the RPI modems had.
I would like you to explain your basis for this statement. Have
you been in contact with the support centers for any PC manufacturers?
or Microsoft supporr centers? You might hear differently.
I've heard plenty and I have been working with no-controller modems
since April of 1995. They are nothing but trouble, regardless of
vendor.
: Once again, I highly recommend this modem.
I cannot recommend any modem that you can't step into standalone
DOS with no drivers loaded and type ATH1 into a com port and have the
modem go off-hook.
Would you buy a modem that didn't have a Hayes-compatible "AT"
command set? No? Well if you bought a no-controller modem,
you did. Only the driver provides the illusion of the modem accepting
AT commands.
Why do you think the competition now puts "Hardware compression" and
other phrases on the outside of the box? They know a marketing
disadvantage in these no-controller modems when they see one.
Yeah, these controller-less modems are popular with some name brand PC
makers (Compaq and AST are two), because it gets them the cheapest possible
bullet for the side of the box, ie "Includes 28.8 modem", even though it
may only be usable with the software load included with that machine,
and are proven to add cost to the support of those same machines.
Frank Durda IV <uhc...@nemesis.lonestar.org>|"The Knights who say "LETNi"
or uhclem%nem...@rwsystr.nkn.net | demand... A SEGMENT REGISTER!!!"
|"A what?"
or ...letni!rwsys!nemesis!uhclem |"LETNi! LETNi! LETNi!" - 1983
Frank Durda IV
"David Reese" (dre...@onramp.net) wrote:
: I think there is question here about how X2 works. My understanding, which
: is admittedly limited, is that X2 kicks in if and only if the modems are
: able to negotiate a full speed connection, either 28.8 or 33.6.
:
: I believe that was the point of the X2 name. There are no intermediate
: speeds. You will either connect at 26.4 or 56k.
:
: Can someone clarify this point?
The modem will attempt to negotiate the 56K reception mode first.
If that fails, it will attempt to try traditional modes, starting
with V.21 and working up to V.34, just like modems do now.
Frank Durda IV <uhc...@nemesis.lonestar.org>|"How do I know? According to
or uhclem%nem...@rwsystr.nkn.net | this NDA, I have to kill
| myself before I tell you."
or ...letni!rwsys!nemesis!uhclem |
David Levin
On 14 Dec 1996 07:05:38 GMT, "David Reese" <dre...@onramp.net> wrote:
>I think there is question here about how X2 works. My understanding, which
>is admittedly limited, is that X2 kicks in if and only if the modems are
>able to negotiate a full speed connection, either 28.8 or 33.6.
>
>I believe that was the point of the X2 name. There are no intermediate
>speeds. You will either connect at 26.4 or 56k.
>
>Can someone clarify this point?
>
>David Reese
>dre...@onramp.net
You do not have to have a perfect 28.8 or 33.6 connection now to take
advantage of the 56K technology that will soon be out. The 56K
technology uses a completely different way of modulating the data then
the 28.8K modems. Please check out two excellent web sites that
describe all of this in more detail http://x2.usr.com/news/needham.html
and http://x2.usr.com/technology/wp.html.
David Levin
Harrison Bergeron
all 50 watts of, Ronny Ong ronn...@airmail.net came up with:
>>
>> >
>> >And more years. Well, speaking personally. Grey-bearded putz, me. But, many
>> >fewer viruses. Anyone got a copy of IBMAV for x86 Linux handy? WHY NOT? :)
>> >
>> I guess no one makes really nice Unix anti virus stuff.
>> Any clues?
>
>really nice spreadsheets, word processors, games, etc. on Unix.
>
I have two spreadsheet programs which I assume are useful to
someone that would find a spreadsheet program to be a useful
piece of software. I have a hard time seeing their usefulness,
except as an electronic ledger sheet with limited numerical
capabilities. For doing real statistics, I use CERNlib soft-
ware which was free.
FrameMaker? It appears to be a nice what-you-see-is-all-you-get
word processor. But for not having nice word processors, I sure
have a lot of books that were typeset by the authors themselves
using TeX or troff and runoff on their own phototypesetters.
K&R, "The C Programming Language" and Stroustrup, "The C++
Programming Language", for example were done in troff. I don't
recall ever seeing a book that was typeset using Word. In fact,
Knuth wrote TeX because he said he was tired of the poor quality
of the textbooks he published. He also offers money for anyone
finding a bug which doubles with each one found. The last one
was 10+ years ago. That seems fairly nice to me. I don't equate
WYSIWIG with "nice"; only the final product and it's hard to
imagine a nice finished product from a limited selection of
true type fonts.
Games, microsoft can keep. I don't play them and I think they're
a waste of time so it really doesn't bother me if ms has the
ONLY games in existence.
Irwin Sabath
Adst...@toobad.bozo (baldeagl) wrote:
>On 14 Dec 1996 17:20:25 GMT, ro...@davids.psyberlink.net (Harrison
>Bergeron) wrote:
>[snip]
Well said Paul.
Long overdue and necessary.
--
Irwin
t.i.n.s.t.a.a.f.l.
Bill Conde
ro...@davids.psyberlink.net (Harrison Bergeron) wrote:
>
> FrameMaker? It appears to be a nice what-you-see-is-all-you-get
> word processor. But for not having nice word processors, I sure
> have a lot of books that were typeset by the authors themselves
> using TeX or troff and runoff on their own phototypesetters.
> K&R, "The C Programming Language" and Stroustrup, "The C++
> Programming Language", for example were done in troff. I don't
> recall ever seeing a book that was typeset using Word. In fact,
> Knuth wrote TeX because he said he was tired of the poor quality
> of the textbooks he published. He also offers money for anyone
> finding a bug which doubles with each one found. The last one
> was 10+ years ago. That seems fairly nice to me. I don't equate
> WYSIWIG with "nice"; only the final product and it's hard to
> imagine a nice finished product from a limited selection of
> true type fonts.
>
David,
It never ceases to amaze me how you continually take the antagonist
position. Most people know that the Microsoft OS's are flawed.
Unfortunately, they are the most popular, worldwide. If one is in
business to make money, it would seem to me to be the smarter choice
to develop a product that can be used by the greatest number of
individuals.
As for your comments on publishing, you have ventured into an area in
which I happen to have over 22 years experience. The overwhelming
majority of typesetting today is done using the MacOS. It's graphic
nature lends itself more readily to the development of typefaces as
well as other graphic images. We are now doing things with a Mac that
would have cost nearly a million dollars less than ten years ago for
the hardware to produce them. As you know, UNIX predates the MacOS, as
well as the popular development of desktop publishing. If it is such a
great publishing OS, why has it continued to languish in the esoteric
provinces of our educational/governmental institutions? Could ease of
use have something to do with it? How about the Mac's WYSIWYG
capabilties?
Granted, UNIX has its pluses; not the least of which is its
flexibility. The best does not always win the battle. How about the
Betamax as a prime example? It is still widely used professionally,
but VHS has become the worldwide standard for the rest of us. UNIX may
be ideal for your applications as a nuclear physicist, but for the
vast majority of us, a GUI is the way to go, and there is no
overriding reason for us to change. It seems to me you could better
spend your time perfecting your OS for those who do use UNIX, rather
than constantly harping on us unwashed GUI-users.
--
Bill Conde
wco...@iadfw.net
Harrison Bergeron
all 50 watts of, Frank Durda IV uhc...@nemesis.lonestar.org came up with:
>David Levin (dle...@dfwmm.net) wrote:
>: Has your video card maker promised you a driver when 97 comes out ?.
>
>Nope, but the Standard VGA drivers Microsoft supplies will work to
>some extent. Yes, I may not be able to use every ounce of accelerated
>stuff in the Matrox-Super-Diamondgrafix-pontoon-ATI-accelerator-chip, but
>it will still do something and be usable.
(which I did pay for, lest anyone think my objection to
microsoft is strictly monetary in nature), doesn't seem
to mind the fact that I've changed c libraries and updated
the kernel a couple of times and lists 410 video cards
supports but, just in case, you can do it by chipset. It
runs 1152x900x16bpp just fine.
>
>: Has your printer maker promised you a driver when 97 comes out ?
>
>Not an issue. Printer "drivers" won't be affected. They are really
>just translators anyway and not real drivers. Besides, I can plug
>that printer into a UNIX, DOS, or even a Cray-XMP and guess what?
>It will print.
formatX into pcl or the native printer language for a formatX
that the printer doesn't support directly. The 10 year old
laser printer I've been loaned seems to print okay without any
driver support. Does postscript, too - directly from netscrape
even.
>
>Now if I have gone and bought some non-standard sound card, yeah I am
>asking for it, but that is exactly what the person buying a
>no-controller modem has done: bought a non-standard modem and the
>risks involved.
>
documentation out of someone. Proprietary - OK, Unsupported OK,
Proprietary AND Unsupported - NOT OK.
[...]
>
>Personally, I would rather blow an extra $30 on a modem with all
>of this brains than suck half (or more) of the processing power out
>of that $2,000 Pentium system.
>
a faster processor is for? Hell, I can wait an extra 45
minutes to get the data fitted. I mean, 30 bucks is 30 bucks.
Since many people inconvenience themselves with the cost and
additional cabling of an external modem to isolate a modem
faux pas - well, to the modem - and avoid a pointless reboot,
it's hard to imagine that someone would save 20-30 bucks and
chance Yet Additional Catastrophic Code (no, I won't say it,
those of you that know, don't need the acronym to file bad
pun charges).
>Winmodem cannot even take the phone off hook without assist from these
>drivers. You consider this an advantage?
>
"Feature", Frank - "Feature". It doesn't need floating point
does it?
>
>Yeah, these controller-less modems are popular with some name brand PC
>makers (Compaq and AST are two), because it gets them the cheapest possible
>bullet for the side of the box, ie "Includes 28.8 modem", even though it
>may only be usable with the software load included with that machine,
>and are proven to add cost to the support of those same machines.
>
Does this mean I can get away with selling a box of graphite
that is labeled: Diamond inside - just pressurize with
compatible hardware?
--
dav...@davids.psyberlink.net _______________________________
Steinberger: | Please pass the grey poupon |
State of the Instrument -------------------------------
David Levin
>Nope, but the Standard VGA drivers Microsoft supplies will work to
>some extent. Yes, I may not be able to use every ounce of accelerated
>stuff in the Matrox-Super-Diamondgrafix-pontoon-ATI-accelerator-chip, but
>it will still do something and be usable. If I don't have a new driver
>from the no-controller modem vendor, the modem does NOTHING. And
>I can guarantee you that Microsoft is not going to get into the
>modem driver business. Video, disk and Sound drivers they do, but not
>modems.
I am glad you know so much about what Microsoft will support. Are you
are on there software development team ? Saying Microsoft or USR will
or will not have drivers for a future OS is pure speculation. Being
that USR is the number one selling modem company right now and the
Winmodem is a big seller, I have little doubt that a driver will be
available.
>
>: Has your printer maker promised you a driver when 97 comes out ?
>
>Not an issue. Printer "drivers" won't be affected. They are really
>just translators anyway and not real drivers. Besides, I can plug
>that printer into a UNIX, DOS, or even a Cray-XMP and guess what?
>It will print. Not so with a no-controller modem. It will just
>sit there consuming electricity and its owners patience.
Wrong, wrong, wrong. There are many printers that are coming out now
that require the use of the Windows GDI and are host-based. Is that
bad, not according to PC Magazine. Check out page 106 of the Nov 5,1996
issue. Many of these printer will not work on another OS.
>Now if I have gone and bought some non-standard sound card, yeah I am
>asking for it, but that is exactly what the person buying a
>no-controller modem has done: bought a non-standard modem and the
>risks involved.
I don't think buying a popular modem from the world's number 1 selling
modem vendor is a risk. I don't know why you think it is, but I won't
be producing an ounce of sweat over this. The product works perfectly
now and I have little doubt it will work in the future.
>I must disagree. Although a "soft" or "controllerless" modem can adapt
>to very select modem technology changes (most require physical changes
>to the DAA and they are just lucking out on 56K), such a modem cannot
>hope to be as portable as a modem that has its own controller and own
>life. In the WinModem had existed in the days of 14,400 bps modems,
>you would not have been able to load new drivers and get 28,800bps.
>An entirely new DAA was required.
If I wanted a 'portable' modem, I would have bought one. I buy a
product based on what I have. The product works for me and I am very
satisfied.
>Do you recall Intels NSP? Did that go anywhere? Using 100% of a
>Intel Pentium 100MHz to do something that could have been accomplished
>with a $10 microprocessor. Even Intel finally realized the economics
>were just stupid and the NSP strategy died. Motorola apparently get
>did not the word, as they demoed at Comdex a modem that was NSP (completely
>operated by the main CPU) and used most of a Pentium 133. Bet Quake
>just crawls on that baby.
>
>Personally, I would rather blow an extra $30 on a modem with all
>of this brains than suck half (or more) of the processing power out
>of that $2,000 Pentium system.
I guess you didn't see my other post. I have done thorough testing
using a 'regular' 28.8 modem and the Winmodem. Using both a highly
compressible text data file and a zipped file, I found the Winmodem
only uses only between 1 and 5% of a Pentium 120 CPU. I think I have
plenty of spare CPU cycles when I am surfing on the Internet, I am not
exactly calculating giant Spreadsheets when I am surfing. I bought a
Pentium to compute not just watch over a device.
>
>: BTW, this modem is not a RPI modem.
>
>True. It is worse. A RPI modem can at least make a phone call
>and connect without special drivers. A HSM (AT&T / Lucent) or USRs
>Winmodem cannot even take the phone off hook without assist from these
>drivers. You consider this an advantage?
>
I don't understand why you continue to think it's so 'BAD' that a modem
requires a driver to operate. If the modems works it doesn't matter how
it works.
>
>: This modem has nowhere near the problems that the RPI modems had.
>
>I would like you to explain your basis for this statement. Have
>you been in contact with the support centers for any PC manufacturers?
>or Microsoft supporr centers? You might hear differently.
>I've heard plenty and I have been working with no-controller modems
>since April of 1995. They are nothing but trouble, regardless of
>vendor.
Have you been in contact with these support centers about the Win modem
? Just because a somewhat similar product has generated alot of calls,
doesn't mean you can lump them all together and say 'ALL' of them will.
This is pure speculation. I can say for my experience and also of my
friends, we have not had one problem with out Winmodems and we are very
satisfied with our modems.
>
>: Once again, I highly recommend this modem.
>
>I cannot recommend any modem that you can't step into standalone
>DOS with no drivers loaded and type ATH1 into a com port and have the
>modem go off-hook.
That is a ridiculous statement. I will have no need to go into DOS with
no drivers to type AT commands as will most people. If I wanted to
type AT commands I have several windows based programs that will allow
me to do this just fine. I don't know if you looked around lately, but
most people nowadays are using Windows to do their work, not DOS.
>Would you buy a modem that didn't have a Hayes-compatible "AT"
>command set? No? Well if you bought a no-controller modem,
>you did. Only the driver provides the illusion of the modem accepting
>AT commands.
No, it's the software in the ROM chip of your 'regular' modem that
provides this capability. Does it matter where the code is out ? NO !!
Is the product hayes-compatible -- YES !!!, what differnence does it
where the code is at and how it works ?? I for one would rather it be
in a driver then in a hard-coded chip.
>Why do you think the competition now puts "Hardware compression" and
>other phrases on the outside of the box? They know a marketing
>disadvantage in these no-controller modems when they see one.
Where the compression is being done doesn't make a differnance if you
have a modern system. Like I said eairly, it only uses 1 to 5% of my
Pentium 120.
The bottom line is the modem works and work very well. Maybe it won't
please die-hard people that are so used to 'smart' modems that they
can't figure out why someone would buy otherwise, but it will satisify
most people. If I can get to point A to point B in the same amount of
time, it doen't matter what form of transportation I am using as long as
their both reliable.
BTW, since you will only be buying stand-alone 'smart' devices, their
are many printers and other devices you need to avoid, so watch out.
David Levin
Allen Eastwood
Well said Paul.
Frankly all the OS wrangling just irritates the hell out of me. Guess
what, I use Unix, Linux Windows and Windows 95 all either at home or
at work and I wish you Unix guys would just can it. That's what
advocacy newsgroups are for.
Adst...@toobad.bozo (baldeagl) wrote:
Frank Durda IV
Harrison Bergeron (ro...@davids.psyberlink.net) wrote:
: Does this mean I can get away with selling a box of graphite
: that is labeled: Diamond inside - just pressurize with
: compatible hardware?
God, I'm too late. The Radio Shack Merchandising aliens already got to
Harrison. They come from the planet Vapam. He thinks like them.
He speaks like them. He's one of them. Harrison must have left one of
those "innocent" advertising flyers in his house when he was asleep.
Once someone is converted, you can't kill them as they cannot die; you
can only fill their mouths in with cement. Makes 'em quieter. :-)
Frank Durda IV <uhc...@nemesis.lonestar.org>|"You've got money? We want it."
or uhclem%nem...@rwsystr.nkn.net | - One of the proposed slogans
| for the new Radio Shack.
or ...letni!rwsys!nemesis!uhclem | Considered unusually honest.
Frank Durda IV
Harold Stevens (ste...@adam.dseg.ti.com) wrote:
: Anyway: who *ARE* these "high priests of proper computing" du jour? Hopper?
: Wozniak? Gates? Negroponte? Mitnick? Andreesen? Canter & Siegel?i
That's easy (in no particular order) Bill Joy, Mr. K and Mr. R, Mike O'Dell,
Rob Kolstad, Michael Karels(sp?), Steve Leineger, Randy Cook, Marshall
McKusick(sp?), and on a "being liberal" day, Bill Jolitz. Yeah, I'll
even include Jobs and Wozniak. I won't include Mr, Honeyman as he
complains too much. :-) If you don't know who these people are or
what they did/do, boy did you miss out.
Most of the people that you mentioned are not priests of computing, they
are priests of marketing. That's two doors down, right next to
Being-Hit-On-The-Head lessons. I've seen Bills and Paul Allens BASIC code.
Paul was the programming brains of the outfit. :-)
It's funny, but if the laws that exist today were in place in 1975, Mr Gates
might be a felon. The original Altair BASIC was written using stolen
computer time. Then Bill resold the BASIC code to others computer makers
despite a contract with MITS, which for some unknown reason was eventually
ruled invalid in a NM court. He got lucky.
: Maybe I'm just simple minded and easily entertained, but me, DOS, text, and
: Unix, well: we all get along just swimmingly. I am sure Windows and I could
: romance as well, if it would get over its sojourn in the land of 640K. Like
: I care it was married at one time to a 16-bit address space. :)
Hey, I went from TOPS-20, a wonderful virtual memory operating system,
down to Z-80 systems with anywhere from 16 to 64K of RAM using TRSDOS.
Made me think that IBM 370/ALC maybe wasn't that bad. Hmm, yeah, 370/ALC
really was that bad - Z80s were better since they at least had a stack. :-)
I use the tool (or OS) that works for tasks I want done. In my case, I
beta test Windows OSes for Microsoft because it gets me free software
(Just signed the latest NDA for the next version, so you can stop
buying the current stuff now). I even went to the Windows '95 launch
party in Redmond - was twenty feet from the thoroughly soiled Jay Leno and
a Bill who stared at the teleprompter (running "Master Script 6.0")
constantly.
However, I do all my work on FreeBSD. No software crashes, no GPFs,
no ping-o-death memory leaks, no viruses. This system has been up
67 days (when I upgraded the OS last). I tend to have to reboot Windows
a tad more often. End of Song.
VI is my friend and I can compose in it ten times faster than in the
Microsoft vision of editing used in all of the Windows products. Give me
a Yank-and-Put keystroke and I'll beat cut & paste mouse clicks into the
mud everytime. On DOS, I use VI too. (I don't use EMACS, reminds me of
LISP and MDL.)
Yeah I can't run Excel on FreeBSD (yet). But I never used Excel when I
could. I boot Billsoft products on my toy computers when I want to play
a game, edit audio or something, but by default, MS products are not
running here.
If Microsoft's own people couldn't get NT to do the job on MS-NBC, and on
their own FTP systems where both had to switch to non-Microsoft operating
systems, that should say a lot. I want my machines to be ruled, not
rule me.
Frank Durda IV <uhc...@nemesis.lonestar.org>|"If you don't have operating
or uhclem%nem...@rwsystr.nkn.net | system religion, get some!"
| :-)
or ...letni!rwsys!nemesis!uhclem |
Frank Durda IV
baldeagl (Adst...@toobad.bozo) wrote:
: The point I've been trying to make through all the "I hate Bill Gates"
: noise is; <most> people use a Windows or Mac based O/S.
Actually, most people don't hate Bill, they hate the Bill machine.
If you sit in a room for a few hours with Bill, you discover he
is obnoxious, still can't comb his hair or tuck his shirt in, is
overbearing, tyranical, and doesn't take input real well. But he is
one guy and insigificant.
It is the machine of thousands of marketeers that march in a cordon
around Bill that generate the insane licensing policies, treat industry
partners as saps, buys and eliminates competition, dumps products
in markets below cost to gain market share (illegal in most countries),
denies what is obvious, is under investigation almost continuously over
the past ten years, etc. These actions are what people who have a
dislike dislike. (Your Highness, your Highness.)
Is Microsoft better at what they do than other companies? In marketing,
they are pretty good. In code development, so-so. There are reasons for
that, such as the need to have 50% of the installed base upgrade every 18
months in order to keep the cash flowing, which means you bring out
software based on dates on calendars, not when it is really ready.
There are a lot of smart people who are burned dry in the first 12 months
because of the right-to-left scheduling and other pessures. This happens
no matter how many free Cokes, volleyball breaks and stock bonuses they
shove into their cells. The question is then whether these fried people
move on to other companies or participate in the clod theory.
Windows '95 slipped several times, until it started seriously affecting
revenue, then a stake was driven in the ground for August 24, 1995.
When June 13th 1995 came around, Beta testers were told "thanks", and
be sure to re-file your unfixed bugs in September for the Nashville
release.
: Only a few use the more esoteric Unix/Linux/et.al O/Ss. Only the most
: technically proficient and knowledgeable would even attempt to use
: Unix anyway. (And yes, I do use it and can find my way around in it.)
Don't forget that Microsoft licensed UNIX in the V32 days and turned
it into a product: XENIX. However, when development began on textual
Windows and the PCs with no memory protection hardware, Microsoft scaled
back the interest in UNIX and went a different way, randomly releasing
and re-gathering control from SCO.
Undoubtedly, UNIX remains a more difficult platform for the average
person to use and be productive with, even with X. I still find X
extremely non-obvious when compared to Machintosh, or its knock-off,
graphical Windows. However, at certain tasks the underlying OS, NOT
THE UI, are better than anything else out there. Someone just needs
to write a consumer-friendly UI.
: And what's the point anyway? All the bitching in the world about
: Gates' market share won't change a thing.
No, but he makes an excellent target. I think I liked him best
on the Letterman show, standing behind the malfunctioning "Quiz Machine",
wearing a white lab coat and a hard hat. :-)
: So UNIX is supposedly "immune" to viruses? So what?
Well, it is a nice selling point. OS/2 seems to list that as a
feature from time to time. It does mean that I don't have to blow
time scanning and sniffing every file that happens to arrive on my
system, which is about 65Meg a day. Face it, DOS/Windows/Machintosh
just never envisioned the need for security of any sort. The
first virus I ever heard of was on the Mac, not DOS/Windows.
: It sure likes to do core dumps when it chokes on input.
Hmm, bad application programming is responsible here. Try SoftRAM '95,
Myst, or almost any version of PowerPoint for the most spectacular crashes.
PowerPoint dying is really annoying, because the room is invariably
full of VIPs, not interested in watching the presenter try to get
out of a GPF loop in some low-level part of Windows.
The problem with Windows is its DOS-upbringing, where writing
anything to anywhere was OK but "you should know better". At least
user A on a UNIX, OS/2, or even to some extent NT system, can't corrupt
the OS so badly that it impacts user B. External influences excluded,
like stepping on the FDDI cable, pouring coffee into the works, etc.
: If it was so perfect, then tell me why IA struggles
: constantly with problems with it on their system?
I suggest you investigate an ISP or other major Internet service that
is using all-NTs or some other operating system and see what their
uptime looks like, and how many customers they can put on the same
amount of hardware. That directly equates to price per month that
the ISP can offer without having totally crappy service.
I'd suggest MS-NBC, but they switched to UNIX (I don't know which) because
the Microsoft developers they shipped on-site could not make NT
work with the load it was being given in the first week of operation.
I believe there is a local ISP that uses NT, but I couldn't verify that as
they appear to be down right now. :-(
: Here's a hint: NO system, no matter how much you love it, is perfect.
Absolutely true. I am a supporter of FreeBSD, but I have sent in
a half-dozen problem reports in the past week. Caused a few
problem reports to be sent in on OSF as well in recent weeks.
I never really liked SCO XE/UNIX. Yes, it is stable, but frozen in 1987.
A lot of new things have come along since then.
And in my recent beta test of the OSR2 of Windows '95, I pulled the
"NO GO" bell on the release candidate (they HATE it when you do this),
citing serious flaws in the TCP/IP stack and its idle detection, along
with about 30 other problem reports, not counting the ones they didn't
fix from the original Win '95 release.
They shipped anyway. Hope you enjoy those bugs. I didn't.
: Now if someone were to come along and build a better company than
: Gates (notice I didn't say product), then they just might knock him
: off his pedestal.
There is no pedestal. We are beyond that. He is building the
equivalent of the Randolph Hearst mansion. That will be where the
tours go in sixty years, assuming he isn't in a liquid oxygen chamber
or cybernetic by then. Might even get declared a National Monument.
Not many people need a house with 50,000 square feet (size reported
in August 1995 and it is still under construction), so it probably won't
go on the market, except as pad sites in a mid-sized shopping mall
that needs a trampoline store. :-)
: I bet he would die laughing (all the way to the bank) if he were to
: read the petty attempts to dissuade people from using his products
: that are posted all over USENET.
He doesn't. He has a group of people that do. Sometimes they even
react. Sometimes problems reported in the newsgroups get fixed
and a press release issued before one call about the problem hits
the support center. I know my former employer was contacted from
Microsoft a time or two about complaints being raised in the newsgroups
that blamed Microsoft for some problem. Microsoft had gone out and bought
one of our PCs and found it to be a flaw in the computer, not their
product. Microsoft wanted us to take action to clear their name.
: Instead of bitching about Gates, why don't you write an O/S the world
: can use; that doesn't crash, can run anything the consumer want's to
: run, is easy to use and works well on a Network? Oh...and don't
: forget. It's needs to be "user friendly".
Why? Even Microsoft has discovered that you can be too "user friendly",
and become insultingly condescending. See Microsoft Bob. See Bob not
sell. They sold something like nine copies. Customers clearly want
something they can understand and use effectively, but not something that
treats them like morons and is built on the assumption that they will
never get smarter or more advanced in their use of the system.
The reason people have flashing 12:00 on their VCR isn't because
they don't know how to set it (most do or are willing to walk through
the procedure in the manual), they just don't want to reset it every
time a microscopic power hit comes along and resets the clock back to
midnight. All that is needed is a VCR that can survive without power
for several minutes without losing the time.
But even if you wrote the most marvelous operating system, with a
friendly GUI that walked the dog and washed the cat, it would have
to be compatible with the Windows API. There is too much inertia
in the consumer market. That same inertia is why the new digital VCR
hasn't taken off. Someone decided to make the tape size different than
VHS or 8mm, and the new players can't play existing tapes, of which
there are millions. Bad move.
: Can't do it [write a replacement OS favored by all]? I didn't think so.
Actually, Java/HTML are supposed to do this for us. It is just
a question of how much entanglement wire Microsoft piles on to
protect its investment in Windows, in the form of ActiveX and
strange little extensions to HTML that the W3C are unlikely to
approve. It will take a long time for Java/HTML to have enough
applications for the goal of making the underlying operating system
and processor unimportant to be reached.
Frank Durda IV <uhc...@nemesis.lonestar.org>|"The Knights who say "LETNi"
or uhclem%nem...@rwsystr.nkn.net | demand... A SEGMENT REGISTER!!!"
|"A what?"
or ...letni!rwsys!nemesis!uhclem |"LETNi! LETNi! LETNi!" - 1983
Frank Durda IV
Harrison Bergeron (ro...@davids.psyberlink.net) wrote:
: While this has been viewed as OS snobbery, I don't think it's
: any more snobbish than the microsoft-centric people that post
: articles with control characters (~^R, ~^S, for example) or
: post an article that for some odd reason is in HTML, or don't
: bother to wrap their lines, or post multiple copies of the
: same article (time after time after time). I see this a being
: very snobbish; to be unaware that - not only are there a large
: number of non-microsoft people reading (I'd say a very large
: number) news, but the protocol doesn't use bolding, underlining,
: HTML, proportional fonts or embedded graphics and if it did,
: they'd be looking at xbm and xpm files rather than bmp's and
: fond designations would look like this.
You ought to see what those Microsoft-isms look like on WebTV, a
hopelessly consumerized device. Might as well turn it off and watch
Perry Mason.
Harrison Bergeron
Frank Durda IV uhc...@nemesis.lonestar.org said:
>baldeagl (Adst...@toobad.bozo) wrote:
>: The point I've been trying to make through all the "I hate Bill Gates"
>: noise is; <most> people use a Windows or Mac based O/S.
>
>Actually, most people don't hate Bill, they hate the Bill machine.
>If you sit in a room for a few hours with Bill, you discover he
>is obnoxious, still can't comb his hair or tuck his shirt in, is
>overbearing, tyranical, and doesn't take input real well. But he is
>one guy and insigificant.
>
making money and stop trying to put himself in a position
to control and meter every byte that goes from point A
to point B by somehow trying to rationalize that any device
that doesn't need some kind of license from microsoft
threatens his right to engage in business.
>: Only a few use the more esoteric Unix/Linux/et.al O/Ss. Only the most
>: technically proficient and knowledgeable would even attempt to use
>: Unix anyway. (And yes, I do use it and can find my way around in it.)
>
And of all of the operating systems I've used, it is by
far the easiest to work in and most logical when trying
to do something unusual and having to base the proceedure
on the consistency of the interface (shell).
>
>Undoubtedly, UNIX remains a more difficult platform for the average
>person to use and be productive with, even with X. I still find X
However, the X market would probably kill a vendor that
presumed that they might want some configuration done
for them. There is nothing that prevents any vendor from
eliminating most of the flexibility and turning out an
interface with a windows-like rigidity, look and feel to
eliminate surprises. Of course, once you elimite the useful
features, the only remaining advantage is portable graphics
(which is a real + once you've ever gotten used to the idea
that the screen you are using and the location of the computer
running your software don't need to be in the same country or
of the same archetecture.) Being able to arrange multiple
screens as a single logical screen is also kind of neat.
>
>: So UNIX is supposedly "immune" to viruses? So what?
>
The point is, that in a Unix (or VMS or any OS with a concept
of permissions) a virus cannot survive and propagate without
having access to the root account. Most Unix people build
quite a bit of the software they use so that one rarely runs
across binaries that don't have a well defined point of origin.
Since the source code usually has the authors names sitting
at the top of the files, the concept of not being able to
go straigt to the person that wrote it provides incentive to
make sure that the code does what it claims, since any code
with bad intentions would be spotted and warnings would be
all over usenet about a day after it was released, if that
long. Also, no one compiles software under the root account
anyway unless there is a compelling reason to do so.
>
>: It sure likes to do core dumps when it chokes on input.
>
I generally find it easier to delete corefiles than
reboot a frozen machine. In addition, with the sources
you have the opportunity to fix the problem without
waiting until someone else gets around to it.
>The problem with Windows is its DOS-upbringing, where writing
>anything to anywhere was OK but "you should know better". At least
>user A on a UNIX, OS/2, or even to some extent NT system, can't corrupt
>the OS so badly that it impacts user B. External influences excluded,
>like stepping on the FDDI cable, pouring coffee into the works, etc.
>
I would hope that NT dealt with the shortcomings of allowing
the world access to the hardware and didn't screw it up. Given
that dave cutler, the creator of what is probably the most
fascist OS to ever be built to the extent that you practically
need a permission to ask for permissions, they won't have much
of an excuse. Of course, VMS sort of suffered a bit in popularity.
>
>: If it was so perfect, then tell me why IA struggles
>: constantly with problems with it on their system?
>
>I suggest you investigate an ISP or other major Internet service that
>is using all-NTs or some other operating system and see what their
>uptime looks like, and how many customers they can put on the same
>amount of hardware.
Yes, find an ISP that can even provide terrible service with
that many customers and I'll admit it may become competitive.
NT has a somewhat different concept of multi-user than the
rest of the world.
>
>: Here's a hint: NO system, no matter how much you love it, is perfect.
>
I'll let you know once Brazil is available.
>
>: Now if someone were to come along and build a better company than
>: Gates (notice I didn't say product), then they just might knock him
>: off his pedestal.
>
I don't think article quality is what separates Scientific
American from People Magazine but I'm not planning on
seeing SA right next to the cash register in the supermarket
any time soon.
>
>: I bet he would die laughing (all the way to the bank) if he were to
>: read the petty attempts to dissuade people from using his products
>: that are posted all over USENET.
>
Who is attempting to dissuade anyone. Do you think I would
direct an antagonistic post toward someone that could be
persuaded? I generally take a productive discussion to email
since it seems most people become bored with a long thread
that fails to include any potential for a scatalogical spinoff.
>
>: Instead of bitching about Gates, why don't you write an O/S the world
>: can use; that doesn't crash, can run anything the consumer want's to
>: run, is easy to use and works well on a Network? Oh...and don't
>: forget. It's needs to be "user friendly".
>
>Why? Even Microsoft has discovered that you can be too "user friendly",
>and become insultingly condescending. See Microsoft Bob. See Bob not
>sell. They sold something like nine copies. Customers clearly want
>something they can understand and use effectively, but not something that
>treats them like morons and is built on the assumption that they will
Good thing. Instead of windows on every platform it would
have been MacBob, XBob, Bobix, VMBOB, and so on.
>time a microscopic power hit comes along and resets the clock back to
>midnight. All that is needed is a VCR that can survive without power
>for several minutes without losing the time.
>
I've always wondered how anyone that can design the ciruitry
in a VCR can miss the usefulness of a battery.
>
>: Can't do it [write a replacement OS favored by all]? I didn't think so.
>
Why would I want to produce a product that would appeal
to the very people that liked an OS to the extent that
I found it easier to write my own? The entire premise
is flawed in concept as the result would gaurauntee I'd
find it unusable.
>Actually, Java/HTML are supposed to do this for us. It is just
>a question of how much entanglement wire Microsoft piles on to
>protect its investment in Windows, in the form of ActiveX and
How many entaglement attories can MS products support before
it ceases to be cost effective? Obviously quite a few.
Frank Durda IV
David Levin (dle...@dfwmm.net) wrote:
: I am glad you know so much about what Microsoft will support. Are you
: are on there software development team ?
Co-developed several projects with Microsoft, including one Microsoft
now denies ever existed. Know people all over Redmond. (Pick a
building #.), Was on Windows '95 (and OSR2 and now Memphis) beta programs
since 1994. Yeah, I know a few things about what is going on up there.
: Saying Microsoft or USR will
: or will not have drivers for a future OS is pure speculation.
No it isn't, sorry. I know what Microsofts policy on modem drivers
is, and I had reps from USR in my office three weeks ago discussing
56K and other USRisms.
If you had more recent meetings with these organizations, then please
give specifics as far as your NDAs allow.
: Being that USR is the number one selling modem company right now and the
: Winmodem is a big seller, I have little doubt that a driver will be
: available.
Oh, I don't doubt they will eventually get around to rolling a new
driver, particularly for those units sold in the previous 90 days or
so, much like the way Sportster 56K upgrades are being handled. Your
unit is older than a date in August? You are out of luck.
You may not be aware, but there have been multiple revisions of the
WinModem and they use different code bases, just like the MT, MP and MS
HSM chipsets from AT&T. You need different drivers for each. AT&T
only supports the MT chipset these days. If you have the older chipsets
in your modem, tough. No new drivers will be produced for those
revisions.
: Wrong, wrong, wrong. There are many printers that are coming out now
: that require the use of the Windows GDI and are host-based. Is that
: bad, not according to PC Magazine. Check out page 106 of the Nov 5,1996
: issue. Many of these printer will not work on another OS.
Yes, I or you or anybody else can come up with something so
hopelessly incompatible that the translation driver isn't sufficient.
If you feel strongly about this and are unafraid, do buy one.
I assume you like the rugged OS life, use off-market operating
systems like UNIX, etc.
I will be happy buying Postscript printers which will work just
fine with the stock Microsoft drivers, which will be available
longer than I will be alive. Or even something that is Epson FX-80
or IBM Proprinter compatible (these two are about the same when it
comes to printer commands and came out nearly ten years apart).
The stock Microsoft drivers works OK here too. No need to call Epson
up and have them say "That's last years model. We don't offer new
drivers for it."
: If I wanted a 'portable' modem, I would have bought one. I buy a
: product based on what I have. The product works for me and I am very
: satisfied.
Again, your choice. I just recommend against a modem with so many
strings attached. If it does the job for you, use it.
: I guess you didn't see my other post. I have done thorough testing
: using a 'regular' 28.8 modem and the Winmodem. Using both a highly
: compressible text data file and a zipped file, I found the Winmodem
: only uses only between 1 and 5% of a Pentium 120 CPU. I think I have
: plenty of spare CPU cycles when I am surfing on the Internet, I am not
: exactly calculating giant Spreadsheets when I am surfing. I bought a
: Pentium to compute not just watch over a device.
You didn't read my original text clearly. The next step beyond
the "Soft" and "no-controller" modem (such as the WinModem), is the
NSP modem, where there is no DSP either, just a set of A-D/D-A
somewhere that are wired to the phone line. The Pentium or
Pentium PRO is completely responsible for generating the V.34
mod/demodulation, performing echo cancellation, constellation stuffing,
and error-correction, compression and command set parsing. These
are the types of modems that use 50% or more of the expensive CPU.
Yes, a HSM, RPI or a WinModem only use between 5 and 30% (burst)
CPU on a 100MHz Pentium. Fine. Let's shave 5% off that $2,000
price and I see $100 of the performance I paid for wasted when an
extra $30 on the modem would have preserved all that performance,
AND never given a worry when I wanted to upgrade OSes or change OSes.
That's my opinion. Clearly you feel good with what you have now,
so go with it.
But if you recommend it to others, do mention that you are gambling
their purchase on *your* hope that the vendor will provide drivers for
new operating systems in the future for free. If they do offer
updates but they aren't free, the first $29.95 update could eat the
savings you had on the original purchase. Yes, a firmware update
for a traditional modem might cost too, but it will be bought to
fix bugs, not just to be able to use a new operating system.
: I don't understand why you continue to think it's so 'BAD' that a modem
: requires a driver to operate. If the modems works it doesn't matter how
: it works.
I've stated the reasons repeatedly in this thread. See above and
previous postings.
: Have you been in contact with these support centers about the Win modem
: ? u
As a matter of fact, I have sat in their call support centers monitoring
calls, and have also sat in the call support center for AST, who also
uses controller-less modems. I was nearly assualted by the support
staff at AST when they found out I was in the group that selected such
a modem for AST to bundle with their PCs (against my wishes). Apart
from some bundled software that was flakey, the soft-modem was the #2
call generator at AST during that time.
: Just because a somewhat similar product has generated alot of calls,
: doesn't mean you can lump them all together and say 'ALL' of them will.
But it was the same problem. People erased the Windows '95 and
installed OS/2, or Linux, or Windows NT, or just want to run in DOS
so Quake will run fast and guess what? The modem didn't work anymore!
Grumble! Call support to get my lousy modem repaired. Or worse,
return the entire computer and get a refund. It happened over and
over again.
These things haven't been out long enough to encounter the point
where old drivers won't work with a new version of the same OS, but
we all know that day comes next summer. Microsoft says so.
: This is pure speculation.
Nope, real life. See above.
: I can say for my experience and also of my
: friends, we have not had one problem with out Winmodems and we are very
: satisfied with our modems.
Fine, I have just experienced a larger sample, that is all. Modems
is what I did for a living for a few years. Still do, to some extent.
: The bottom line is the modem works and work very well.
Please add the words "under the environment I choose to run under" and
proceed. The moment you step from those confines, your opinion
of the product should decline.
: BTW, since you will only be buying stand-alone 'smart' devices, their
: are many printers and other devices you need to avoid, so watch out.
See above, not an issue.
Frank Durda IV <uhc...@nemesis.lonestar.org>|"The Knights who say "LETNi"
or uhclem%nem...@rwsystr.nkn.net | demand... A SEGMENT REGISTER!!!"
|"A what?"
or ...letni!rwsys!nemesis!uhclem |"LETNi! LETNi! LETNi!" - 1983
Jim Sellers
In article <32b2c145...@news.why.net>, dle...@dfwmm.net says...
Hello David,
I am really glad your WinModem works so well for you. I DO work in a
Microsoft Support Center dealing with modems. A large percentage of our
calls deal with Winmodems or RPI or TI chipset modems that will not connect
or once connected will not stay connected. I could not in good faith ever
recommend a WinModem to anyone based on problems I help people work with
every day.
Again, I am glad yours works well, but I can GUARANTEE not everyone has the
same experience that you have had. (ie...it works)
Happy holidays and enjoy,
Jim
baldeagl
On Sun, 15 Dec 1996 23:52:40 GMT, uhc...@nemesis.lonestar.org (Frank
Durda IV) wrote:
[snip]
I'm not even going to <begin> to try and argue with you, Frank. Your
knowledge of the industry is obviously far superior to mine. I just
have a few little questions.
>
>Is Microsoft better at what they do than other companies? In marketing,
>they are pretty good. In code development, so-so. There are reasons for
>that, such as the need to have 50% of the installed base upgrade every 18
>months in order to keep the cash flowing, which means you bring out
>software based on dates on calendars, not when it is really ready.
That's one view. But is it based upon facts? It seems to me there is
a logical flow of development...from bare bones DOS, to a DOS that
could actually do something, to a 16 bit Windows based upon that DOS,
to a 32 bit Windows based upon that DOS, to a 32 bit NT which divorces
itself from DOS.
Couldn't it just be the logical progression of an industry, rather
than a cynical attempt to extract money from consumers?
It seems to me that MS has often been accused of "piracy" of other's
ideas. Those ideas were brought about by the demand of the
marketplace, not by some cynical need to generate more revenue.
Memory management and drive compression are two that come easily to
mind.
[snip]
>
>Windows '95 slipped several times, until it started seriously affecting
>revenue, then a stake was driven in the ground for August 24, 1995.
>When June 13th 1995 came around, Beta testers were told "thanks", and
>be sure to re-file your unfixed bugs in September for the Nashville
>release.
Isn't this, again, a matter of perception? I rather think the "early"
release of Win 95 was actuated by the threat of IBM's OS/2 Warp
release. I think Gates saw that as a serious threat to his empire and
felt he had to release early to stay competitive. Isn't that an
equally fair interpretation?
[snip]
> Someone just needs
>to write a consumer-friendly UI. [referring to Unix]
And if someone does? Perhaps they will be the next Bill Gates.
Remember, at one time, IBM ruled this world. Perhaps tomorrow it will
be some guy who, today, is hacking code on Unix at the tender age of
14. That's life, isn't it?
[snip]
>
>: So UNIX is supposedly "immune" to viruses? So what?
>
>Well, it is a nice selling point. OS/2 seems to list that as a
>feature from time to time. It does mean that I don't have to blow
>time scanning and sniffing every file that happens to arrive on my
>system, which is about 65Meg a day. Face it, DOS/Windows/Machintosh
>just never envisioned the need for security of any sort. The
>first virus I ever heard of was on the Mac, not DOS/Windows.
So you blame Gates for not being able to see some 10 years into the
future? When he "wrote" DOS (or more accurately stole CP/M because
the author refused to license it to IBM), how was he supposed to know
that a GUI interface (which didn't even exist then) would have
problems with viruses? How was he to know that the dang thing would
be anything more than a stand alone machine used for spreadsheets and
word processing? Who the hell ever even thought there would be an
Internet, as we know it today?
How much security do you need for a machine that sits in a guy's
office, separated from the rest of the world, automating chore's that
used to take him hours to do? Remember, back then a network was a
mainframe and a bunch of terminals.
At the time, all Gates was trying to do was save his butt from losing
the IBM deal because he didn't have an OS to go with his software.
Blame the guy that wrote CP/M (just can't seem to remember his name
<g>) for letting the golden goose get away. After all, it's he who
would have control of the market today, not Gates, if he had just had
a little more sense of the potential.
>: It sure likes to do core dumps when it chokes on input.
>
>Hmm, bad application programming is responsible here.
IOW, bad apps exist for every platform? I'm shocked. <G>
What is Windows, if not a bunch of apps stacked on top of DOS?
When's the last time you crashed DOS?
[snip]
>
>The problem with Windows is its DOS-upbringing, where writing
>anything to anywhere was OK but "you should know better".
Again, aren't you blaming Gates for not being prescient? The guy may
be a good businessman and a so-so programmer, but don't hold him to
the standard of a prophet.
[snip]
>
>: If it was so perfect, then tell me why IA struggles
>: constantly with problems with it on their system?
>
>I suggest you investigate an ISP or other major Internet service that
>is using all-NTs or some other operating system and see what their
>uptime looks like, and how many customers they can put on the same
>amount of hardware. That directly equates to price per month that
>the ISP can offer without having totally crappy service.
I'm extremely happy with IA. I especially appreciate their candor
regarding system problems. Their downtime is minimal. The point is;
even the best equipment/software combinations are not infallible.
[snip]
>
>They shipped anyway. Hope you enjoy those bugs. I didn't.
I haven't found a bug yet.............that I enjoyed. Regardless of
the OS it exists on.
[snip]
>
>But even if you wrote the most marvelous operating system, with a
>friendly GUI that walked the dog and washed the cat, it would have
>to be compatible with the Windows API. There is too much inertia
>in the consumer market.
Ah, but didn't they say the same thing about Big Blue just a few years
back?
Give some guy with determination and grit a little time, so good luck
(which is really what got Gates where he is), some serendipitous
timing and a loyal following, and 20 years from now, people could be
talking about Windows in the past tense (just as they do now of Apple,
God rest its soul.)
>That same inertia is why the new digital VCR
>hasn't taken off. Someone decided to make the tape size different than
>VHS or 8mm, and the new players can't play existing tapes, of which
>there are millions. Bad move.
Aren't you comparing apples to oranges here? VCR equates to PC.
Tapes equate to software.
I doubt anyone will upset the PC apple cart. But there's no doubt in
my mind someone could come along and blow Gates and his gang out of
the water.
Can you envision someone making a better video tape that would run on
the same VCRs we have now?
Don Lee
Frank Durda IV <uhc...@nemesis.lonestar.org> wrote in article
<E2HBn...@nemesis.lonestar.org>...
> baldeagl (Adst...@toobad.bozo) wrote:
>
> : If it was so perfect, then tell me why IA struggles
> : constantly with problems with it on their system?
>
> I suggest you investigate an ISP or other major Internet service that
> is using all-NTs or some other operating system and see what their
> uptime looks like, and how many customers they can put on the same
> amount of hardware. That directly equates to price per month that
> the ISP can offer without having totally crappy service.
>
> I'd suggest MS-NBC, but they switched to UNIX (I don't know which)
because
> the Microsoft developers they shipped on-site could not make NT
> work with the load it was being given in the first week of operation.
>
> I believe there is a local ISP that uses NT, but I couldn't verify that
as
> they appear to be down right now. :-(
I've wondered if there were any ISP's using NT. Are you joking that there
is
one and its down or is there really one out there?
If you know of one I'd be curious who it is.
Don
baldeagl
On 16 Dec 1996 13:32:14 GMT, ro...@davids.psyberlink.net (Harrison
Bergeron) wrote:
[snip]
> >
> >: It sure likes to do core dumps when it chokes on input.
> >
> I generally find it easier to delete corefiles than
> reboot a frozen machine. In addition, with the sources
> you have the opportunity to fix the problem without
> waiting until someone else gets around to it.
Ahhhh, Harrison. Therein lies the problem.
You are an intelligent, Ph.D. physicist who likes to fiddle with
programming.
Try telling Joe Bob that he needs to get the source, fix the code and
recompile to get his 'puter to work, and see how fast the monitor
flies out the back window of his house into the already mounting pile
of useless artifacts of over-self-infatuated engineering types who
think the whole damn world should still use a slide rule.
I'm really glad that you can fix the source code to your liking.
Joe Bob could care less.
[snip]
> >
> I don't think article quality is what separates Scientific
> American from People Magazine but I'm not planning on
> seeing SA right next to the cash register in the supermarket
> any time soon.
No, but SA's what separates <you> from Joe Bob. :-)
[snip]
> >
> I've always wondered how anyone that can design the ciruitry
> in a VCR can miss the usefulness of a battery.
The same way you can miss the usefulness of a GUI interface and point
and click computing. <G>
Harrison Bergeron
baldeagl Adst...@toobad.bozo:
>On Sun, 15 Dec 1996 23:52:40 GMT, uhc...@nemesis.lonestar.org (Frank
>Durda IV) wrote:
>
>>
>>Is Microsoft better at what they do than other companies? In marketing,
>>they are pretty good. In code development, so-so. There are reasons for
>>that, such as the need to have 50% of the installed base upgrade every 18
>>months in order to keep the cash flowing, which means you bring out
>>software based on dates on calendars, not when it is really ready.
>
>That's one view. But is it based upon facts? It seems to me there is
>a logical flow of development...from bare bones DOS, to a DOS that
>could actually do something, to a 16 bit Windows based upon that DOS,
>to a 32 bit Windows based upon that DOS, to a 32 bit NT which divorces
>itself from DOS.
>
>Couldn't it just be the logical progression of an industry, rather
>than a cynical attempt to extract money from consumers?
>
>It seems to me that MS has often been accused of "piracy" of other's
>ideas. Those ideas were brought about by the demand of the
>marketplace, not by some cynical need to generate more revenue.
>
>Memory management and drive compression are two that come easily to
>mind.
>
things like HIMEM.SYS and EMM386.EXE. The 386 was released in
1985. It was capable of addressing 32 bit addresses. If you
also remember the stacker issue, Microsoft solved the the
lawsuit against them simply by countersuing stacker for using
a system call out of DOS that hadn't been published in any
of Microsoft's literature. After that they bought stacker. If
Microsoft borrows from others, fine, but they have no business
preventing others from doing what they've done. This is not
meeting the needs of the consumer. A company that has interest
in their customers as opposed to simply positioning themselves
to monopolize the industry does spend thier money on attornies
that live at the DoJ. They fix their software.
>And if someone does? Perhaps they will be the next Bill Gates.
>Remember, at one time, IBM ruled this world. Perhaps tomorrow it will
>be some guy who, today, is hacking code on Unix at the tender age of
>14. That's life, isn't it?
>
to reduce compatibility problems rather than increse
them inorder get rid of their competitors. Introducing
incompatibilities is fine if it serves a purpose not
addressed by what is available and widely in use. By
purpose I mean, purpose to the consumer, not a private
business agenda.
>So you blame Gates for not being able to see some 10 years into the
>future? When he "wrote" DOS (or more accurately stole CP/M because
>the author refused to license it to IBM), how was he supposed to know
>that a GUI interface (which didn't even exist then) would have
>problems with viruses? How was he to know that the dang thing would
>be anything more than a stand alone machine used for spreadsheets and
>word processing? Who the hell ever even thought there would be an
>Internet, as we know it today?
>
>How much security do you need for a machine that sits in a guy's
>office, separated from the rest of the world, automating chore's that
>used to take him hours to do? Remember, back then a network was a
>mainframe and a bunch of terminals.
>
1) The means to protect the system were available in 1985.
There is nothing available to day that wasn't available
then in terms of the chips ability to provide a protected
interface. I'll dig out my 386 reference manuals if you
want me to provide citations. Since DOS wasn't multi-user
they could have provided a 2 boot modes and reserved a
maitainence mode to allow for operations requiring any
privileged access, like allowing things to write to the
hard disk directly. That would have eliminated most of
the problems of installing a boot sector virus from an
infected floppy.
2) The GUI interface and viruses intrinsically have nothing to
do with each other. It's a matter of design.
3) Standalone machines are the machines with the most viruses.
Floppy to floppy transfer via an infected machine seems
pretty efficient.
4) The internet has little to do with virus transfer. It's not
internet's fault that allows people to obtain virus infected
programs more easily. It's the fault of the operating system
that trusts all programs without prejudice.
>At the time, all Gates was trying to do was save his butt from losing
>the IBM deal because he didn't have an OS to go with his software.
>
>Blame the guy that wrote CP/M (just can't seem to remember his name
><g>) for letting the golden goose get away. After all, it's he who
>would have control of the market today, not Gates, if he had just had
>a little more sense of the potential.
>
was fairly useful. This means the lesson to be learned
here is that you are better off looking for dupes that
write code without having an attorey for a partner to
make sure you own it when you are done than you are
write good code because the public doesn't care if you
are ethical and therefore ethics and doing a good job
get in the way of making money. That's the message you
send to all of the business men of the future (your kids).
>>: It sure likes to do core dumps when it chokes on input.
>>Hmm, bad application programming is responsible here.
>
>IOW, bad apps exist for every platform? I'm shocked. <G>
>
up evey would be competitor through litigation or threat
thereof, I have choices not to use the bad stuff and the
good drives out the bad in short order.
>What is Windows, if not a bunch of apps stacked on top of DOS?
>When's the last time you crashed DOS?
>
time I used dos.
>>The problem with Windows is its DOS-upbringing, where writing
>>anything to anywhere was OK but "you should know better".
>
>Again, aren't you blaming Gates for not being prescient? The guy may
>be a good businessman and a so-so programmer, but don't hold him to
>the standard of a prophet.
>
wrong. I hold him only to the standard he has bestowed
upon himself. If he stands up and says he is clueless
about the what will happen in five years, then I won't
hold to anything beyond that.
[...]
>>But even if you wrote the most marvelous operating system, with a
>>friendly GUI that walked the dog and washed the cat, it would have
>>to be compatible with the Windows API. There is too much inertia
>>in the consumer market.
>
>Ah, but didn't they say the same thing about Big Blue just a few years
>back?
>
invitation for everyone to come by for a party which will
feature a Bill Gates pinata filled with windows disks.
>Give some guy with determination and grit a little time, so good luck
>(which is really what got Gates where he is), some serendipitous
>timing and a loyal following, and 20 years from now, people could be
>talking about Windows in the past tense (just as they do now of Apple,
>God rest its soul.)
>
I give Billie Sol a little more credit for benevolence.
Harrison Bergeron
baldeagl Adst...@toobad.bozo:
>On 16 Dec 1996 13:32:14 GMT, ro...@davids.psyberlink.net (Harrison
>Bergeron) wrote:
>
>[snip]
>> >
>> >
>> I generally find it easier to delete corefiles than
>> reboot a frozen machine. In addition, with the sources
>> you have the opportunity to fix the problem without
>> waiting until someone else gets around to it.
>
>
>You are an intelligent, Ph.D. physicist who likes to fiddle with
>programming.
>
to despise programming. I often told people I though it
was the lowest possible form of entertainment in existence.
I did it because I had to. Our system admins didn't admin,
and pc software was terrible. Once I got over Ultrix and
discovered that all programming wasn't VMS/Ultrix/DOS/WIN
hell, I actually found it could be fun. As an aside, any
system admin reading this should realize that not admining
means someone will do it for you. If you're lucky they'll
let you know.
>
>Try telling Joe Bob that he needs to get the source, fix the code and
>recompile to get his 'puter to work, and see how fast the monitor
>flies out the back window of his house into the already mounting pile
>of useless artifacts of over-self-infatuated engineering types who
>think the whole damn world should still use a slide rule.
>
people want to fix their own code or not. I just don't
want to be prevented from doing it to my own. In the case
of an OS, they better damn well be able to fix it since in
many cases broken software translates to unauthorized access.
Do you want to have a hole in your OS at UTD that you can't
fix and the vendor doesn't deem important enough to get around
to this month? No system administrator has any excuse for
not being able to understand some basic programming.
>I'm really glad that you can fix the source code to your liking.
>
>Joe Bob could care less.
>
>
>The same way you can miss the usefulness of a GUI interface and point
>and click computing. <G>
>
I point and click when it is the most efficient way and most
intuitive way to do so - FOR ME. I can use the mouse to reboot
if I want (at least I'm pretty sure I can). I can do it MY way.
I don't want to have a computer that's useful to Joe Bob, if
it's not useful to me. If microsoft suceeds in beating all of
their legal problems, I may end up with Joe Bob's pc, running
Microsoft JoeBoB. Which means I will use it to write a JoeBoB
virus.
(Since you gave me the idea, should I list you in the license
when I unleash it ? ;)
>What do you want to fix today?
Joe Bramblett
>X-Newsreader: slrn (0.9.1.1 BETA UNIX)
9.2.0's out, and Davis has been posting with 9.2.1, so it should be
showing up soon.
In article <slrn5bcva...@davids.psyberlink.net>, Harrison Bergeron wrote:
>baldeagl Adst...@toobad.bozo:
> >
> >Memory management and drive compression are two that come easily to
> >mind.
> >
> What you fail to realize, is that there was never a need for
> things like HIMEM.SYS and EMM386.EXE. The 386 was released in
> 1985. It was capable of addressing 32 bit addresses.
Besides which, QEMM did a far better job of memory management in DOS 5
than anything M$ ever put in. Their MemMaker util in DOS 6 was nothing
but a slower way of typing control-alt-delete. I tried it on several
different machines, and only found one that didn't immediately reboot.
It froze.
--
Joe Bramblett
bo...@joeb.psyberlink.net
airnews...@airmail.net
bo...@127.0.0.1
In <slrn5bd2h...@joeb.psyberlink.net>, bo...@joeb.psyberlink.net (Joe Bramblett) writes:
>>X-Newsreader: slrn (0.9.1.1 BETA UNIX)
>9.2.0's out, and Davis has been posting with 9.2.1, so it should be
>showing up soon.
>
>In article <slrn5bcva...@davids.psyberlink.net>, Harrison Bergeron wrote:
>>baldeagl Adst...@toobad.bozo:
>> >
>> >mind.
>> >
>> What you fail to realize, is that there was never a need for
>> things like HIMEM.SYS and EMM386.EXE. The 386 was released in
>> 1985. It was capable of addressing 32 bit addresses.
>
>than anything M$ ever put in. Their MemMaker util in DOS 6 was nothing
>but a slower way of typing control-alt-delete. I tried it on several
>different machines, and only found one that didn't immediately reboot.
>
If you'll pardon the os snobbery, the above is a good reason to run os/2 if you _have_
to run dos software :-)
baldeagl
On 17 Dec 1996 11:23:26 GMT, ro...@davids.psyberlink.net (Harrison
Bergeron) wrote:
[snip]
> >
> I used
> I I
> I I
> I
> I
> I'm I
> I
> my
> I
> I
> FOR ME. I
> I I'm I I can do it MY way.
> I
> me.
> I
> I
See my answer to Joe Bramblett's post, in this same thread, for an
explanation of the above. <g>
What do you want to fix today?
baldeagl
On Tue, 17 Dec 1996 07:58:19 GMT, airnews...@airmail.net (Joe
Bramblett) wrote:
[snip]
>
>Quite a bit. If that guy has one braindead moment, he could end up
>wasting those hours trying to recover from it.
Of course, no one using (x)nix has EVER had a braindead moment. <g>
>
>Why do you think I do most stuff as bofh instead of root on my
>machine?
Ahhhh.....because you want to?
[snip]
>
>>Can you envision someone making a better video tape that would run on
>>the same VCRs we have now?
>
>No, but let Mic^H^H^H somebody's marketing department get ahold of it,
>and they'll tell you they have.
I see.......so now MS is responsible for all the hardware upgrades
you've had to do also, right? Let's see.......we could jump all over
the CD ROM guys. They've gone from 2X to 10X in about one year. Talk
about forcing their customer base to constantly upgrade.
And what about those nasty guys at Intel? No sooner do you get the
processor seated and they're announcing another faster one. And now
the buggers have announce the new MMX chips which are completely
incompatible with previous versions.
Boy, and you accuse MS of being greedy????
What about those nasty guys who came out with zip drives and made
floppys almost obsolete? We'll have to go after them too, dirty
buggers.
Man, the more I think about it, it's a damn conspiracy to rip us off.
The whole computer industry is in cohoots to steal us blind.
Man, better quit buying stuff now, before they <really> get their
hooks into us.
Irwin Sabath
Adst...@toobad.bozo (baldeagl) wrote:
<snip>
>Couldn't it just be the logical progression of an industry, rather
>than a cynical attempt to extract money from consumers?<
My vote is for cynicism --everyday.
Nice article, Paul.
Agree or disagree, even us technically-challenged have no difficulty
in understanding it.
--
Irwin
t.i.n.s.t.a.a.f.l.
David L. Cathey
In article <slrn5b3tv...@davids.psyberlink.net>, ro...@davids.psyberlink.net (Harrison Bergeron) writes:
>all 50 watts of, David Levin dle...@dfwmm.net came up with:
> >On Fri, 13 Dec 1996 13:52:41 GMT, uhc...@nemesis.lonestar.org (Frank
> > Just because some modems are finally using drivers (just
> >like almost every device that hooks up to your computer), doesn't mean
> >that it's 'BAD'.
>
> Yes it does. Why burden the cpu with something that can be
> done cheaply all by itself? This isn't done for performance,
> it's done for the cash it will generate. Since the can't keep
> plugging you for 200-300 on a modem that offers little over
> it's predecessor, they'll nickel and dime you to death on
> dubious "upgrades" in software. Anyone trying to improve their
> computing speed is going to offload as much work as they can.
I agree. Over the last 10-15 years, we've gone from dumb devices
to intelligent devices (boxes with their own brains). The reason is to
offload the main CPU, rather than interupt it to death with every little
character of data.
Now, if they built a Flash-RAM modem, one with it's own brain but
you uploaded software to it (much like many terminal servers and routers
do today), then THAT would be the ticket. Then you could take your
modem to your Win95 laptop, upload the new release, then walk it over to
your Livingston Portmaster or whatever, and you're ready to go. Advantage
of PROM's, but the easy of software upgrades.
> One more giant leap into the technological dark ages. (Does
> it use 12ax7's too?)
Is there a PDP-8 driver for that modem? If someone knows where
a paper tape for that is, let me know...
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
David L. Cathey |Inet: dav...@montagar.com
Montagar Software Concepts |Fone: (972)-578-5036
P. O. Box 260772, Plano, TX 75026-0772 |http://www.montagar.com/~davidc/
baldeagl
On Tue, 17 Dec 1996 17:11:36 GMT, tdsh...@airmail.net (T.D. Shadow)
wrote:
[snip]
>
>Neither he nor BG writes code. They are simply CEOs. If you want to
>argue who's company is better, why bring up paychecks or houses?
Penis envy; that's why!
(You know....mine is bigger than yours......)
baldeagl
On 17 Dec 1996 15:20:01 GMT, ro...@davids.psyberlink.net (Harrison
Bergeron) wrote:
[snip]
>Like microsoft's marketing, it's completely misleading.
The essence of good communication is brevity.
You would do well to apply the principle liberally.
You missed or ignored the point entirely. Why discuss it further?
Joe Bramblett
In article <32bba916...@news.airmail.net>, baldeagl wrote:
>On Tue, 17 Dec 1996 07:58:19 GMT, airnews...@airmail.net (Joe
>Bramblett) wrote:
>
>Of course, no one using (x)nix has EVER had a braindead moment. <g>
I do every now and then, but simply by doing most stuff as something
other than root, I can avoid doing anything really nasty. Anybody that
sits down in front of a DOS machine can do a del \*.*
>I see.......so now MS is responsible for all the hardware upgrades
>you've had to do also, right? Let's see.......we could jump all over
>the CD ROM guys. They've gone from 2X to 10X in about one year. Talk
>about forcing their customer base to constantly upgrade.
Ok, then what about the old CDROM I've got here. No drivers for
anyhthing but MS.
baldeagl
On 18 Dec 1996 01:01:14 GMT, bo...@joeb.psyberlink.net (Joe Bramblett)
wrote:
[snip]
>
>I do every now and then, but simply by doing most stuff as something
>other than root, I can avoid doing anything really nasty. Anybody that
>sits down in front of a DOS machine can do a del \*.*
They can also do an a:setup :-)
>
>>I see.......so now MS is responsible for all the hardware upgrades
>>you've had to do also, right? Let's see.......we could jump all over
>>the CD ROM guys. They've gone from 2X to 10X in about one year. Talk
>>about forcing their customer base to constantly upgrade.
>
>Ok, then what about the old CDROM I've got here. No drivers for
>anyhthing but MS.
So write one yourself. You're the one running Linux, because you
don't want someone else's writing your code for you. Shouldn't take
you too long to write a driver code and make the baby hum on your box.
<g>
Harold Stevens
In article <32b92cac...@news.airmail.net>, Adst...@toobad.bozo (baldeagl) writes:
[Snip...]
|> You missed or ignored the point entirely. Why discuss it further?
Umm...before bailing, could you pass the Grey Poupon, please? :)
Regards, Weird
wy...@ti.com
bo...@127.0.0.1
In <32b557f6...@news.airmail.net>, tdsh...@airmail.net (T.D. Shadow) writes:
>
>>Aside from the issue of typesetting, there is composition. While I
>>don't know specifically what Stephen King, John Grisham, or even
>>Howard Stern pound their gazillion-selling novels into, I have a much
>>easier time envisioning Word or WordPerfect than vi or emacs.
>
>Stephen King uses the dedicated (Word Processor) machine made by WANG
>in 1982. Howard Stern uses OS/2, but I'm not sure what wp-software. I
>can safely say it's NOT a MS product. He has more distain for it than
>even Harrison :) !!
if I wasn't terribly attached to os/2, I'd quit using it after this :-)
FWIW, my favorite word processor is vi for unix or qedit for dos or os/2.
Simplicity, simplicitiy, simplicty :-)
William Rumbley
In article <1996Dec17.142315.22751@montagar>, dav...@montagar.com (David L. Cathey) writes:
|>
|> Now, if they built a Flash-RAM modem, one with it's own brain but
That would be a USR Courier. :-)
Nice box.
--
wrum...@nortel.com | Any opinions expressed are my own.
Nortel Wireless Systems | ''''''''''''''''''''''''''''''''''
Richardson, TX 75083-3871 | There is a spirit in man.
Joe Bramblett
In article <32c07765...@news.airmail.net>, baldeagl wrote:
>>
>>Ok, then what about the old CDROM I've got here. No drivers for
>>anyhthing but MS.
>
>So write one yourself. You're the one running Linux, because you
>don't want someone else's writing your code for you.
Uhhh...I don't have a problem with someone else writing it, I just
want them to show it to me before I run it on my machine. I trust my
friends to unload the shotguns after a day of hunting, but that
doesn't mean I'll toss them in the car without checking each one
myself.
> Shouldn't take
>you too long to write a driver code and make the baby hum on your box.
><g>
A) It's a single speed, it doesn't hum, it coughs.
B) Proprietary interface. I doubt they'd even tell me what the pinout
is so I could fix the controller card (If the company even has
anybody who still knows)
ML
David Levin wrote:
>
> On Thu, 12 Dec 1996 22:53:46 -0600, Michael LaPorta <mlap...@gte.net>
> wrote:
>
> >Guess it's a shame that only 5% of the people in the US can handle the
> >33,6 connect rate ( 3429 symbol rate ). Don't hold your breath on the
> >57,6......
>
> But most people get close to the maximum of the technology (24k, 26.4k,
> 28.8k, 31.2k). I see no reason why people won't get reasonable close to
> the 56K connect rate.
But will the majority of the "got a computer for Christmas, and I expect
my 33,600 bps modem to connect at 33,600" accept this?
This is kinda like in CA where the bandwith is rolled off so severly on
the lines, that ISDN users can only get 56k, instead of 64k on one
channel.
but all this was covered a couple of days ago.....
ML
Harrison Bergeron
Frank Durda IV uhc...@nemesis.lonestar.org:
>
>On a side note, how much longer can Microsoft get away with calling it
>Windows NT? Remember what the "N" stands for? "New". According
>to FTC regulations, a product can only be marked as "New" for
>six months after its introduction. (This rule was originally targeted
>manufacturers who left the "New Formula" and other nonsense on their
>packaging year after year.) So this product really needs to now be
>called "Windows T" or "Windows OT" or "Windows Full Strength" or
>"Windows, NOT Bob", or something like that. Call Janet Reno. :-)
>
According to cutler: V M S +1 - > W N T
besides, there is always verioning you to death.
--
dav...@davids.psyberlink.net | Those that do not understand Unix are
Steinberger: | comdemned to reinvent it - poorly.
State of the Instrument | -- unknown
Irwin Sabath
uhc...@nemesis.lonestar.org (Frank Durda IV) wrote:
>
>: Larry Ellison is the second richest man.
>
>In the world? I thought it was a guy in the Orient, followed by
>some investor from mid-America whose name escapes me (Buffet?).
>Perhaps you mean in the PC business.<
Relying on memory (always hazardous): Buffet was No .2 in this year's
Fortune 400 (U.S. only). Ellison was up there as well, but not close
to Gates.
>
>
>: No one bitches about this ego-centric billionare with a mansion.
>
>I guess that is because Orson Wells is dead. :-) <
William Randolph Hearst. (Rosebud)
--
Irwin
t.i.n.s.t.a.a.f.l.
baldeagl
On Thu, 19 Dec 1996 04:55:30 GMT, uhc...@nemesis.lonestar.org (Frank
Durda IV) wrote:
>baldeagl (Adst...@toobad.bozo) wrote:
>: Of course, no one using (x)nix has EVER had a braindead moment. <g>
>
>Well, there was that time I renamed the /shlib directory on a SCO system.
>The Windows equivalent would be renaming the \WINDOWS directory and
>then trying to do stuff or trying to reboot. Not pretty. Everything dies
>instantly. Hard to bounce back from. :-(
You???? I can hardly believe it. Was that after one of those 36
hours without sleep troubleshooting marathons?
Larry Weiss
Frank Durda IV wrote:
>
> It even drew the attention of Doonesbury in the December 9, 10 and 11th
> strips. What I want to know is who is the other person depicted in the
> strip who is trying to build the bigger house to out-do Gates?
Re: news:clari.living.comics.doonesbury
If you find out, let me know! I'm puzzled too.
Frank Durda IV
[5] On Thu, 19 Dec 1996 04:55:30 GMT, uhc...@nemesis.lonestar.org
[5] (Frank Durda IV) wrote:
[5]Well, there was that time I renamed the /shlib directory on a SCO system.
[5]The Windows equivalent would be renaming the \WINDOWS directory and
[5]then trying to do stuff or trying to reboot. Not pretty. Everything dies
[5]instantly. Hard to bounce back from. :-(
[6]baldeagl wrote:
[6]You???? I can hardly believe it. Was that after one of those 36
[6]hours without sleep troubleshooting marathons?
I vaguely recall a manager standing behind me at the time, pacing around
saying stuff like "Don't spend any time on that, it isn't part of your
official project list, but we need it fixed right now, but we should get
rid of this system and use a DOS-based solution, but we require UUCP to mail
things to and from Microsoft, and aren't you done yet? Do you know
how much money we are losing while this system is down? No, I can't
authorize the purchase of a replacement hard 700Meg hard drive, you'll
just have to do with the space you still have (80Meg). And while you are in
there, can you add these newsgroups and set up connections to these
companies? Geez, if it takes this long, I'll assign an intern to delete
stuff for you so you can work on your official projects...arrrrggh gurgle."
The jury let me off. :-)
Frank Durda IV <uhc...@nemesis.lonestar.org>|"Hmm, is Microsoft scared of UNIX?
or uhclem%nem...@rwsystr.nkn.net | Then why did they change the
| application license to bar using
or ...letni!rwsys!nemesis!uhclem | Microsoft apps on Win emulators?"
Frank Durda IV
baldeagl (Adst...@toobad.bozo) wrote:
: Of course, no one using (x)nix has EVER had a braindead moment. <g>
Well, there was that time I renamed the /shlib directory on a SCO system.
The Windows equivalent would be renaming the \WINDOWS directory and
then trying to do stuff or trying to reboot. Not pretty. Everything dies
instantly. Hard to bounce back from. :-(
That's when you find out how many executables SCO shipped that were
statically linked in that release. (answer: One. The kernel).
Frank Durda IV
>There is no pedestal. We are beyond that. He is building the
>equivalent of the Randolph Hearst mansion.
T.D. Shadow (tdsh...@airmail.net) wrote:
: I'm neither a friend nor foe of Big Billy the Billionare, but one
: thing ALWAYS seems to irk me.
:
: Who gives a damn about his house? How petty is *that*?
Dunno, but clearly he is trying to make a statement about something.
He spent most of his time on Leno the other night talking about the
house instead of the paperback version of his book which now comes
with a CD-ROM that actually works, unlike the one that came with the
hardcover version of "The Road to My House", or something like that. :-)
It even drew the attention of Doonesbury in the December 9, 10 and 11th
strips. What I want to know is who is the other person depicted in the
strip who is trying to build the bigger house to out-do Gates?
: Larry Ellison is the second richest man.
In the world? I thought it was a guy in the Orient, followed by
some investor from mid-America whose name escapes me (Buffet?).
Perhaps you mean in the PC business.
: No one bitches about this ego-centric billionare with a mansion.
I guess that is because Orson Wells is dead. :-)
Or maybe his house has less square footage than a football field
(Bills is larger) and so Larrys house won't make it into the records
books. :-)
Frank Durda IV <uhc...@nemesis.lonestar.org>|"If you make it go out of
or uhclem%nem...@rwsystr.nkn.net | focus, it will become clearer."
| - Tandy/AST Director of
or ...letni!rwsys!nemesis!uhclem | Software Development.
Frank Durda IV
Frank Durda IV <uhc...@nemesis.lonestar.org> wrote in article
[2]I'd suggest MS-NBC, but they switched to UNIX (I don't know which)
[2]because the Microsoft developers they shipped on-site could not make
[2]NT work with the load it was being given in the first week of operation.
[2]
[2]I believe there is a local ISP that uses NT, but I couldn't verify that
[2]as they appear to be down right now. :-(
"Don Lee" (dl...@esd.dl.nec.com) wrote:
[13]I've wondered if there were any ISP's using NT. Are you joking that
[13]there is one and its down or is there really one out there?
[13]If you know of one I'd be curious who it is.
Yes, there is at least one ISP in the Dallas Fort Worth Metroplex trying
to run everything on NT systems. Granted they do have digital
modems/terminal servers instead of using modems in the NT boxes, but mail,
news, FTP, local WWW, authentication and all other functions are on on
NT systems.
I see no point in identifying them, since that's their choice and if
they can make it work, more power to them. In fact, Microsoft would
be interested to locate anybody that has solved this enigma. There
are a few other ISPs across the country that also claim to be all-NT,
and some just have select customer functions on NT. Others may have the
odd NT box somewhere doing billing or some such off-the-front-line
operations, but no NT in front-line duties.
And yes, this ISP was down when I tried to contact their FTP system
when I wrote the original post. I don't know why they were down and
am not concerned. I am not a customer.
I personally don't think NT should be used in these types of operations
where the immense overhead of the GUI is neither needed or used. Why
waste 20+ Meg of RAM just on a GUI layer that 99.99999% (7 nines) of the
people utilizing the system will never use or even see, and the sysadmins
will only use the GUI to shut down, reboot, and fix those pesky
"the %s log is full" messages, or possibly start a system backup.
For sites that run "dark" computer rooms, the GUI and associated baggage
is a total waste since the machines are administered remotely anyway.
Apart from reliability, all Internet users care about is how fast the
equipment they are connected to transfers packets, and retrieves and stores
data on disk. That's it. There is no GUI friendliness factor. No Windows
API needed by the customer applications. All that stuff is wasted here
and just adds potential points of failure to an environment that already
has enough things that can go wrong.
And that is the basic problem with NT in this particular application.
Because of the Win/GUI overhead, it takes more machines, more horsepower
and more RAM to provide similar network-based performance and customer
count volumes when using NT vs other operating systems on the same hardware.
Look at www.freebsd.org sometime. This box consists of a single Pentium
(possibly PRO) running (at last check) at 166MHz, and it supports a
staggering 5,000 simultaneous FTP connections plus WWW access, and has
three T3s plugged into that computer to give it 100Mbit aggragated Internet
connectivity. No one has shown me a NT box that can get anywhere near this
level of brute packet and disk performance, that didn't cost at least
$100,000 more. If you FTP anonymous to www.freebsd.org, it will give you
a precise description on the hardware setup they use in the greeting
message.
NT is undoubtedly better at other things, but Internet ISP operations
doesn't seem to be its strong suit.
- - - - - - -
On a side note, how much longer can Microsoft get away with calling it
Windows NT? Remember what the "N" stands for? "New". According
to FTC regulations, a product can only be marked as "New" for
six months after its introduction. (This rule was originally targeted
manufacturers who left the "New Formula" and other nonsense on their
packaging year after year.) So this product really needs to now be
called "Windows T" or "Windows OT" or "Windows Full Strength" or
"Windows, NOT Bob", or something like that. Call Janet Reno. :-)
Frank Durda IV <uhc...@nemesis.lonestar.org>|"You see? It's not my
or uhclem%nem...@rwsystr.nkn.net | fault! It's this
| 'General Protection' persons
or ...letni!rwsys!nemesis!uhclem | fault!" - Ratbert
Frank Durda IV
>Look at www.freebsd.org sometime. This box consists of a single Pentium
>(possibly PRO) running (at last check) at 166MHz, and it supports a
>staggering 5,000 simultaneous FTP connections plus WWW access, and has
>three T3s plugged into that computer to give it 100Mbit aggragated Internet
>connectivity. No one has shown me a NT box that can get anywhere near this
>$100,000 more.
ro...@127.0.0.1 wrote:
: It is a pro. The last time I checked it was 150MHz or something like that
: with multiple 2940 controllers.
I can believe that. It has something like 72GB of hard disk space,
plus CD-ROMs. You can only get so many drives per SCSI card.
Bottom line, this FreeBSD machine is doing tons of work for being one
machine with one CPU. Please show me a NT box with any number of
processors performing equally well under similar Internet-type load
that costs less than $100K. I don't know of any such configuration,
and you can buy a lot of hardware for under 100K.
Frank Durda IV <uhc...@nemesis.lonestar.org>|"The Knights who say "LETNi"
or uhclem%nem...@rwsystr.nkn.net | demand... A SEGMENT REGISTER!!!"
|"A what?"
or ...letni!rwsys!nemesis!uhclem |"LETNi! LETNi! LETNi!" - 1983
Frank Durda IV
: Frank Durda IV uhc...@nemesis.lonestar.org:
>On a side note, how much longer can Microsoft get away with calling it
>Windows NT? Remember what the "N" stands for? "New". According
>to FTC regulations, a product can only be marked as "New" for
>six months after its introduction. (This rule was originally targeted
>manufacturers who left the "New Formula" and other nonsense on their
>packaging year after year.) So this product really needs to now be
>called "Windows T" or "Windows OT" or "Windows Full Strength" or
>"Windows, NOT Bob", or something like that. Call Janet Reno. :-)
>
Harrison Bergeron (ro...@davids.psyberlink.net) wrote:
: According to cutler: V M S +1 - > W N T
: besides, there is always verioning you to death.
Uh, yeah. Interesting bit of revisionist history. Totally false,
of course. I still have documents from Microsoft as well
as advertising that clearly shows the product to be
Windows New Technology, or Windows NT for short.
David L. Cathey
In article <59b243$h...@library.airnews.net>, bo...@127.0.0.1 writes:
>>If Microsoft's own people couldn't get NT to do the job on MS-NBC, and on
>>their own FTP systems where both had to switch to non-Microsoft operating
>>systems, that should say a lot. I want my machines to be ruled, not
>>rule me.
>
>I was aware of this some time ago (and laughed hysterically)...I'd love to
>know exactly what operating system they wound up using. I'll ROFL
>if it's Solaris.
Actually, several of the systems they use run OpenVMS. Which
geneology of WNT makes some sense. After all, they can't afford some
of these clusters to go down.
ro...@127.0.0.1
In <E2qDn...@nemesis.lonestar.org>, uhc...@nemesis.lonestar.org (Frank Durda IV) writes:
>>Look at www.freebsd.org sometime. This box consists of a single Pentium
>>(possibly PRO) running (at last check) at 166MHz, and it supports a
>>staggering 5,000 simultaneous FTP connections plus WWW access, and has
>>three T3s plugged into that computer to give it 100Mbit aggragated Internet
>>connectivity. No one has shown me a NT box that can get anywhere near this
>>level of brute packet and disk performance that didn't cost at least
>>$100,000 more.
>
>ro...@127.0.0.1 wrote:
>: It is a pro. The last time I checked it was 150MHz or something like that
>: with multiple 2940 controllers.
>
>I can believe that. It has something like 72GB of hard disk space,
>plus CD-ROMs. You can only get so many drives per SCSI card.
The 2940UWs that I use will do the full 15 drives per card. 9gb*15=135.
take a machine with 4 pci slots and you could have 540gb attached to
one box :-) One of those 8 pci slot alr boxes would make a nifty
ftp server.
>
>Bottom line, this FreeBSD machine is doing tons of work for being one
>machine with one CPU. Please show me a NT box with any number of
>processors performing equally well under similar Internet-type load
>that costs less than $100K. I don't know of any such configuration,
>and you can buy a lot of hardware for under 100K.
Yup. And it costs $20 :-) Shame Oracle doesn't run under it :-/
Mark A Harrison
baldeagl (Adst...@toobad.bozo) wrote:
: That's one view. But is it based upon facts? It seems to me there is
: a logical flow of development...from bare bones DOS, to a DOS that
: could actually do something, to a 16 bit Windows based upon that DOS,
: to a 32 bit Windows based upon that DOS, to a 32 bit NT which divorces
: itself from DOS.
: Couldn't it just be the logical progression of an industry, rather
: than a cynical attempt to extract money from consumers?
Actually, the progression went bare bones DOS, Unix, AT&T won't
give good licensing terms, scratch Unix, DOS that could actually
do something (which version would that be?).
Frank could fill you in on the historical details, he was there...
: I doubt anyone will upset the PC apple cart. But there's no doubt in
: my mind someone could come along and blow Gates and his gang out of
: the water.
Steve Jobs, call your (new) office... :-)
Mark.
Harrison Bergeron
Mark A Harrison mhar...@utdallas.edu:
>
>Actually, the progression went bare bones DOS, Unix, AT&T won't
>give good licensing terms, scratch Unix, DOS that could actually
>do something (which version would that be?).
>
decree prevented them from marketing it, so the provided it
to universities. They didn't advertise, market or support
it.
>
>Steve Jobs, call your (new) office... :-)
If there's a hell, Bill will be stuck on an island with
a laptop and lotus notes and an infinite supply of
batteries.
Bryce Wray
In article <slrn5btt9...@davids.psyberlink.net>, dav...@davids.psyberlink.net wrote:
> Mark A Harrison mhar...@utdallas.edu:
> >
> >Actually, the progression went bare bones DOS, Unix, AT&T won't
> >give good licensing terms, scratch Unix, DOS that could actually
> >do something (which version would that be?).
> >
> It wasn't AT&T's choice to do much of anything. A consent
> decree prevented them from marketing it, so the provided it
> to universities. They didn't advertise, market or support
> it.
> >
> >Steve Jobs, call your (new) office... :-)
>
> If there's a hell, Bill will be stuck on an island with
> a laptop and lotus notes and an infinite supply of
> batteries.
>
OK, fine; just as long as it's a really HOT island!
______________________________________________________________________
I hid from spammers until I discovered the joys of flaming 'em.
Bryce Wray ** bw...@gte.net
http://home1.gte.net/bwray/index.htm
** Spammers, beware: **
Pursuant to US Code, Title 47, Chapter 5, Subchapter II, Sec. 227, any
and all unsolicited commercial e-mail sent to this address is subject
to a download and archival fee in the amount of $500 (U.S.); e-mailing
denotes acceptance of these terms.
Frank Durda IV
Mark A Harrison (mhar...@utdallas.edu) wrote:
[55]Actually, the progression went bare bones DOS, Unix, AT&T won't
[55]give good licensing terms, scratch Unix, DOS that could actually
[55]do something (which version would that be?).
[55]
[55]Frank could fill you in on the historical details, he was there...
Um, well, PC-DOS appears in 1981, well after UNIX can claim to have
existed. Granted that PC-DOS (what we call MS-DOS today or just 'DOS
if you believe that all operating systems come from Microsoft) has
roots in a CP/M-86 port done by some guys in a garage in Seattle (which
Bill bought), which is similar (down to the JMP 5 vector) to CP/M, which
dates back to around 1977 or 1978.
Even if you take the 1977 date, the same year the Apple and the TRS-80
appeared, UNIX was already around, at least in the confines of AT&T.
Tandys original DOS for its Model I was written by a guy called
Randy Cook. Reasonable in design, but so-so in execution, its interfaces
and file system structures were quickly copied by several other new
companies, like Apparat, Logical Systems, Misosys, and others, who
sold on the promise of a stable operating system with fewer bugs, and
more features than what Tandy offered. Tandy and Cook separated on
rather violent terms, and Tandy essentially threw away all of his work and
started over for subsequent machine models. The new machines were
always different anyway, as the innovation-killing "PC-compatible"
notion hadn't arrived on the scene yet.
Apple had their disk operating system (DOS) by 1978, but I don't know the
details of its development.
And the Altair from MITS, beat Tandy and Apple by three years for
the platform and almost two years for a disk operating system. But the
time MITS decided to start selling the systems assembled and improved the
peripherals they were buying, the Apple and TRS-80s systems were out
and being sold in far more places than a PO box in New Mexico. And
there was SWTPC, IMSAI, Heathkit and a whole series of other microcomputer
makers, each with their own operating system, or CP/M.
But why wasn't AT&T selling UNIX? Well, there was this guy called President
Lyndon Johnson. On his last day as president, he signed the papers
that caused the government to file suit against IBM (and indirectly AT&T)
for being a monopoly, eg a big company completely controlling markets
and using predatory and dumping marketing methods to kill competition...
sorta like what Microsoft is today.
Anyway, one of the things that came from the years of litigation that
followed was that IBM agreed to never enter the telecommunications
industry, and AT&T agreed to never enter the computer industry.
For some reason, in exchange for this agreement, it was decided to let
these two companies continue to dominate the market segments they already
controlled, but prevent them from entering other markets. Apparently the
government of the day just wasn't ready to break up AT&T or Big Blue.
For now.
Therefore, UNIX, which looked a lot like a computer product, could not
be commercially sold by AT&T. They could make it available to
other educational and research firms, much as they license patents and
other trade secrets. This is where Berkeley BSD UNIX got its start.
Around 1981, AT&T finally got a bit bolder and licensed UNIX to others for
resale in the commercial sector (Microsoft XENIX, Charles River), who sold
it to Tandy and anyone else willing to take this operating system. Back
then, you could get the AT&T V32 license for just $5,000 (that's what
Tandy paid), and the BSD source license for $500. You needed both if
you were working on XENIX because Microsoft had put in bits of both.
Yeah, there were royalties per copy to Microsoft of about $85, but that
was it. Six years later, an AT&T System V Release 4 source license for
a single computer cost $50,000.
The agreement that blocked AT&T and IBM from invading each others markets,
was dissolved by Judge Green when the AT&T system was broken up a decade
later. AT&T argued that the constraint should be lifted since they were
no longer a monopoly, and IBM argued that if AT&T was released from
the deal, they should be released too. IBM argued all these PC makers
(Compaq, Franklin, Tandy, Zenith etc) were providing them with ample
competition even though it was only in one market of their business.
AT&T almost immediately had PCs on the market (and bought NCR when their
first models flopped), and IBM went out and bought Rolm to get a start
in the telecom business. Despite the activity, neither has done very
well in the others territory so far.
In my opinion, the thing that gave UNIX it's biggest break was the
Berkeley UNIX releases in the '80s. Berkeley took a system that
ran on PDP-11s, and did a port to the virtual memory environment of the
VAX-11/780, a reasonably modern computer that many schools and businesses
had or could afford, then they set their undergrads and grad students the
thousands of tasks of fixing the broken, incomplete or simply missing
pieces. The result over time was that these additions provided a far
more complete and usable operating system and user interface than what
came out of AT&T.
For those of us marketing UNIX based on AT&T V7, then System III, then
System V, each new AT&T release meant re-applying many of the functions and
features of the BSD releases that AT&T failed to integrate into their base
code. As an example, without these non-AT&T features, the only editor UNIX
would provide would be ed. (For DOS-only readers, think of using EDLIN
for writing all of your letters and software.) TCP/IP and the Internet as
we know it would not exist. Berkeley got DARPA grants to design and add
networking to BSD UNIX, and that funding helped keep development at Berkeley
alive until around four or five years ago. I remember getting an
announcement that the last computer in that group was being shut down
and sent over to be used by the campus police department.
BSD Unix, while popular as the development and research platform for many
companies, only ran on a tiny number of computer models, and no one had
bothered to expend the effort to migrate it onto other platforms, such
as Sparc or Intel-based systems. The government was perfectly willing
to continue buying and using older and more expensive platforms, so there
wasn't much money provided to research making BSD run on other platforms.
This was the case until 1991 when a primitive port appeared (BSD386) in
a series of articles in Dr. Dobbs magazine. That work eventually spawned
BSDi, a commercially available system supported by some of the people
who created the original VAX BSD system, who were no longer needed at
Berkeley since the grant money ran out. In addition, FreeBSD, NetBSD,
and OpenBSD are all based on BSD386 and are freely available.
Linux, is based on the manuals for System V Release 4, and not the code.
It also includes a lot of the BSD enhancements, although by the time
Release 4 came out, AT&T finally started including a lot of the BSD
code in the product they sold. Because Linux is based on the manuals
and not the actual code, it doesn't have the robustness of some
of the other free operating systems in certain areas like networking.
(A lot of the code has been in use for fifteen years, so most of the bugs
have been found.) Despite this, Linux does have a substantial following,
particularly in Europe.
The fact that AT&T started adding the BSD stuff to their distribution hurt
the AT&T legal case years later when they sued over the BSDi and FreeBSD
UNIX systems claiming they contained AT&T code. The University of California
countersued, claiming AT&T was using the Berkeley copyrighted code and
documentation in the AT&T product without permission or proper credit.
UCB asked the courts to order all AT&T product containing BSD material
to be recalled and destroyed, including from sublicensees, such as SCO,
Digital, IBM, Data General, Kodak, etc. Oops.
AT&T sorta goofed, forgot to actually file their own copyrights, then tried
to pre-date them when they did file them, tried to claim copyright on some
clearly non-copyrightable stuff, such as "#define TRUE 1 and
#define FALSE 0" (really! but so did Microsoft), and AT&T apparently hadn't
kept too close attention of how much of the original AT&T UNIX code had
been replaced by ten years worth of undergrad projects and voluntary
contributions from people all over the world. When it got down to it in
court, only five modules (parts of five files) were agreed to still contain
code that AT&T could call its own out of thousands of files in the UNIX
distribution. Those five modules quickly were re-written with non-AT&T
code, and all was well. AT&T agreed to provide proper credit in the
boot up messages and manuals for the Berkeley code. Thus, all the legal
issues were resolved (essentially a lengthy and expensive stalemate was
the result), and all of these systems current versions are completely legal.
FreeBSD 2.0 was the first "clean" free releases.
In general, no one has really tried to market UNIX to consumers or
put the effort into making it consumer friendly. Originally, it was
because the cost of the minimum hardware put it out of the consumers
reach. Some people point to "X" as a better GUI that makes UNIX easier
to use, but it seems to make it easier only for people already familiar
with computer operational concepts. It's primary goal still seems to
have been to simply allow graphical applications to be used under UNIX,
not to be more user friendly than your typical Bourne shell. X is certainly
less obvious than Microsoft Windows from the beginners point of view.
At least there are a few words on the screen to fall back on for guidance
in Microsoft Windows, like "Help".
SCO attempted to make the installation process easier, but managed to annoy
experienced users, and drown beginners in a endless series of menus
that still assume you know when to press F4 or some other key to
get something to happen. Plus, once installation is done, you have
the same UNIX environment as always.
Tandy did a migration shell for XENIX, which provided the same TRSDOS
commands and syntaxes. Despite wanting to get rid of this semi-stable
piece of software over the years, we consistently found a small number
of vocal (and fixed in their ways) people who wanted trsshell included
in subsequent operating system releases.
I don't see a user-friendly shell on the horizon for UNIX, apart from
the clones of Windows 3.x and Windows '95 GUI, and now that Microsoft
has just changed the license for Microsoft applications to forbid their
use on non-Microsoft operating systems (this clause sounds a tad
monopolistic, doesn't it?), it is illegal for someone to run Word '97 on
WABI or WINE, or one or the other Windows emulations than can be used
on one of the BSD operating systems or Linux. Something like telling
a resturant they can't give a customer a Coke in a glass that has a Pepsi
logo on it. Personally, I don't think this would hold up in court, but
Microsoft has more money than we all do.
Mz. Reno, where are you? Still looking for monopolistic activities?
Anyway, UNIX is probably destined to remain a back-room platform, getting
the job done out of sight and being a persistent thorn in the side of the
monopolies.
Frank Durda IV <uhc...@nemesis.lonestar.org>|"Ah, is the whittle bitty
or uhclem%nem...@rwsystr.nkn.net | Microsoft afrayed of those nasty
| API emulators? Bad emulators!
or ...letni!rwsys!nemesis!uhclem | Don't you go scaring Bill!"
(c) 1996, ask before reprinting.
Samuel S. Thomas
In article <59lt8t$p...@library.airnews.net>, ro...@127.0.0.1 says...
>
>In <E2qDn...@nemesis.lonestar.org>, uhc...@nemesis.lonestar.org (Frank
Durda I
>V) writes:
>>>Look at www.freebsd.org sometime. This box consists of a single Pentium
>>>(possibly PRO) running (at last check) at 166MHz, and it supports a
>>>staggering 5,000 simultaneous FTP connections plus WWW access, and has
>>>three T3s plugged into that computer to give it 100Mbit aggragated Internet
>>>connectivity. No one has shown me a NT box that can get anywhere near this
>>>level of brute packet and disk performance that didn't cost at least
>>>$100,000 more.
>>
>>ro...@127.0.0.1 wrote:
>>: It is a pro. The last time I checked it was 150MHz or something like that
>>: with multiple 2940 controllers.
from the README (amazing how little readership a file with this name can get)
file at ftp.cdrom.com:
Welcome to wcarchive.cdrom.com, a 150Mhz Pentium Pro system with 512MB of
memory and 72GB of hard disk storage. This server is running FreeBSD 2.1.0
(The latest version of FreeBSD is available in pub/FreeBSD).
512MB of RAM!! hmm, I thought 640K was enough for anybody!?!?!?
>>I can believe that. It has something like 72GB of hard disk space,
>>plus CD-ROMs. You can only get so many drives per SCSI card.
>
>The 2940UWs that I use will do the full 15 drives per card. 9gb*15=135.
>take a machine with 4 pci slots and you could have 540gb attached to
>one box :-) One of those 8 pci slot alr boxes would make a nifty
>ftp server.
4 slots=405GB...9GB*15*3=405GB
add one PCI 100B-TX ethernet card, and you're ready to go.
>>Bottom line, this FreeBSD machine is doing tons of work for being one
>>machine with one CPU. Please show me a NT box with any number of
>>processors performing equally well under similar Internet-type load
>>that costs less than $100K. I don't know of any such configuration,
>>and you can buy a lot of hardware for under 100K.
It is also their www server. It has sendmail and POP3 Daemons running,
too. Adding a newsfeed would be trivial.(for people who know what they're
doing, anyway)
other Daemons that run nicely on my FreeBSD box:
BIND (Nameserver address:127.0.0.1 :)
GATED (throw in a couple of 10/100BT cards, and you're now a router)
>Yup. And it costs $20 :-)
damn, I paid $24.95 for mine, but I don't feel cheated :)
>Shame Oracle doesn't run under it :-/
hmmm, I think that's a feature :)
bottom line:FreeBSD, it's an ISP in a box :)
--
Name: Samuel S. Thomas
E-mail: sth...@onramp.net (Samuel S. Thomas)
http://rampages.onramp.net/~sthomas
remember: THINK!
-----BEGIN PGP PUBLIC KEY BLOCK-----
Version: 2.6.2
mQCNAzJTV3IAAAEEANFerrIuGYvYlFtnvLzfEfG2f44mYsbSA5cDYLEDc4OCNMpT
gzcwJ6I/N6OdxDy7/9iYOxycbPSv6TEspwi7LG5natd1J1j3KdvN6fpTAwEh5oKm
jX+ql1NlgqlzXNc/nNH63pt+rWdLZF1fJkkQ9VBU+EYV5+m7DW1yAZTdRoeNAAUR
tCVTYW11ZWwgUy4gVGhvbWFzIDxzdGhvbWFzQG9ucmFtcC5uZXQ+
=GAY+
-----END PGP PUBLIC KEY BLOCK-----
Kevin Berry
Ronny Ong wrote:
> >has just changed the license for Microsoft applications to forbid their
> >use on non-Microsoft operating systems (this clause sounds a tad
> >monopolistic, doesn't it?), it is illegal for someone to run Word '97 on
> >WABI or WINE, or one or the other Windows emulations than can be used
> >on one of the BSD operating systems or Linux. Something like telling
>
> seen this interpretation substantiated. What the language states is
> that the apps must run on a "licensed" copy of a Microsoft OS. The
> emphasis here is mine, but that's how I read it. I've seen otherwise
[snip]
I wouldn't think it would matter anyway. MS Office is a poor product compared
to current versions of the Corel and Lotus competitors. They also have much
fewer bugs, too.
--
mailto:rbe...@arlington.net or kevin...@chrysalis.org
http://www.chrysalis.org/kevinb
Note: All words from my account are my own and do not represent the
views of StarText, or any of its affiliates. Check for my PGP keys on
the servers as 0xFF92A22D, 0x122C2199, 0x08E821A9.
Key fingerprint for key 0xff92a22d = CDE4096152DFBB9C-6192ACE4290CFC04.
ro...@127.0.0.1
>Welcome to wcarchive.cdrom.com, a 150Mhz Pentium Pro system with 512MB of
>memory and 72GB of hard disk storage. This server is running FreeBSD 2.1.0
>(The latest version of FreeBSD is available in pub/FreeBSD).
>
if you ftp over to the system right now, it says 200MHz p6. max 1500 users.
at the moment there are 816 connected.
>512MB of RAM!! hmm, I thought 640K was enough for anybody!?!?!?
>
>>>I can believe that. It has something like 72GB of hard disk space,
>>>plus CD-ROMs. You can only get so many drives per SCSI card.
>>
>>The 2940UWs that I use will do the full 15 drives per card. 9gb*15=135.
>>take a machine with 4 pci slots and you could have 540gb attached to
>>one box :-) One of those 8 pci slot alr boxes would make a nifty
>>ftp server.
>
>4 slots=405GB...9GB*15*3=405GB
>add one PCI 100B-TX ethernet card, and you're ready to go.
>
I guess you'd need a pci nic to do 100megabits. i was thinking of
4 scsi cards. a while back i found 4 1740 eisa controllers and
stuck them in one machine so each drive would have one
controller. i used the machine largely for sim city and offline mail
reader.
>>>Bottom line, this FreeBSD machine is doing tons of work for being one
>>>machine with one CPU. Please show me a NT box with any number of
>>>processors performing equally well under similar Internet-type load
>>>that costs less than $100K. I don't know of any such configuration,
>>>and you can buy a lot of hardware for under 100K.
>
>It is also their www server. It has sendmail and POP3 Daemons running,
>too. Adding a newsfeed would be trivial.(for people who know what they're
>doing, anyway)
adding a news feed would also be a waste of bandwidth and diskspace.
>
>other Daemons that run nicely on my FreeBSD box:
>
>BIND (Nameserver address:127.0.0.1 :)
cat /etc/resolv.conf
127.0.0.1
127.0.0.1
127.0.0.1
>GATED (throw in a couple of 10/100BT cards, and you're now a router)
>
>>Yup. And it costs $20 :-)
>
>damn, I paid $24.95 for mine, but I don't feel cheated :)
>
you can ftp it if you like, but that is only for people who feel they
should inflict punishment on themselves.
>>Shame Oracle doesn't run under it :-/
>hmmm, I think that's a feature :)
>
>bottom line:FreeBSD, it's an ISP in a box :)
sort of. if you ever got very big, you'd have to move to something
a bit more scalable. i am getting to be where i really hate solaris
after all the neato stuff you get with freebsd. the only thing
keeping me from running bsd for a web server is you cant mirror
drives, do smp (at the moment) and there is some way cool
things out for solaris as far as far as databases, streaming
video, audio, secure web servers, etc. solaris is rock solid and
scalable way beyond what i need, but bsd is so cool. and comes
with many more games :-P I guess I could buy an indy, which
comes with some cool cames and the interface doesnt suck
anywhere near as badly as openwindows....those cool little
juggling balls would almost make it worthwhile :-P
ro...@127.0.0.1
In <32c3822c...@news.airmail.net>, ronn...@airmail.net (Ronny Ong) writes:
>The Mac faithful rushing to embrace NeXTStep would probably disagree.
>Of course, from a business standpoint, it's proprietary and destined
>to become more so, so its applicability to Unix at large is
>nonexistent. Therein lies the rub: unless someone is willing to
>develop and distribute it for FREE, the much-denied fragmentation of
>Unix strongly discourages wide adoption.
What would really be cool is a chrp compliant nextstep, especially
an smp version 8-) nextstep is the most beautiful operating system
ever created by man. if they brought back newsgrazer and all the
neat stuff you got with black hardware , I'd be a very happy camper.
i personally think the beos is going to be more popular on the mac.
the people that write it are incredibly clueful people.
>
>>has just changed the license for Microsoft applications to forbid their
>>use on non-Microsoft operating systems (this clause sounds a tad
>>monopolistic, doesn't it?), it is illegal for someone to run Word '97 on
>>WABI or WINE, or one or the other Windows emulations than can be used
>>on one of the BSD operating systems or Linux. Something like telling
>
>Other than in presumptuous MS flames and press editorials, I haven't
>seen this interpretation substantiated. What the language states is
>that the apps must run on a "licensed" copy of a Microsoft OS. The
>emphasis here is mine, but that's how I read it. I've seen otherwise
>law-abiding companies purchase a single Windows 95 upgrade package and
>proceed to install it in 50 machines. The same companies typically
>bought a single copy of Windows 3.1 a few years ago and installed it
>on the same 50 machines. If they get raided by SPA, the fine is based
>upon the Microsoft's estimated retail price of 49 copies of Windows
>95. Less than $10K, big deal. Now change it so that the penalty is
>based on 49 copies of Windows 95 as well as 50 copies of the Microsoft
>Office 97 they were running on Windows 95. Even if they paid for those
>copies of Office 97, they're illegal because they're running on an
>unlicensed OS. Now the liability is several times more.
>
>I don't know about WINE, but WABI apparently still requires Microsoft
>Windows binaries, as does OS/2 if you include Win-OS/2 support, so if
>you've got a legal copy of WABI or OS/2, you automatically have a
>licensed copy of a Microsoft OS, too.
>
>
i looked at wabi that shipped with solaris. it wanted me to stick windows
disks in the machine. rather than further desecrate (sp?) the machine
i killed the install app.
Harrison Bergeron
Samuel S. Thomas bo...@onramp.net stated:
>
>
R3_6Alpha_2 in ELF binary format along with gdc, ripquery
and ospf_monitor (should anyone have trouble finding an
elf compiled copy let me know. It's compiled with libc 5.3.18
and is NOT compiled as a static executable, so it may not
work with earlier versions of libc)
>remember: THINK!
Optimist.
--
dav...@davids.psyberlink.net |
Steinberger: | Bah Humbug
State of the Instrument |
Phil Howard
On 27 Dec 1996 05:47:14 GMT ro...@127.0.0.1 wrote:
| I guess you'd need a pci nic to do 100megabits. i was thinking of
| 4 scsi cards. a while back i found 4 1740 eisa controllers and
| stuck them in one machine so each drive would have one
| controller. i used the machine largely for sim city and offline mail
| reader.
If you stuck in 3 Adaptec AHA-3985's and ran ribbon cables all around,
you could run as many 3(cards)*3(channels)*15(drives)*9gig = 1.215 TERAbytes
I've seen some motherboards out there with 8 PCI slots and dual P-PRO.
| >It is also their www server. It has sendmail and POP3 Daemons running,
| >too. Adding a newsfeed would be trivial.(for people who know what they're
| >doing, anyway)
|
| adding a news feed would also be a waste of bandwidth and diskspace.
The news feed would also kill even 512meg of RAM unless they have swap
turned OFF. News servers DO NEED TO BE SEPARATED!
| sort of. if you ever got very big, you'd have to move to something
| a bit more scalable. i am getting to be where i really hate solaris
| after all the neato stuff you get with freebsd. the only thing
| keeping me from running bsd for a web server is you cant mirror
You're better off with hardware mirroring, anyway, and that can be done
with Mylex or DPT SCSI controllers.
| drives, do smp (at the moment) and there is some way cool
| things out for solaris as far as far as databases, streaming
| video, audio, secure web servers, etc. solaris is rock solid and
Not really. It has the same I/O to swap buffer instability problem
that many other systems have (but IRIX does not appear to have it).
Linux and FreeBSD also have it. But one quick solution is to pile
in the RAM (512meg!) and turn off the swap space. Solaris 2.6 may
have a workaround option for this problem.
--
Phil Howard KA9WGN +-------------------------------------------------------+
Linux Consultant | Linux installation, configuration, administration, |
Milepost Services | monitoring, maintenance, and diagnostic services. |
phil at milepost.com +-------------------------------------------------------+
m...@ohmy.com
In <5a6j8i$3...@sunsrv12.clr.com>, ph...@milepost.com (Phil Howard) writes:
>On 27 Dec 1996 05:47:14 GMT ro...@127.0.0.1 wrote:
>
>| I guess you'd need a pci nic to do 100megabits. i was thinking of
>| 4 scsi cards. a while back i found 4 1740 eisa controllers and
>| stuck them in one machine so each drive would have one
>| controller. i used the machine largely for sim city and offline mail
>| reader.
>
>If you stuck in 3 Adaptec AHA-3985's and ran ribbon cables all around,
>you could run as many 3(cards)*3(channels)*15(drives)*9gig = 1.215 TERAbytes
is there freebsd support for it? :-)
>
>I've seen some motherboards out there with 8 PCI slots and dual P-PRO.
>
you can get 4 ppros on a board if you want to buy the entire machine
from intel.
>
>| sort of. if you ever got very big, you'd have to move to something
>| a bit more scalable. i am getting to be where i really hate solaris
>| after all the neato stuff you get with freebsd. the only thing
>| keeping me from running bsd for a web server is you cant mirror
>
>You're better off with hardware mirroring, anyway, and that can be done
>with Mylex or DPT SCSI controllers.
>
I suppose. I haven't had the best of luck with dpt in the past.
>| drives, do smp (at the moment) and there is some way cool
>| things out for solaris as far as far as databases, streaming
>| video, audio, secure web servers, etc. solaris is rock solid and
>
>Not really. It has the same I/O to swap buffer instability problem
>that many other systems have (but IRIX does not appear to have it).
>Linux and FreeBSD also have it. But one quick solution is to pile
>in the RAM (512meg!) and turn off the swap space. Solaris 2.6 may
>have a workaround option for this problem.
I was farily close to buying an SGI until I found out the drives had
to be bought from sgi due to some special chip or something they
add to a normal scsi drive to make you buy drives from them. that
and the io numbers i've seen for the buck haven't impressed me
much.
Now that you bring this up, I have an excuse to buy 512mb of
ram :-)
Mark A Harrison
Frank Durda IV (uhc...@nemesis.lonestar.org) wrote:
: Mark A Harrison (mhar...@utdallas.edu) wrote:
: [55]Actually, the progression went bare bones DOS, Unix, AT&T won't
: [55]give good licensing terms, scratch Unix, DOS that could actually
: [55]do something (which version would that be?).
: [55]
: [55]Frank could fill you in on the historical details, he was there...
: Um, well, PC-DOS appears in 1981, well after UNIX can claim to have
: existed.
Yes, of course. But Microsoft's involvement with Unix follows
the timeline I mentioned, doesn't it?
Mark.