Various online Speedtests are averaging 15M down, 800k up. Some
speedtests report upload as higher, but note there is compression
going on.
In the past when I worked from home, upload speeds were a crucial
point. Now though, our home internet service is mostly web browsing,
email, etc.
I really appreciate cost comparisons folks can offer. From what I
read, having cable TV plus Internet is cheaper than cable TV plus DSL
and a landline. I would be happy to be corrected.
One thing that surprised me is that VISI no longer offers an
"off-network" option for Usenet. You have to be a DSL subscriber.
Since we haven't nuked our DSL yet, I can type this now. I asked the
VISI tech whether Octanews was still around - and now they apparently
have only payed plans - no more low-bandwidth free service.
Unlike folks who "Read Playboy for the articles", I really only use
Usenet for text newsgroups. Even $5 a month seems kinda harsh for
occasional tapping of 1k messages. You'd think they'd at least offer a
discount if you killfile Lutsen.
As I mentioned, one of the motivations is that our landline is useless
- we use our cells and only ever get calls from telemarketers. A
blogger I read mentioned that it'd be awesome if the government
actually enforced the "do-not-call" list. I think it'd be awesome if
the government enforced monopoly laws. "I don't like Comcast/QWest..
I'll just switch to the other cable/phone company.. oh, wait.."
Hell, with our mandatory but unlimited data plan, maybe I should just
tether a Sprint cellphone to the home network.
BLink
--------------------------
"The worst thing about censorship is [redacted]"
> Usenet for text newsgroups. Even $5 a month seems kinda harsh for
> occasional tapping of 1k messages. You'd think they'd at least offer a
> discount if you killfile Lutsen.
:o)
> Unlike folks who "Read Playboy for the articles", I really only use
> Usenet for text newsgroups. Even $5 a month seems kinda harsh for
> occasional tapping of 1k messages. You'd think they'd at least offer a
> discount if you killfile Lutsen.
You can get text only Usenet service from individual.net for 10 Euros a
year. I've been using them for several years and am quite happy with them.
What do you (or anyone) need 15M for? Seems to be the way of things these
days - ISP's offer ever-faster; I'd much rather they either offered
cheaper or offered better service for the same price. Or did you just go
for that much in order to get a reasonable upload speed?
> One thing that surprised me is that VISI no longer offers an
> "off-network" option for Usenet. You have to be a DSL subscriber. Since
> we haven't nuked our DSL yet, I can type this now. I asked the VISI tech
> whether Octanews was still around - and now they apparently have only
> payed plans - no more low-bandwidth free service.
As someone said, news.invidiual.net are very cheap and they were always
excellent in terms of service. Their payment gateway people (and this was
the UK one, the US operation may be better) were a bunch of muppets,
though, so I ended up leaving them earlier this year.
I've been using Eternal September since (previously Motzarella) - the
service is pretty darn good considering the price (i.e. nothing). Once in
a while it's a little sluggish to first fetch new articles, and once in a
while it won't let me post something and I have to try again a few
seconds later, but overall I'm impressed.
> BLink
> --------------------------
news.individual.net is about $10/year for text only. Very reliable. I
used it for years before getting DSL and iphouse's free news. FWIW.
--
-Barb, Mother Superior, HOSSSPoJ
http://web.me.com/barbschaller - Who Said Chickens Have Fingers?
10-30-2009
> I've been using Eternal September since (previously Motzarella) - the
> service is pretty darn good considering the price (i.e. nothing). Once in
> a while it's a little sluggish to first fetch new articles, and once in a
> while it won't let me post something and I have to try again a few
> seconds later, but overall I'm impressed.
same here. very reliable
Regards,
Mike O'Brien
life is full of choices. choices have consequences
Telecommuting. If you are going to work with large CAD files, cutting
upload and download times in half makes a difference of hours instead of
mere minutes per telecommuting week.
Though I hear that there may soon be a more efficient method of distributing
large files, supposedly involving "fractals" in some form. Fractals can do
everything! ;-)
I also have been use them and been quite happy.
Good excuse to go and get a coffee or have lunch or whatever though :-) I
certainly wouldn't want to be messing around with large files over an
ancient 2400-baud modem or anything, but I have a hard time finding things
where something as high as 15M is *vital* over a fraction of that speed -
and if it is then maybe telecommuting isn't the answer...
Hmm, fond memories of the good ol' days when project deadlines loomed
though, driving a few hundred miles with a few data CDs or tapes in the
back of the car to a customer site because it was faster and more reliable
than having the computer network do it :-)
> Though I hear that there may soon be a more efficient method of
> distributing large files, supposedly involving "fractals" in some form.
> Fractals can do everything! ;-)
Do you mean compression of large data files so they can be moved around
more quickly, or distribution/access to portions of those files (so that
less data gets moved to access a particular part)?
(I know someone who worked on fractal compression to reduce data sizes in
the early 90s - it proved to be a bit of a dead-end at the time though,
because the compression levels attainable just weren't enough to justify
the cost or the time taken. The custom hardware he designed bristled with
memory and processors though, and was quite the work of art :-)
cheers
Jules
That won't work. But you COULD cancel your $5/month account and sign
up with Google Groups for free.
> Hmm, fond memories of the good ol' days when project deadlines loomed
> though, driving a few hundred miles with a few data CDs or tapes in the
> back of the car to a customer site because it was faster and more reliable
> than having the computer network do it :-)
My preferred method of getting lots of data from California used to be
FedEx. I would get a 10.5" reel of tape with fresh data in my hot little
hands early every morning. FedEx never failed me.
It's percentages. The number of client projects you can handle will be
seriously hampered when your backbone speed is cut in half. Architecture
involves a lot of data transfer, and it's not a "coffeebreak/lunchbreak"
deal unless you're only working with a very small number of projects.
But lets look at another application, digital video footage at 29.97 frames
per second and 1080p resolution. You can count on any non-trivial media
project being distributed over multiple workstations and/or locations, and
that means that your bottleneck will be the speed of the data transfer
backbone that you are working on. You can do the math and figure out the
number of frames for a minute of footage, or ten minutes or a half hour.
This isn't "take a coffee break" stuff, it's an all day process with tight
deadlines.
>
>> Though I hear that there may soon be a more efficient method of
>> distributing large files, supposedly involving "fractals" in some form.
>> Fractals can do everything! ;-)
>
> Do you mean compression of large data files so they can be moved around
> more quickly, or distribution/access to portions of those files (so that
> less data gets moved to access a particular part)?
>
> (I know someone who worked on fractal compression to reduce data sizes in
> the early 90s - it proved to be a bit of a dead-end at the time though,
> because the compression levels attainable just weren't enough to justify
> the cost or the time taken.)
Blasphemy! If you've been following the newsgroup for the past month or so,
you already know that fractals are the most efficient and cost-effective way
to implement EVERYTHING, whether it be better mass transit, or the
organization of loosely-associated volunteer efforts!! ;o)
Just kidding of course. The applications where fractals help me out the
most are animation timesavers, such as generating random-appearing
computer-generated landscapes, or improving the billowy appearance of cloud
animations -- fractal noise does a nice job enhancing such textures.
>
> "Jules" <jules.rich...@remove.this.gmail.com> wrote in message
> news:pan.2009.11.19....@remove.this.gmail.com...
>> On Thu, 19 Nov 2009 08:13:34 -0600, catpandaddy wrote:
>>> "Jules" <jules.rich...@remove.this.gmail.com> wrote in message
>>> news:pan.2009.11.19....@remove.this.gmail.com...
>>>>
>>>> What do you (or anyone) need 15M for?
>>>
>>> Telecommuting. If you are going to work with large CAD files, cutting
>>> upload and download times in half makes a difference of hours instead of
>>> mere minutes per telecommuting week.
>>
>> Good excuse to go and get a coffee or have lunch or whatever though :-) I
>> certainly wouldn't want to be messing around with large files over an
>> ancient 2400-baud modem or anything, but I have a hard time finding things
>> where something as high as 15M is *vital* over a fraction of that speed -
>> and if it is then maybe telecommuting isn't the answer...
>
> It's percentages. The number of client projects you can handle will be
> seriously hampered when your backbone speed is cut in half. Architecture
> involves a lot of data transfer, and it's not a "coffeebreak/lunchbreak"
> deal unless you're only working with a very small number of projects.
I suppose I've always worked with systems that are quite distributed - so
the bulk of the data's essentially central and I just download the
fragments that I need across the network, rather than a clone of some
enormous data-set (the few times I have really needed to do the latter
I've managed to either time it with a visit somewhere where I can copy
the data 'locally' to disk, or I've just set it off downloading
overnight so it's ready for when I need it the next day)
> But lets look at another application, digital video footage at 29.97
> frames per second and 1080p resolution. You can count on any
> non-trivial media project being distributed over multiple workstations
> and/or locations, and that means that your bottleneck will be the speed
> of the data transfer backbone that you are working on.
Yes, but then you would have a nice fat backbone between sites - but I'm
not sure it's a situation where you'd want people working at home on the
data anyway, partly for security reasons (TV and media companies at least
are *very* paranoid about unauthorized access to data, IME) and partly
because media-based projects tend to be highly collaborative and so don't
benefit so much from lots of folk working in isolation.
> You can do the
> math and figure out the number of frames for a minute of footage, or ten
> minutes or a half hour. This isn't "take a coffee break" stuff, it's an
> all day process with tight deadlines.
:-)
Last time I screwed with that kind of stuff was a project for delivering
movies to movie theaters via network - this was well over a decade ago
now, so it was really pushing the technology of the era. Some of the
numbers did get scary pretty quick (we had hard disks and DAT tapes full
of data flying back and forth all over the place). The network links were
all ATM at what, 155Mbps, and they could barely cope. Lots of fun for a
techie to mess around with, though! :-)
> Blasphemy! If you've been following the newsgroup for the past month or
> so, you already know that fractals are the most efficient and
> cost-effective way to implement EVERYTHING, whether it be better mass
> transit, or the organization of loosely-associated volunteer efforts!!
> ;o)
Ha ha! I've always admired it as an idea - recommended it myself on
several occasions - but there is a danger of thinking that everything is a
scaled-up or scaled-down version of something else, and losing sight of
"local issues".
> Just kidding of course. The applications where fractals help me out the
> most are animation timesavers, such as generating random-appearing
> computer-generated landscapes, or improving the billowy appearance of
> cloud animations -- fractal noise does a nice job enhancing such
> textures.
Heh, remember the "fractal flight" animation on the humble ol' Amiga in
the late 80s or so - along with some fractal terrain-generating program (I
have a horrible feeling it may have been called Vista)? I think that was
the first time I had anything "fractally" running on a machine I had at
home, but it was pretty impressive for the day and showed where things
were heading...
cheers
Jules
>>>> "Jules" <jules.rich...@remove.this.gmail.com> wrote in message
>>>> news:pan.2009.11.19....@remove.this.gmail.com...
>>>>>
>>>>> What do you (or anyone) need 15M for?
You answered your own question below when you said:
> numbers did get scary pretty quick (we had hard disks and DAT tapes full
> of data flying back and forth all over the place). The network links were
> all ATM at what, 155Mbps, and they could barely cope.
Barely coping at 155Mbps... kinda makes a mere 15Mbps seem inadequate,
doesn't it!
I know it's probably apples and oranges, but the day when it is possible to
do a 100Mbps render farm entirely over home-grade internet can't get here
soon enough for me. That would be a thousand shades of awesome. Well,
technically it would be 1024 shades, I guess.
It makes it seem entirely inadequate, that's the thing. 30M or 60M
probably isn't fast enough either - at least not if any kind of quality is
expected at the far end. Of course I'm sure the day will come where nobody
really remembers what good-quality broadcast over the wire could look
like, at which point crappy quality seems just awesome because "hey,
it's new technology!"...
> I know it's probably apples and oranges, but the day when it is possible
> to do a 100Mbps render farm entirely over home-grade internet can't get
> here soon enough for me. That would be a thousand shades of awesome.
> Well, technically it would be 1024 shades, I guess.
I think it's possible to do it now. All you need to do is centralize the
model that you're working on; the machines making up the farm just see a
small window of the rendered scene and only need to retrieve the parts
(meshes, textures, objects etc.) that will appear in that final window. I
don't think the actual amount of data transferred to each machine is
*that* big, even though the entire model might be enormous.
From the user's point of view the local software just needs to know where
the model is and what machines are part of the farm. The 'farm' machines
just need to transfer the data to do their bit when asked, and the spit
their rendered chunk back to the 'controller' for display.
I suspect a lot of the effects houses could do it over the public 'net
with the software they have right now - they just don't because of
security considerations and a belief that it's better to have all their
designers under one roof rather than spread all over the place.
(I've set up networked raytracing before, just not over the 'net, and it's
kind of fun hooking up a bank of machines and watching each of them do
their bit and contribute toward the final scene :-)
cheers
Jules
That's more cynical than I'm willing to go, even with my skeptical side
getting a vote. Even analog transmissions took a long time to really start
looking good, and I remember the first time I saw Mpeg-1 compression back in
the early 1990s.
Today's implementations of HDTV are best thought of as equivalent to where
analog television was back sometime between the 1940s and the 1970s. I
remember the first time I saw how blocky a home-recorded VHS videotape
looked when played on an HD television screen -- the screen resolution was
so much higher than the tape resolution that the flaws had nowhere to hide.
What really puts things in perspective is that this time around, the
improvments in quality that took the better part of a century with standard
resolution television, we are achieving in much less time, and we're doing
it with exponentially higher requirements.
It sounds like we should collaborate sometime.
If you know which tech told you that, please let me know. We still
offer the service, authenticated, at offnetwork-news.visi.com. If your
auth doesn't work, please contact Support to open a ticket. I'll make
sure they know it's still active. The old tool to enable offnetwork
access is no longer available, but we can set it up for you or update
your auth.
--
Jeff Kozel
Manager, Customer Services
VISI
Support: 612.395.9010
Main: 612.395.9000
Email: sup...@visi.com
Web: www.visi.com
I was curious about this as I am accessing Usenet off network from VISI
at this moment. Seems to work fine.
>Brian Link wrote:
<snip>
>> One thing that surprised me is that VISI no longer offers an
>> "off-network" option for Usenet. You have to be a DSL subscriber.
>> Since we haven't nuked our DSL yet, I can type this now. I asked the
>> VISI tech whether Octanews was still around - and now they apparently
>> have only payed plans - no more low-bandwidth free service.
>
>If you know which tech told you that, please let me know. We still
>offer the service, authenticated, at offnetwork-news.visi.com. If your
>auth doesn't work, please contact Support to open a ticket. I'll make
>sure they know it's still active. The old tool to enable offnetwork
>access is no longer available, but we can set it up for you or update
>your auth.
I don't recall the fellow's name, but I called around 9pm last night,
I think.
Thanks for pointing this out.. I just took it as another sign that
VISI's morphed into a web-hosting and business-access outfit rather
than an ISP for any old Joe.
I remember a job interview I had last year, where someone looked at my
email address and said "oh, you work for VISI?". Um, no. "Oh, well
your business must use them". Um, no, I'm just a guy who likes
internet access and I get it through VISI.
"wow, they do that?"
does it have articles?
Might be interesting!
btw as resident fractal expert, you don't happen to know of any good
[modern] fractal terrain/landscape generators, do you?
>
> "Jules" <jules.rich...@remove.this.gmail.com> wrote in message
> news:pan.2009.11.20....@remove.this.gmail.com...
>> On Thu, 19 Nov 2009 23:43:25 -0600, catpandaddy wrote:
>>>>>>>> What do you (or anyone) need 15M for?
>>>
>>> You answered your own question below when you said:
>>>
>>>> numbers did get scary pretty quick (we had hard disks and DAT tapes full
>>>> of data flying back and forth all over the place). The network links
>>>> were
>>>> all ATM at what, 155Mbps, and they could barely cope.
>>>
>>> Barely coping at 155Mbps... kinda makes a mere 15Mbps seem inadequate,
>>> doesn't it!
>>
>> It makes it seem entirely inadequate, that's the thing. 30M or 60M
>> probably isn't fast enough either - at least not if any kind of quality is
>> expected at the far end. Of course I'm sure the day will come where nobody
>> really remembers what good-quality broadcast over the wire could look
>> like, at which point crappy quality seems just awesome because "hey,
>> it's new technology!"...
>
> That's more cynical than I'm willing to go, even with my skeptical side
> getting a vote.
:-)
> Even analog transmissions took a long time to really start
> looking good, and I remember the first time I saw Mpeg-1 compression back in
> the early 1990s.
Yeah, maybe they did here and the setup/evolution was a lot different.
Remember I grew up with three national analog TV channels (later four,
then five - but I couldn't get that last one because the frequencies were
blocked by a local astronomy research place). That was PAL TV though
rather than NTSC, so the color encoding was (IMHO) better and it gave an
extra 20% or so screen resolution.
I can't really do a fair comparison with US analog though as I don't think
I've seen an NTSC setup at its best yet (and probably never will now, I
suppose!)
> Today's implementations of HDTV are best thought of as equivalent to
> where analog television was back sometime between the 1940s and the
> 1970s. I remember the first time I saw how blocky a home-recorded VHS
> videotape looked when played on an HD television screen -- the screen
> resolution was so much higher than the tape resolution that the flaws
> had nowhere to hide.
Hmm, I'm not sure where it'll go, but TV companies seem very good at
pushing quantity over quality (particularly if it gets them more ad
revenue) - and to be honest, the majority (perhaps now used to things like
cell-phone cameras rather than real cameras) seem to silently accept it.
> What really puts things in perspective is that this time around, the
> improvments in quality that took the better part of a century with
> standard resolution television, we are achieving in much less time, and
> we're doing it with exponentially higher requirements.
I suppose the interesting thing is whether those potential improvements
actually get passed onto the consumer - or if faster links and more
bandwidth just translate to extra poor-quality channels.
On the display side, lots of resolution is great - so long as the data's
there to best make use of it, and so long as the picture looks as good as
it does with the technology that's being made obsolete (brightness,
contrast, pixel aliasing, switching frequency etc. etc.). I'm not sure
we're quite "there" yet.
cheers
Jules
In this day and age: All of them!
Erm, that is to say, all of the modelers around today have excellent terrain
creation tools, and they all rock.
Hmm, maybe I'll take another prod at Blender then!
Although this site has some good info on the processes:
http://local.wasp.uwa.edu.au/~pbourke/fractals/noise
... rustling up some code doesn't look too difficult; I only need terrain
as though viewed from directly above, projected onto a rectangle. The only
downside is that it won't automatically handle rivers, but I'm sure I
could do a bit of post-processing (the end result just needs to be a
static image).
cheers!
Jules
Actually, it's not silent at all. Most people, when given the choice
of more channels vs higher quality, choose more channels.
(Keep in mind that anyone on this newsgroup doesn't count as "most
people.")
Not that the networks aren't happy to oblige, mind you. But people
really do prefer quantity.
Heck, I even agree with most people on this one.
Craig
Just keep in mind if you do, that release 2.5 won't be out until a few
months into 2010, and it's a significant update. A lot of new features have
been added incrementally over the years, and it's caused too much sprawl in
the various control panels. All of the updates from v2.31 through v2.49 are
being reintegrated from the ground up. Once the new major upgrade is out of
beta testing, you'll probably be able to jump right in with just the online
documentation, and there should be a comprehensive users manual in print
version soon after.
Which choice offers the highest titty quotient?
I'm sure there's a farming channel for you.