Question Re: Size of Desktop Computers - Page 2

Last message on previous page:
Archived from groups: comp.sys.laptops (More info?)

plated metal wrote:

> J. Clarke wrote:
>> plated metal wrote:
>>
>>
>>>J. Clarke wrote:
>>>
>>>>kiss_my_ass wrote:
>>>>
>>>>
>>>>
>>>>>its like cellphones taking over landlines, in a couple of years or less
>>>>>laptops will become the norm of the day.
>>>>
>>>>
>>>>In what universe have cellphones taken over landlines? I'm sorry, but
>>>>one cannot get broadband on a cellphone.
>>>>
>>>
>>>broadband will be wireless in a few years (trust me, i'm working on it
>>>right now!)
>>
>>
>> And of course wired broadband will be standing still the whole time.
>
> you have a point, but it's fundamentally limited by the impedance of the
> wires - the capacitance severely limits the max frequency signal you can
> chug down them, and the more bits per hertz you try to stuff, the
> tighter the signal to noise ratio. and so much of the copper is old old
> old, and the connections to it decrepit. Plus, its far cheaper to put in
> wireless infrastructure in china and india and the like. once a 20 buck
> mmm-wave radio is possible, well, it's all over...

Right now none of the cell-phone providers can provide me with a usable
signal in my house which is midway between New York and Boston. And I'm to
believe that in ten years they're going to be giving me broadband over that
nonexistent signal? Don't make me laugh.

In any case, there is also coaxial cable coming into nearly ever house in
the country, and that is capable of working at the same frequencies as
cellular.

As for a 20 buck "mmm-wave radio", do you mean one that operates at a 1
millimeter wave length or is an "mmm wave" something like a "tetrion
particle"? If you mean "mm" that would be what, 3 teraHertz? So who is
using that frequency for RF transmission at any price anywhere? What makes
you think that such devices will be available for 20 bucks any time in the
next billion years?

>>>and that'll be the end of telcos as they exist right now -
>>>they won't have a (near) monopoly on the dominant way of getting
>>>reliable data into your house or business.
>>
>>
>> Or so you told the venture capitalist.
>
> shhhhhhhhhhhhhh! for crying out loud, whose side are you on?

The side of reason, not the side of rah-rah enthusiasm for every harebrained
scheme that some college kid or Bill Gates wannabee comes up with.

>>>there'll be a hell of a
>>>shake-out, the usual lobbying of congress/parliaments to preserve
>>>monopolies by ridiculous anti-capitalist stealth (you ain't seen nothing
>>>with the way the music industry is resisting technology change).
>>>
>>>i reckon in 10-15 years the copper wires between your house and the
>>>local telephone exchange will be unused.
>>
>>
>> Or else it will be carrying 10-100 Mb/sec using the same technology
>> you're using to get T1 speed out of broadband.
>
> nup. dude, the wireless will be doing gigabits/sec.

So let's see, you're going to have cellular service doing billions of bits
per second. That means that you are going to have a bandwidth of at least
4 GHz. To carry several hundred of these at one time you're up in the
hundreds of Ghz--sounds like you really do intend to run in the teraHertz
range. Now, what kind of power are you going to have to put in a cell
phone to get it to reach the cell at those bands going through things like
cars and trees and rain and snow and houses and the like? Hmmm?

I see that there are a few initiatives looking at the possibility of this
for short range communication and one of the benefits touted _is_ that it
is by its nature a short-range technology. That means that either there's
a cell on every tree and in every house or it's not going to be used for
cell phones. It's also being looked at for spectroscopy and the
spectroscopists seem to think that those frequencies are strongly absorbed
by air.

Seems to me that you've read something you didn't really understand about
some initiative or other that's trying to make a laboratory curiosity that
might someday with much work be made into a short-range communication
system similar to 802.11g or the like but at higher frequency and you've in
your head assumed that all wireless is created equal and turned that into a
next-generation cellular system.

Sorry, but this is only slightly more credible than cold fusion.

> -p

--
--John
to email, dial "usenet" and validate
(was jclarke at eye bee em dot net)

Archived from groups: comp.sys.laptops (More info?)

>broadband will be wireless in a few years (trust me, i'm working on it
>right now!) and that'll be the end of telcos as they exist right now -

So you are saying that the wireless phones of the
future will just be small "computers" catching a
broadband signal off nearest tower?

And that ALL communications will occur with such a
device whether it be voice, email, data, etc?

Archived from groups: comp.sys.laptops (More info?)

In "Re: Question Re: Size of Desktop Computers", "J. Clarke"
<jclarke.usenet@snet.net.invalid> wrote:


[...]
>Would I go back to that era if I could? I don't think so. Am I glad to
>have been there? You bet.

I also recall the PDP-8 I started on with a great deal of nostalgic affection,
but it wasn't exactly small-form-factor in anything except OS size...!

Archived from groups: comp.sys.laptops (More info?)

pokemonn2@hotmail.com wrote:

> jakdedert: "Actually, the laptop designer has to go to fairly
> extraordinary lengths to
> fit that much computing power into such a small package. Things have
> to be
> designed to much tighter tolerances, with much more complex
> engineering,
> employing more expensive materials. "
>
> You couldn't be more wrong jakdedert. The Intel Sonoma chipset along
> with the Pentium M can be incorporated into all new desktop computers
> with ease.

At what price? You think Intel is going to sell these at bargain prices?

> This chipset uses much less energy and actually runs faster
> than the Pentium 4 due to fewer "stops" and greater efficiency in
> processing power.

According to who? Where are the benchmarks that show this?

> Also, DDR2 memory uses less voltage (and hence less
> energy)

Not a valid assumption. Lower voltage != less energy.

> than DDR1 not to mention that DDR2 runs faster.

Where are the benchmarks that show this?

> These are 2
> simple examples that I just gave that computer manufacturers can
> instantaneously incorporate into their systems to make them run more
> efficiently.

Not "instantly". They'd need to design the system around those components.
And how much "more efficiently" would they run?

> If they actually cared, and did research and development,
> I am extremely confident that they could find many many more ways to
> make their systems run more efficiently.

Why would they want to?

> In the end, it all boils down to cost vs. benefit.
>
> That's right jakdedert. It does, and what I'm proposing is going to
> lower costs and everyone will benefit.
> jakdedert: Yet another person who completely missed the point of my
> original post. And I do mean completely.

Look, smart guy, if you're so certain that the market is clamoring for the
kind of machine you describe, then get some venture capital and start
providing them and make yourself rich.

I'm not holding my breath to see you deliver.

--
--John
to email, dial "usenet" and validate
(was jclarke at eye bee em dot net)

Archived from groups: comp.sys.laptops (More info?)

Paul Rubin wrote:

> "J. Clarke" <jclarke.usenet@snet.net.invalid> writes:
>> > This chipset uses much less energy and actually runs faster
>> > than the Pentium 4 due to fewer "stops" and greater efficiency in
>> > processing power.
>>
>> According to who? Where are the benchmarks that show this?
>
> I assume that means at the same clock speed. I saw a benchmark
> yesterday but don't feel like digging it up. It's not a controversial
> claim. A PM-1.6 will benchmark quite a bit better than a P4-1.6. But
> P4's are available at much higher frequencies (there's no 3 ghz PM),
> so the fastest Pentium boxes in absolute terms are still P4's.

Why don't you let him tell us what he means?

He's been arguing for abandoning the P4 and making all machines like laptops
for many, many posts now.

As for "at the same clock speed", a PIII beats a P4 at the same clock speed,
so I suppose we should abandon the P4 and go back to the PIII?

--
--John
to email, dial "usenet" and validate
(was jclarke at eye bee em dot net)

Archived from groups: comp.sys.laptops (More info?)

ptb@lab.it.uc3m.es (Peter T. Breuer) writes:
> > As for "at the same clock speed", a PIII beats a P4 at the same
> > clock speed, so I suppose we should abandon the P4 and go back to
> > the PIII?
>
> One would hope so. The P4 was an attempt to win the hype-war over
> clock speeds. It turned out that lusers just bought the chip with the
> faster clock, no matter that it was slower! So intel had the bright
> idea of dropping everything like on-the-fly optimization, branch
> prediction, speculative execution, everything that increased the gate
> depth in the chip and made it fast, and simply crank up the clock rate
> on a skeleton architecture and sell it!

Yes, getting rid of those "brainiac" features let them crank up the
clock rate to speeds completely unreachable by the P3. You can buy
3.8 ghz P4's these days. Maybe you could get the same throughput from
a P3 at a "mere" 3 ghz, but they never made a 3 ghz P3 and they couldn't
if they wanted to.

I do agree with the idea of a Pentium M desktop machine since even for
desktop machines, power consumption, fan noise, etc. are issues. I've
heard Intel is putting AMD's x86-64 extensions into the Pentium 4. I
hope they'll do the same thing for the Pentium M. I'd love to buy a
64 bit Thinkpad.

Archived from groups: comp.sys.laptops (More info?)

J. Clarke <jclarke.usenet@snet.net.invalid> wrote:
>
> Why don't you let him tell us what he means?
>
> He's been arguing for abandoning the P4 and making all machines like laptops
> for many, many posts now.
>
> As for "at the same clock speed", a PIII beats a P4 at the same clock speed,
> so I suppose we should abandon the P4 and go back to the PIII?
>

The P-M can indeed whup up on the P4 in some benchmarks, even throwing clock
speed considerations out the window:

http://www.anandtech.com/cpuchipse [...] =2342&p=11

-In Business Winstone, a 1.7GHz P-M beats all P4s, including P4s at over twice
its clock speed.
-In WinZip, a 2.0GHz P-M beats all P4s but the 3.8GHz.
-In Adobe Premier, a 2.0GHz P-M beats all P4s.
-In Doom 3, the 2.0GHz P-M beats all P4s but the 3.8Ghz and the gaming-oriented
Extreme Edition chips.
-In compiling the Quake 3 source code, the 2.0Ghz P-M beat all P4s except the
3.46GHz Extreme Edition.
-In Maya 3D Rendering, the P-M 1.7Ghz beat all P4s.

I know that the P-M performs extremely poorly in other benchmarks. Its FP units
are underpowered. But otherwise it is a pretty powerful chip, especially
considering that it has no on-die memory controller like the Athlon 64.

And BTW, the P-M is largely based on the P3.

Archived from groups: comp.sys.laptops (More info?)

wbw <wbwither@y-a-h-double-o.com> writes:
> I know that the P-M performs extremely poorly in other benchmarks.
> Its FP units are underpowered. But otherwise it is a pretty
> powerful chip, especially considering that it has no on-die memory
> controller like the Athlon 64. And BTW, the P-M is largely based on
> the P3.

It seems to me that the PM is worse at floating point than the P3. I
better check, but I believe my 1.2 ghz PM runs the Ogg Vorbis encoder
slower than a 750 mhz P3.

Archived from groups: comp.sys.laptops (More info?)

me@privacy.net writes:
> >I do agree with the idea of a Pentium M desktop machine since even for
> >desktop machines, power consumption, fan noise, etc. are issues.
>
> I like the idea also
>
> Does anyone make such a desktop already?

I think there are some PM desktops. There's also Via-based desktops
that are small and fanless but quite a bit slower. See www.mini-itx.com
for some.

Archived from groups: comp.sys.laptops (More info?)

Paul Rubin <http://phr.cx@nospam.invalid> wrote:
> ptb@lab.it.uc3m.es (Peter T. Breuer) writes:
> > My desktop runs a P3 1GHz down-clocked to 750MHz, with no fan, just a
> > copper heat sink. Nice and silent.
>
> Interesting. I don't think I want to buy any more 32-bit computers
> but I wonder if I can downclock an Athlon 64 with one of those Zalman
> heat sinks. There's also the issue of the power supply fan. I have a
> Quietpc noise enclosure for a 3.5" hard drive but have not tried it
> yet. I'll probably stick with FDB laptop hard drives which are
> already very quiet.
>
> Alternatively, maybe I can put the box in the closet. Do you happen
> to know the maximum permissible length of a DVI cable?

Mine is in the cupboard under the desk. There's a pair of openings cut
in the back of the enclosure (one up, one down) to let air circulate in
and out. My DVI is in the drawer above the cupboard part. The cable is
a normal IDE cable. At 33MHz and flat they can be about a metre, I'm
pretty sure, but I didn't need or use anything special. For higher
speed bus, you may have more restrictions. Terminating on the master
device should avoid reflections.

There seem to be special X cases around now that are small and square
and neat. I just used a tower .. microtower? Minitower? Anyway, a
small case. I had to look for one with the smallest depth, front to
back, and remove trimmings. Then one has to do some wonders to stop the
cables sticking out the back to far.

Peter

Archived from groups: comp.sys.laptops (More info?)

Paul Rubin wrote:

> ptb@lab.it.uc3m.es (Peter T. Breuer) writes:
>> > As for "at the same clock speed", a PIII beats a P4 at the same
>> > clock speed, so I suppose we should abandon the P4 and go back to
>> > the PIII?
>>
>> One would hope so. The P4 was an attempt to win the hype-war over
>> clock speeds. It turned out that lusers just bought the chip with the
>> faster clock, no matter that it was slower! So intel had the bright
>> idea of dropping everything like on-the-fly optimization, branch
>> prediction, speculative execution, everything that increased the gate
>> depth in the chip and made it fast, and simply crank up the clock rate
>> on a skeleton architecture and sell it!
>
> Yes, getting rid of those "brainiac" features let them crank up the
> clock rate to speeds completely unreachable by the P3. You can buy
> 3.8 ghz P4's these days. Maybe you could get the same throughput from
> a P3 at a "mere" 3 ghz, but they never made a 3 ghz P3 and they couldn't
> if they wanted to.
>
> I do agree with the idea of a Pentium M desktop machine since even for
> desktop machines, power consumption, fan noise, etc. are issues. I've
> heard Intel is putting AMD's x86-64 extensions into the Pentium 4. I
> hope they'll do the same thing for the Pentium M. I'd love to buy a
> 64 bit Thinkpad.

Why not just buy one of the numerous AMD64 machines that are currently
available? While I like Thinkpads there's nothing magic about them.

--
--John
to email, dial "usenet" and validate
(was jclarke at eye bee em dot net)

Archived from groups: comp.sys.laptops (More info?)

"J. Clarke" <jclarke.usenet@snet.net.invalid> writes:
> > Interesting. I don't think I want to buy any more 32-bit computers
> > but I wonder if I can downclock an Athlon 64 with one of those Zalman
> > heat sinks.
>
> You could go with a Mobile Athlon 64, fits the same socket. But why
> downclock unless you're on a supertight budget?

To be able to run without a fan. Can I use a Mobile Athlon 64 at full
speed with just a passive heat sink?

> > There's also the issue of the power supply fan.
>
> Antec Phantom 350.

Fanless p/s? I haven't used a desktop machine in years, so I'm a bit
out of date.

> FDB desktop drives are also very quiet. Disk noise is not the driver when
> attempting to build a silent machine. A Tivo is quiet enough for any
> living room and it has two run-of-the-mill desktop drives, but its major
> noise source is its fan. Tivo uses the same vibration mounts as used on
> some models from Antec and several other vendors including IBM.

Once you get rid of the fan noise, disk noise becomes noticable. This
machine is for my bedroom, which needs to be quieter than a living
room needs to be.

> > Alternatively, maybe I can put the box in the closet. Do you happen
> > to know the maximum permissible length of a DVI cable?
>
> Supposedly about 9 meters. You can get DVI repeaters that let you go
> hundreds of feet.

Wow, I didn't know that. That's great, this may solve the whole problem.

> If you're serious about quiet, you might want to look into a Zalman TNN
> case. Not cheap, but it will handle nearly any processor and video board
> and needs _no_ fans. A cheaper alternative is the Zalman Reserator liquid
> cooling system, which is often criticized because it's not designed for
> overclocking but for quiet.

Again, an interesting idea. That case is about $1300 but with $1700
worth of motherboard, HD, 21" LCD, etc., the result will probably
outperform any $3000 laptop. The liquid cooled system sounds more
financially sensible though.

Thanks, I will save your post for when I actually want to buy something.

Archived from groups: comp.sys.laptops (More info?)

Peter T. Breuer wrote:

> J. Clarke <jclarke.usenet@snet.net.invalid> wrote:
>> Paul Rubin wrote:
>> > "J. Clarke" <jclarke.usenet@snet.net.invalid> writes:
>> >> > This chipset uses much less energy and actually runs faster
>> >> > than the Pentium 4 due to fewer "stops" and greater efficiency in
>> >> > processing power.
>>
>> Why don't you let him tell us what he means?
>
> He told you. See above.

He said nothing about whether he meant 'at the same clock speed' or in
absolute terms.

>> He's been arguing for abandoning the P4 and making all machines like
>> laptops for many, many posts now.
>
> One doesn't have to argue - it's been self-evident since the start of
> the P4.

Self-evident to who?

>> As for "at the same clock speed", a PIII beats a P4 at the same clock
>> speed, so I suppose we should abandon the P4 and go back to the PIII?
>
> One would hope so. The P4 was an attempt to win the hype-war over
> clock speeds. It turned out that lusers just bought the chip with the
> faster clock, no matter that it was slower! So intel had the bright
> idea of dropping everything like on-the-fly optimization, branch
> prediction, speculative execution, everything that increased the gate
> depth in the chip and made it fast, and simply crank up the clock rate
> on a skeleton architecture and sell it!

And thus they got clock speeds so high that their chips outperformed those
with all the optimizations.

> It worked - all the idiots go buy hyper-fast P4s at exorbitant prices,
> unaware that it's better to buy a slow, power-savvy sophisticated chip
> than a fast, power hungry chip that in the end runs things slower.

If in fact the "fast, power hungry chip" did "in the end run things slower",
which is not the case.

> It
> nicely occupied the market niche for ignorant savages, and made other
> hardware demands that squeezed more monies out of the masses with
> extraordinary ease.
>
> Now they have other architctures ready at more sensible clock speeds,
> they are going to have to think of some other marketing hype.

They already have. PCI Express for example.

And meanwhile, how do you explain AMD64?

> Peter

--
--John
to email, dial "usenet" and validate
(was jclarke at eye bee em dot net)

Archived from groups: comp.sys.laptops (More info?)

Paul Rubin wrote:

> "J. Clarke" <jclarke.usenet@snet.net.invalid> writes:
>> > Interesting. I don't think I want to buy any more 32-bit computers
>> > but I wonder if I can downclock an Athlon 64 with one of those Zalman
>> > heat sinks.
>>
>> You could go with a Mobile Athlon 64, fits the same socket. But why
>> downclock unless you're on a supertight budget?
>
> To be able to run without a fan. Can I use a Mobile Athlon 64 at full
> speed with just a passive heat sink?

If the heat sink is big enough. Thermaltake 103 or Sonic Tower or the
Silverstone Nitrogon SST, for example. They do need some airflow but you
can with a good case get by with a single large slow fan for everything.

>> > There's also the issue of the power supply fan.
>>
>> Antec Phantom 350.
>
> Fanless p/s? I haven't used a desktop machine in years, so I'm a bit
> out of date.

Yep. Fanless.

>> FDB desktop drives are also very quiet. Disk noise is not the driver
>> when
>> attempting to build a silent machine. A Tivo is quiet enough for any
>> living room and it has two run-of-the-mill desktop drives, but its major
>> noise source is its fan. Tivo uses the same vibration mounts as used on
>> some models from Antec and several other vendors including IBM.
>
> Once you get rid of the fan noise, disk noise becomes noticable. This
> machine is for my bedroom, which needs to be quieter than a living
> room needs to be.

Actually, my bedroom has a certain amount of ambient noise (don't ask). I
find that a little "white noise" helps me sleep. Regardless of this, the
Seagate and Samsung drives I've used make no audible noise at all unless
they are seeking. And if you're asleep they shouldn't be seeking. The
first time I fired up a recent model Barracuda it was so quiet I thought it
was dead.

>> > Alternatively, maybe I can put the box in the closet. Do you happen
>> > to know the maximum permissible length of a DVI cable?
>>
>> Supposedly about 9 meters. You can get DVI repeaters that let you go
>> hundreds of feet.
>
> Wow, I didn't know that. That's great, this may solve the whole problem.
>
>> If you're serious about quiet, you might want to look into a Zalman TNN
>> case. Not cheap, but it will handle nearly any processor and video board
>> and needs _no_ fans. A cheaper alternative is the Zalman Reserator
>> liquid cooling system, which is often criticized because it's not
>> designed for overclocking but for quiet.
>
> Again, an interesting idea. That case is about $1300 but with $1700
> worth of motherboard, HD, 21" LCD, etc., the result will probably
> outperform any $3000 laptop. The liquid cooled system sounds more
> financially sensible though.
>
> Thanks, I will save your post for when I actually want to buy something.

--
--John
to email, dial "usenet" and validate
(was jclarke at eye bee em dot net)

Archived from groups: comp.sys.laptops (More info?)

"J. Clarke" <jclarke.usenet@snet.net.invalid> writes:
> > To be able to run without a fan. Can I use a Mobile Athlon 64 at full
> > speed with just a passive heat sink?
>
> If the heat sink is big enough. Thermaltake 103 or Sonic Tower or the
> Silverstone Nitrogon SST, for example. They do need some airflow but you
> can with a good case get by with a single large slow fan for everything.

Hmm, this is getting interesting. I couldn't find the Thermaltake 103
but found the Sonic Tower:

http://www.plycon.com/cpuc/tt_sonictower.htm

Looks like it's going to need a biiiiiiiig case. I hoped to avoid
that but maybe it's worth it. Got any suggestions? I don't need a
lot of drives or anything like that inside. I do need to use a
specific AGP video card to run my old SGI 1600SW monitor.

> >> Antec Phantom 350.
> >
> > Fanless p/s? I haven't used a desktop machine in years, so I'm a bit
> > out of date.
>
> Yep. Fanless.

Wow! They're not even that expensive. Standard ATX form factor?

> of this, the Seagate and Samsung drives I've used make no audible
> noise at all unless they are seeking. And if you're asleep they
> shouldn't be seeking. The first time I fired up a recent model
> Barracuda it was so quiet I thought it was dead.

Hmm, I still find even laptop drives make noticable noise in a quiet
room, but certainly with the QuietPC enclosure, I can get the disk to
be quiet enough. So I basically just need a case with a large slow
fan. Those used to be hard to find, for reasons I've never
understood, but hopefully it's easier now. This industry seems to
really be couthing up. Back in the old days, Steve Jobs was the only
guy who had half a clue about making computers usable in living
quarters.

Archived from groups: comp.sys.laptops (More info?)

"J. Clarke" <jclarke.usenet@snet.net.invalid> writes:
> > and the built-in Broadcom wifi has no Linux support at all.
>
> This is true of many laptops--the Broadcome wifi chip is very popular for
> some reason. To get it working under Linux you use ndiswrapper and the
> Windows XP driver.

I'm trying to stick to a 100% no-closed-source policy on personal
machines; I'd otherwise probably buy a Powerbook rather than mess with
this stuff at all. Right now I'm using a Netgear (Orinoco) wifi pc
card instead of the internal wifi in my x40 for this reason. I do
unfortunately have to use Windows for work, but that's on a separate
machine, provided by the company.

> > Also, apparently the Compaq's fan is rather noisy.
>
> I think that's going to be true on any AMD64 machine.

Hmm, so we come back around to the notion of waiting for Intel to
put x86-64 extensions into the Pentium M.

> > Aside from that, IBM has been very pro-Linux, so other things being
> > equal, I prefer to buy from them on general principles.
>
> I did feel that way, but with IBM selling to the Chinese (assuming that the
> SEC approves that deal) those general principles no longer apply.

Well, the parameters will change, though I'm not clear on exactly what
the changes will be.

> However if you're interested in a 6800 series get it fast--the manufacturer
> (Arima) has a new model out and Gateway seems to be going to that one which
> is an unknown.

Hmm, well, there's always ebay.

> Incidentally, the 6800s are so well known by this time that a geek can spot
> one across a room--a friend gave one to his drop dead gorgeous daughter
> (the kind that strikes even non-geeks dumb) before she left for college and
> geeks are actually stirring up the courage to talk to her because they want
> so badly to see her machine.

Cool, I'll see if I can find a pic. I generally do prefer low-key styling
like the Thinkpads, though the Apple machines look good too.

Archived from groups: comp.sys.laptops (More info?)

"J. Clarke" <jclarke.usenet@snet.net.invalid> writes:
> You need to decide which is more important to you, politics or a working
> laptop. If politics is more important then you're going to have to make
> some compromises on hardware.

Yes, I don't mind that, I'm used to it, though the compromise I'm used
to making is simply having to spend more money to get hardware that
works with Linux (e.g. good modems vs. Winmodems). I'm more perturbed
by the situation where Linux-friendly hardware simply doesn't exist at
any price.

> There are no full-function 3d accelerated drivers for either ATI or
> Nvidia video and there are no real alternatives to those two brands
> for example. You can get by with the default SVGA drivers but you
> lose performance.

I think the Thinkpad A20p and maybe the 770 use ATI and work ok. Maybe
there are issues with 3d acceleration or something. I don't use 3d at
all so it's not an issue for me.

> > Hmm, so we come back around to the notion of waiting for Intel to
> > put x86-64 extensions into the Pentium M.
>
> Which gets around to waiting forever. Personally I'd rather put up
> with a little more fan noise than contribute further to the Intel
> near-monopoly while we're talking politics.

True, but there's already a political sacrifice (the Broadcom chip) in
the Amd64 machine. Well, I'm not in a big hurry, and I hope before
too long, x86-64 will be in everything, the way SSE is.

> And while IBM may be pro-linux they are not pro-linux to the extent that
> they will compromise their hardware designs to be linux friendly. They
> seem to favor the "Linux on the server/Windows on the desktop" model.

They used to ship some Thinkpads with Linux pre-installed; I don't
know if they still do.

Archived from groups: comp.sys.laptops (More info?)

"J. Clarke" <jclarke.usenet@snet.net.invalid> writes:
> Or one can use Linmodems.

Not sure what that is.

> > I'm more perturbed by the situation where Linux-friendly hardware
> > simply doesn't exist at any price.
>
> Define "Linux friendly".

Hardware whose programming interfaces are documented by the vendor and
the docs are available to the public without needing to sign secrecy
agreements. That means Linux developers can write and release drivers
without having to reverse engineer stuff.

> The Thinkpad 770 most assuredly does _not_ use ATI video. It uses an S3
> chip that was a dog when it was new. And while the 770 is usable under
> Linux there are functions that are not supported at all. Try to get the
> DVD decoder to work for example. And don't tell me about software
> decoders--the 770 doesn't have enough CPU power to run one.

Only the fanciest of the 770's had the dvd decoder. Mine doesn't have
it. In other regards the S3 chip has been ok under Linux. I don't
use the 770 any more but had no complaints about the graphics speed
when using it. All I really run that uses the screen is Firefox and
Emacs though, so I'm not too concerned about extreme video
performance.

I do remember something about not having enough video ram to support
24 bit color at full resolution, which was a screw, but not the fault
of the chipset as far as I know. And ram is cheap enough now that
nobody does that any more.

> Further, the ATI video that existed when the 770 was current is a far cry
> from the ATI video that exists today.

Right, it's the A20p that uses ATI then (ATI Mobility-128).

> There's a bit of a difference between Broadcom and Intel. Broadcom does not
> have the power to decree that, for example, "AGP is Dead, Long Live PCI
> Express" and have it happen. Even IBM was never that powerful.

Is there some problem with PCIX? It sounds like an improvement over
AGP. I like the idea of using it as a high speed interconnect between
multiple machines.

> That does not mean that all components in the machine are supported
> under Linux, it just means that IBM has decided that for
> applications for which Linux is appropriate those functions are not
> needed. OS/2 is one of the supported operating systems on the 770.
> That does not mean that there was ever an OS/2 driver for the DVD
> decoder.

Yeah, there was also the MWave modem in the 770, another useless
appendage. I used a PCMCIA card modem.

Archived from groups: comp.sys.laptops (More info?)

http://www.anandtech.com/cpuchipse [...] spx?i=2342

There's a good comparison here. The 2.0ghz Pentium M rivals the P4
3.2ghz in many applications. The weak level 2 cache is what does it in
though - too bad.

By the way John, you like pretty much everyone else here has missed my
point. It's my hope that everyone can change their "thinking" habits
and think "efficiency". I'm not out to make a buck so you can quit the
drivel about becoming rich. If you look at my posts, I NEVER stated
that the market was clamoring for the kind of machine I am describing.
I am aware that the majority of people could care less about having a
computer that takes up little space and uses a minimal amount of
electricity which is a shame. We need to produce more "eco-friendly"
friendly computers.

Archived from groups: comp.sys.laptops (More info?)

stevenqrdh@gmail.com wrote:
>If you look at my posts, I NEVER stated
>that the market was clamoring for the kind of machine I am describing.
>I am aware that the majority of people could care less about having a
>computer that takes up little space and uses a minimal amount of
>electricity which is a shame. We need to produce more "eco-friendly"
>friendly computers.

Remind us again why producing slow, expensive computers that can't be
expanded or modified and that no-one is going to buy is a good thing,
please?

Your business plan needs some work. 8*}

Archived from groups: comp.sys.laptops (More info?)

Check this out:

www.breezepc.co.uk

Archived from groups: comp.sys.laptops (More info?)

William P. N. Smith wrote:

> joebuckley2@mac.com wrote:
>>www.breezepc.co.uk
>
> Yeah, very impressive:
>
> 403 permission denied

The link worked well enough for me, but all it got to was yet another
cloneworks.

--
--John
to email, dial "usenet" and validate
(was jclarke at eye bee em dot net)

Archived from groups: comp.sys.laptops (More info?)

K. Bauer wrote:

>
> <William P. N. Smith> schrieb im Newsbeitrag
> news:3up561t52telpjdbutgkumfkaff4ca3nli@4ax.com...
>> joebuckley2@mac.com wrote:
>>>www.breezepc.co.uk
>>
>> Yeah, very impressive:
>>
>> 403 permission denied

Did you have a point or did you just feel that we might benefit from seeing
this again?

--
--John
to email, dial "usenet" and validate
(was jclarke at eye bee em dot net)