Fun with Windows XP
Paul Robson
Gentoo is compiling underneath me. Doing some stuff with VS.Net.
I had innumerable instability problems, the majority of which if not all
were solved by reducing the optimisation settings in the BIOS. This meant
it ran rather slowly. Too many things did not work , even with reinstalls.
The printer would suddenly just stop printing - I think it was connected
with a USB problem which only seems to exist on WinXP. Resetting the
machine caused it to work again. It would work fine for an hour, then stop
again. No reason could be found.
Some things never worked. The VS.Net shell running application was one, so
was the VS.Net updater. They crashed with odd error messages, no reports in
the event log, no real way of finding out why. I asked on the vs.net NGs
and no-one there knew either. I reinstalled it and it still didn't work.
VS.Net worked fine, though it was pretty sluggish really. This is only an
XP1900. It was quite a nice, albeit a bit overloaded.
Surprised by the general sluggishness. The machine booted quickly enough -
well quicker than Gentoo - about 2/3 of the time, but apps just didn't
multitask very well. I think the boot time is a cheat, because a lot of
times when it was loaded up, the screen would come up but the system didn't
work for another 10-20 seconds, while tray icons and other stuff were being
loaded I think. The speed tended to be erratic - sometimes it was nippy,
sometimes like treacle, and there appeared to be no pattern.
In all, it crashed totally about once every 2 hours of usage. I couldn't
figure out why.
Occasionally one would crash. XP claims you can kill it off using the
taskmanager which you get up with C/A/D. It does not work properly.
Sometimes the kill app task would appear as a new task but not kill the app
it was supposed to be killing.
At the same time I was creating a large scanned document. Windows crashed
regularly with this : it was about 50 pages of scans that was put together
to form a book in OOo. Crashed as in stopped ; if it simply keeled over I'd
blame OpenOffice.
CDWriting did not work reliably - hardly at all flat out, and about 98%
reliably at 4x (its a 12x drive) even with DMA off (which Linux also
requires on my hardware).
At one point there was a really wierd crash which caused the registration
request icon to come back - it looks like a key. After a while it went away
again. Then it came back and it needed reregistering. I've no idea why.
I did check the hardware - I've a copy of SuSE which has a memory tester
built into the install disk. Everything now seems to work fine.
The worst problem was the shutdown, which frequently simply didn't. It went
through a shutdown routine, lots of HD activity, then just stopped.
Probably about a quarter of the shutdowns ended like this. I did after
searching find that disabling an NVIDIA Service would help. It didn't.
What was the upside ? I liked Autoroute. Money was a bit frilly but not too
bad. Hearts I liked :) VS.Net had lots of bells and whistles as you would
expect.
Not keen to go back, even though I may not have the choice at a later date
:( Think I'll keep two machines, because using XP has just been a PITA.
Paul Robson
> Had to convert my machine to WinXP Pro for a couple of weeks. As I speak,
> Gentoo is compiling underneath me. Doing some stuff with VS.Net.
>
... the zip drive didnt work properly either. Saving big OO files caused a
directory too deep error - sometimes.
Peter Jensen
> What was the upside ? I liked Autoroute.
What, exactly, is that?
--
PeKaJe
Genius is pain.
-- John Lennon
Rui Malheiro
Hash: SHA1
Em Quinta, 27 de Novembro de 2003 12:43, Paul Robson escreveu:
> Not keen to go back, even though I may not have the choice at a later date
> :( Think I'll keep two machines, because using XP has just been a PITA.
Did you evaluate the possibility of setting a virtual machine on the Gentoo
host to run XP as guest OS? As long as you have a good amount of memory
(512MB should be fine) it realy is worth trying. Then you could limit the
usage of XP to your .net related tasks and do everything else on Gentoo,
like OOo and such.
- --
Rui Malheiro
"Um outro mundo é possível"
.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)
iD8DBQE/xfdO6xZVe81Ht3gRAtD1AKCUDObuuVkdm5L+htOPrgnCH+2rtACguf7v
t0uVExdH+HJV962c8GX1Sm0=
=Z5+5
-----END PGP SIGNATURE-----
Paul Robson
> Paul Robson wrote:
>
>> What was the upside ? I liked Autoroute.
>
> What, exactly, is that?
>
A route planner. Type in start and end and it works a route out for you. Bit
like multimap.com
Linønut
> Had to convert my machine to WinXP Pro for a couple of weeks. As I speak,
> Gentoo is compiling underneath me. Doing some stuff with VS.Net.
>
> I had innumerable instability problems, the majority of which if not all
> were solved by reducing the optimisation settings in the BIOS. This meant
> it ran rather slowly. Too many things did not work , even with reinstalls.
> The printer would suddenly just stop printing - I think it was connected
> with a USB problem which only seems to exist on WinXP. Resetting the
> machine caused it to work again. It would work fine for an hour, then stop
> again. No reason could be found.
>
> Some things never worked....
One of our XP machines had almost as many problems. A clean reinstall
helped, but it took the guy weeks before he lost patience and did it.
My machine has worked okay, but here are some problems:
Visual C++ .NET loses the editing cursor for about 1/2 minutes on
occasion. Nothing to do for it but wait.
The Explorer window will flicker like a sunnuvabitch as it refreshes
it self. Very unprofessional looking.
On occasion, XP will refuse to detect my thumb drive when I insert it
(this has also happened on another XP machine). The only recourse is
to reboot, after which XP will detect it.
I bought a music CD (gasp!) that has a concert DVD. WiMP and
Intervideo DVD both crash when I select "Play the Video". (mplayer
exits, too, at this same point, but xine plays this video just fine.)
Just as with Win 2000, deleting items exhibits a very long pause in
which that friggin' waste-paper animation moves with no other action.
Just as with Win 2000 and Win 2003 Enterprise Server, file-sharing is
extremely putzy and often annoying to wait for.
Tracking down hardware drivers was more difficult than expected.
Other than the last problem, these issues are annoyances that have me
sputtering daily about "monopoly crapware", and convince me that
Windows will ultimately take second place to the free UNIXen, at least
on the desktops of professionals who are in the know.
--
No, I won't fix your Windows computer!
Peter Jensen
Oh ... I was thinking automatic routing of PCB's. My professional bias
is showing. Just like when people mention 'bugs' or 'cookies', my
initial mental image is often somewhere completely different :-)
--
PeKaJe
Life is a series of rude awakenings. -- R.V. Winkle
kevin bailey
that way i can still use my debian machine for faster web
browsing/email/etc and carry out most of the work using various tools.
when win2k plays up or requires yet another reboot at least it is just in
the window - and you can install multiple virtual machines if you need
win98 for example - or take a snapshot before you install any new software
- that way you can revert if the installation causes problems,
also - if the win2k virtual machine has problems i can still browse web and
NG's to look for answers.
kev bailey
Sinister Midget
Hell, that & solitaire make it more than worth the cost and
frustration! That's not even counting the bonuses, like Klez, Swen,
CodeRed, spyware, rights transference to MICROS~1, the needed third
party programs to add some minor functionality (like popup killing and
printing to pdf), etc.
--
CodeRed - Innovative Microsoft peer-to-peer software.
Sinister Midget
A few months ago I reported similar problems and worse. I recall being
called inept and stupid by the likes of such brilliant minds as
Dickbell and his brother, WJ "Dumb" Bell. When I pointed out that I
wasn't in charge of the machines any more (though I had been until the
really smart MCSEs got over being scared of eX-Pee) I was called inept
for not being able to hold onto things, as well as stupid and
incompetent for not fixing things in the first place. I mentioned that
one machine worked a little better than the others (just barely better,
but better nonetheless) and that was the only one I'd had an
opportunity to figure a few things out with before control was given to
the really learned MCSE maintainers, and that the MCSE-maintained
machines weren't as well off as that lone machine, and they were being
kept-up by schooled and learned MICROS~1 scholars. That didn't even faze
the bright-minded smart Windoze users, and they made sure I knew it was
my fault that the MCSEs weren't able to make things work any better
than I'd done with the short stint I'd had my hands on the machines.*
In that light, I have to say you must be stupid and inept because your
MICROS~1 scholar can't do his job properly (although I doubt there
really is a "properly" for anyone to do), and it's all your fault he
doesn't know anything that will fix all of the problems you didn't
create.
* I still intend to take a picture of a machine locked-up with the
entire screen turned white. In fact, I might make it into a short
movie. The problem has diminished in the last few weeks after we found
a way to not stress the machines too much by running software mostly on
the server (an old 500MHz machine with 512M on W2K and doing all of the
work is working better than multiple 1.2GHz machines using eX-Pee and
doing their own little bit of processing - pathetic). However, it
doesn't bode well for MICROS~1 again since, as always, it requires the
use of third-party applications to work around the crappy design of
Winders.
--
.NET is basically bill gates' snake-oil solution to all those
registry/DLL hell problems. Bill Gates is a true genius. He's made
installing and maintaining windows apps so ridiculously difficult and
expensive that businesses and consumers will actually buy into the idea
of having their applications on someone else's server.
Linønut
> In that light, I have to say you must be stupid and inept because your
> MICROS~1 scholar can't do his job properly (although I doubt there
> really is a "properly" for anyone to do), and it's all your fault he
> doesn't know anything that will fix all of the problems you didn't
> create.
Actually, I installed my machine myself, and didn't allow anyone else to
touch it. As a result, I've had by far the **fewest** problems with XP.
If you want something done right, do it yourself.
> * I still intend to take a picture of a machine locked-up with the
> entire screen turned white.
I took a couple snapshots of one of the other XP boxes, showing the XP
version of the Blue Screen of Death. I'll have to post them in Linųnut
Territory.
Duane Attaway
> Had to convert my machine to WinXP Pro for a couple of weeks. As I speak,
Great read. It reminded me when I used Windows 95, 98, NT3.5, NT4.0,
etc... I liked Windows and this story brought back much nostolgia as it
was such a friendly animal.
Windws was like a pet, one that I loved very much. But this animal
would pee through the house, bring in feas, tear open the garbage
and fing it everywhere, and rip any visitors a new asshole.
I sort of miss Windows and enjoy being over at a friend's house to play
with it. Its fun to watch them play with it too and mess with the
settings to get things working. Unfortunately, I wasn't smart enough to
keep it running on my computers and had to settle for the GNU utilities
running on the Linux kernel. I often do miss the challenges Windows
gave me.
Thanks again for the story!
Paul Robson
> A few months ago I reported similar problems and worse. I recall being
> called inept and stupid by the likes of such brilliant minds as
> Dickbell and his brother, WJ "Dumb" Bell. When I pointed out that I
> wasn't in charge of the machines any more (though I had been until the
> really smart MCSEs got over being scared of eX-Pee)
One of the really disappointing things was alt.os.windows-xp which I posted
a couple of problems on - the large files/Zip drive problem. I got no
answer and as far as I could see nor did anyone else, it's full of crap
from the likes of Kaditcha Man about how Linux users are gay. Why a linux
user would bother with an XP user group is beyond me.
This does not contrast very well with alt.os.linux.mandrake / slackware /
redhat / suse etc. where any query will generate a stack of answers,
suggestions, or at worst, a "no idea mate, sorry" response.
Sinister Midget
The ones where the skunks-at-the-picnic don't hang out aren't much
better for help, though they're a bit more civil. Mostly they just
provide a few stock answers to the multitude of problems. And, like
Dickbell et al, if you tell them it didn't work, some will start
accusing you of making it up or just calling some names. But at least
some of them /do/ try.
There /are/ a few that have a bit more than a clue. Wading through the
others is a bit difficult sometimes.
--
BugBear - Innovative Microsoft peer-to-peer software.
The Onion Man
> Not keen to go back, even though I may not have the choice at a later date
> :( Think I'll keep two machines, because using XP has just been a PITA.
What really gets me is when people who have paid good money for such a POS
expect help & sympathy. I find it hard to convey to them that I am not
interested, I don't care and I never will.
--
Ian
He knows his onions.
The Onion Man
>> What was the upside ? I liked Autoroute.
>
> What, exactly, is that?
I still have one of the original versions of Autoroute before it was
bought by MS.
It's a DOS program. It was quite fast on a P133, runs lightening fast on
any more modern machine and apart from the fact that its maps are a
little out of date, far out performs any other route software I've ever
used. It doesn't do fancy things like recommend pubs, parks or playpens to
stop off at on your journey, it just gives you the shortest route from A
to B. Perfect!
The Onion Man
>> A route planner. Type in start and end and it works a route out for
>> you. Bit like multimap.com
>
> Oh ... I was thinking automatic routing of PCB's. My professional bias
> is showing. Just like when people mention 'bugs' or 'cookies', my
> initial mental image is often somewhere completely different :-)
Cup of Java anyone? Freshly brewed it'll go nicely with those Choc chip
cookies....
Sugapablo
> redhat / suse etc. where any query will generate a stack of answers,
> suggestions, or at worst, a "no idea mate, sorry" response.
Agreed. I have found a great deal of help over the past year and a half
from alt.os.linux.mandrake, comp.os.linux.misc, and my local LUG
(WPLUG).
But I think it's the nature of what Linux is, a community. People not
only contribute code, but they contribute help for others wading through
the "New World".
Now granted, you might get RTFMed, but at least with Linux there usually
IS a FM. :)
--
[ Sugapablo ]
[ http://www.sugapablo.com <--music ]
[ http://www.sugapablo.net <--personal ]
[ suga...@12jabber.com <--jabber IM ]
Milo T.
> Had to convert my machine to WinXP Pro for a couple of weeks. As I speak,
> Gentoo is compiling underneath me. Doing some stuff with VS.Net.
>
> I had innumerable instability problems, the majority of which if not all
> were solved by reducing the optimisation settings in the BIOS. This meant
> it ran rather slowly. Too many things did not work , even with reinstalls.
> The printer would suddenly just stop printing - I think it was connected
> with a USB problem which only seems to exist on WinXP. Resetting the
> machine caused it to work again. It would work fine for an hour, then stop
> again. No reason could be found.
[litany of problems deleted].
Hmmm... so let's see... you have obvious hardware problems of some kind
with your system (you shouldn't need to reduce the optimization settings in
the BIOS), and you're complaining about XP crashing a lot?
Did it occur to you that the problem you have might not be just a "memory"
problem? SuSe could quite happily test the memory until kingdom come; it
just wouldn't make any difference.
OK; I'm going to stick my neck out on this one: You built this machine
yourself, and it has an AMD processor in it.
John Bailo
> On Thu, 27 Nov 2003 12:43:34 +0000, Paul Robson wrote:
>
>
>>Had to convert my machine to WinXP Pro for a couple of weeks. As I speak,
>>Gentoo is compiling underneath me. Doing some stuff with VS.Net.
>>
>>I had innumerable instability problems, the majority of which if not all
>>were solved by reducing the optimisation settings in the BIOS. This meant
>>it ran rather slowly. Too many things did not work , even with reinstalls.
> OK; I'm going to stick my neck out on this one: You built this machine
> yourself, and it has an AMD processor in it.
Wait! You're saying that he can't build his own machine and he
can't use an AMD chip and still be XP compatible????
Ok, stick your neck out.
CHOP !!!
There, now I'll be eating roast SCOoke for Thanksgiving.
Linux Sucks
Paul Robson <auti...@autismuk.muralichucks.freeserve.co.uk> wrote:
>In all, it crashed totally about once every 2 hours of usage. I couldn't
>figure out why.
You should stop using computers if Windows XP crashes once every 2
hours.
You are totally stupid and this post proves it.
Milo T.
> Milo T. wrote:
>> On Thu, 27 Nov 2003 12:43:34 +0000, Paul Robson wrote:
>>
>>
>>>Had to convert my machine to WinXP Pro for a couple of weeks. As I speak,
>>>Gentoo is compiling underneath me. Doing some stuff with VS.Net.
>>>
>>>I had innumerable instability problems, the majority of which if not all
>>>were solved by reducing the optimisation settings in the BIOS. This meant
>>>it ran rather slowly. Too many things did not work , even with reinstalls.
>
>> OK; I'm going to stick my neck out on this one: You built this machine
>> yourself, and it has an AMD processor in it.
>
> Wait! You're saying that he can't build his own machine and he
> can't use an AMD chip and still be XP compatible????
No. The true diagnosis comes later, if he confirms the above.
I'm running on a 1.4GHz athlon here. Runs fine. Zero problems.
I just have a sneaky feeling, having built a fair number of my own systems,
that I've seen this behavior before. And it can be one of two things (in
the typical case).
Ori
>
> I just have a sneaky feeling, having built a fair number of my own systems,
> that I've seen this behavior before. And it can be one of two things (in
> the typical case).
I see.
Enlighten us please...
Ori
>
> I took a couple snapshots of one of the other XP boxes, showing the XP
> version of the Blue Screen of Death. I'll have to post them in Linųnut
> Territory.
>
Is that the best you can do?
I still delight in reproducing one error for my friends:
The full text of this error message is as follows:
"No Error
No Error"
It has about a 90% chance of occuring whenever I try printing over the
network from CUPS (on Slackware) to a networked printer (on Windows XP pro)
The other ten percent it gives me an "Emulation error".
I've never heard of this problem before, and can't find a solution
(although I'll bet I'm misconfiguring something). although I must admit
I've been too lazy to do any real searches yet.
Milo T.
No. Not until I have more data.
Paul Robson
> On Thu, 27 Nov 2003 12:43:34 +0000, Paul Robson wrote:
>
>> Had to convert my machine to WinXP Pro for a couple of weeks. As I speak,
>> Gentoo is compiling underneath me. Doing some stuff with VS.Net.
>>
>> I had innumerable instability problems, the majority of which if not all
>> were solved by reducing the optimisation settings in the BIOS. This meant
>> it ran rather slowly. Too many things did not work , even with
>> reinstalls. The printer would suddenly just stop printing - I think it
>> was connected with a USB problem which only seems to exist on WinXP.
>> Resetting the machine caused it to work again. It would work fine for an
>> hour, then stop again. No reason could be found.
> [litany of problems deleted].
>
> Hmmm... so let's see... you have obvious hardware problems of some kind
> with your system (you shouldn't need to reduce the optimization settings
> in the BIOS), and you're complaining about XP crashing a lot?
Well, that would be fine if the machine didn't run Linux perfectly before
and afterwards on the same optimisation settings. I did wonder seriously if
the USB Ports were broken somehow but now they work fine again. If it
wasn't for that I'd agree the hardware was at fault.
Advice on reducing the optimisations is (I think) on the BSOD. It offers
this as a fixing suggestion. They are now all back on again.
> Did it occur to you that the problem you have might not be just a "memory"
> problem? SuSe could quite happily test the memory until kingdom come; it
> just wouldn't make any difference.
Yes, but that was an easy simple test I could do, and because there was no
real repeatable pattern to the crashes. I couldn't find any sequence of
actions that caused them to reliably keel over ; it just died periodically.
It wasn't overheating. This made me think memory.
> OK; I'm going to stick my neck out on this one: You built this machine
> yourself, and it has an AMD processor in it.
75%. It does have an Athlon XP in it. It was built for me by a local
supplier, who I've probably bought 20-30 machines off over the years (not
all for me personally obviously !). It's about a year old I think, has run
Gentoo for most of its life, without a problem. The H/W is a VIA KT266/A
Mobo, Nvidia Geforce II GFX and XP1900, most of the other stuff is on the
mobo. It is not a particularly complex piece of hardware.
Probably about a third of the crashes were USBUHCI.SYS related which I
presume is the XP equivalent of the similarly named kernel module. I
couldn't find anyone else with a similar problem, the newsgroups are full
of "Linux is for gayboys" posts. It's like my vsshell (sic?) problem.
No-one has any suggestion and there's no clue from the error or the log
precisely why it doesn't work.
XP is fine until you get problems, then its a bloody nightmare. The
printer : I had a similar problem with CUPS. Turning the logging on full
revealed ghostscript was crashing. Running it manually revealed it
segfaulted. I rebuilt it (Gentoo) and everything worked again (powerout
problem originally I think). I cannot see how you can do this kind of
thing in XP. You can reinstall the drivers. You can reinstall Windows. I'd
agree the services thing is much better but still way short.
Even in the helpful newsgroups people had no answers. MSDE worked okay, but
did not use the correct name of the database, so every time I wanted to use
it I had to type the computer name and database name into the dialog box
and start it by hand. I asked in the msde newsgroup how I could set this
default so it just starts (I know that it will autostart). I got a few
basic answers - which I'd already checked - but no real help.
If you think this is imaginary look on DejaNews for MSDE and autismuk, the
thread is called "Sorry this is a dumb question but I can't find the
answer". Incidentally the zip drive thread - no answers - you can see this
by dejaing for xp zip autismuk "Zip drive doesn't copy big files".
I know it sounds like a typical "XP is crap" post but it's actually what
happened.
Paul Robson
Okay, ask away , privately if you like because no-one else has come up with
an answer & I'd really like to know. Remove muralichucks and you have my
correct email address.
Pete Goodwin
[snip]
> OK; I'm going to stick my neck out on this one: You built this machine
> yourself, and it has an AMD processor in it.
I built my own AMD system and it works just fine with Windows XP.
--
Pete Goodwin, using XanaNews 1.15.8.2 on Windows 2000.
"If the dogs are barking at your heels, you know you're leading the
pack."
Peter Köhlmann
> Milo T. wrote:
>
> [snip]
>
>> OK; I'm going to stick my neck out on this one: You built this machine
>> yourself, and it has an AMD processor in it.
>
> I built my own AMD system and it works just fine with Windows XP.
>
Sure, 2NIC Pete, sure
For suitably small values of "works"
--
Real programmers don't comment their code. It was hard to write,
it should be hard to understand.
Milo T.
> Milo T. wrote:
>
> [snip]
>
>> OK; I'm going to stick my neck out on this one: You built this machine
>> yourself, and it has an AMD processor in it.
>
> I built my own AMD system and it works just fine with Windows XP.
Same here.
John Bailo
Great -- now trot off to the Gate$ mansion
to tell Bill and Steve and
to get your fees and the new autographed
pictures of Michael Jackson waiting for you
in the Pete Townsend memorial shrine.
Milo T.
> Well, that would be fine if the machine didn't run Linux perfectly before
> and afterwards on the same optimisation settings. I did wonder seriously if
> the USB Ports were broken somehow but now they work fine again. If it
> wasn't for that I'd agree the hardware was at fault.
>
> Advice on reducing the optimisations is (I think) on the BSOD. It offers
> this as a fixing suggestion. They are now all back on again.
>
>> Did it occur to you that the problem you have might not be just a "memory"
>> problem? SuSe could quite happily test the memory until kingdom come; it
>> just wouldn't make any difference.
>
> Yes, but that was an easy simple test I could do, and because there was no
> real repeatable pattern to the crashes. I couldn't find any sequence of
> actions that caused them to reliably keel over ; it just died periodically.
> It wasn't overheating. This made me think memory.
Don't be so sure about the overheating...
>> OK; I'm going to stick my neck out on this one: You built this machine
>> yourself, and it has an AMD processor in it.
>
> 75%. It does have an Athlon XP in it. It was built for me by a local
> supplier, who I've probably bought 20-30 machines off over the years (not
> all for me personally obviously !). It's about a year old I think, has run
> Gentoo for most of its life, without a problem. The H/W is a VIA KT266/A
> Mobo, Nvidia Geforce II GFX and XP1900, most of the other stuff is on the
> mobo. It is not a particularly complex piece of hardware.
>
> Probably about a third of the crashes were USBUHCI.SYS related which I
> presume is the XP equivalent of the similarly named kernel module. I
> couldn't find anyone else with a similar problem, the newsgroups are full
> of "Linux is for gayboys" posts. It's like my vsshell (sic?) problem.
> No-one has any suggestion and there's no clue from the error or the log
> precisely why it doesn't work.
OK... my guess so far is this:
1. Check that you have a 350W power supply in that case. Make sure it's a
real 350W power supply, certified by AMD (check on the web). That'll cause
a lot of problems, and caused a lot of crashes on one system I built -
they're *very* picky about PSU quality. (Depending on your CPU, you may get
away with only 300W).
2. If that doesn't clear up a lot of the issues, crack the fan off the CPU
die. Check for spots where the grease is thin and not getting good contact
on the back of the fan. Get some isopropyl alcohol, and clean the thermal
grease off the CPU (you'll need to be very careful, and use a lint free
cloth). Replace the grease with fresh grease, ensuring good coverage and
replace the CPU fan.
As for the USB problems:
Flash the BIOS with the latest BIOS from the manufacturer's website
(Asus?). Download and install the motherboard drivers from the website
while you're at it. You're looking for new VIA X-in-1 drivers for the
motherboard, which will update the USB firmware, Northbridge and a few
other bits and pieces.
As to why you don't see this with Linux?
Linux does tend to run CPUs without giving them quite as extensive a
workout. *shrugs*.
John Bailo
> Linux does tend to run CPUs without giving them quite as extensive a
> workout. *shrugs*.
True. Because linux has not spent the last
20 years trying to warp hardware into a lock
in with the LoseDOS 1900 and YPay application
loaders.
Linux is just fast and efficient, so it sheds
the dross.
Paul Robson
>
> OK... my guess so far is this:
>
> 1. Check that you have a 350W power supply in that case. Make sure it's a
> real 350W power supply, certified by AMD (check on the web). That'll cause
> a lot of problems, and caused a lot of crashes on one system I built -
> they're very picky about PSU quality. (Depending on your CPU, you may get
> away with only 300W).
>
It's a 300W seems to be certified for Pentium.
> 2. If that doesn't clear up a lot of the issues, crack the fan off the CPU
> die. Check for spots where the grease is thin and not getting good contact
> on the back of the fan. Get some isopropyl alcohol, and clean the thermal
> grease off the CPU (you'll need to be very careful, and use a lint free
> cloth). Replace the grease with fresh grease, ensuring good coverage and
> replace the CPU fan.
>
I'll give it a go, thanks. I have actually run with the side of the case off
to try and solve the overheating (?) issues if such exist.
> As for the USB problems:
>
> Flash the BIOS with the latest BIOS from the manufacturer's website
> (Asus?).
Appears to be the latest.
>Download and install the motherboard drivers from the website
> while you're at it. You're looking for new VIA X-in-1 drivers for the
> motherboard, which will update the USB firmware, Northbridge and a few
> other bits and pieces.
>
I was running 449 which as of last check was the most up to date, it's from
September I think.
>
> As to why you don't see this with Linux?
>
> Linux does tend to run CPUs without giving them quite as extensive a
> workout. shrugs.
Eh ? You sure 'bout this ? One problem I've heard of with Linux similarly is
that it sometimes shows up duff RAM because it flogs it harder than
Windows. Given that a CPU is well a CPU and runs more or less consistently,
and the Windows code can't be compiled for more than a Pentium (on box)
whereas the Gentoo is built of an Athlon XP, how'd you come to that
conclusion ?
Milo T.
> Milo T. wrote:
>
>>
>> OK... my guess so far is this:
>>
>> 1. Check that you have a 350W power supply in that case. Make sure it's a
>> real 350W power supply, certified by AMD (check on the web). That'll cause
>> a lot of problems, and caused a lot of crashes on one system I built -
>> they're very picky about PSU quality. (Depending on your CPU, you may get
>> away with only 300W).
>>
>
> It's a 300W seems to be certified for Pentium.
OK; you might want to double check that it's on AMD's "good" list... if
not, see if you can borrow an AMD-compliant PSU from someone. That was the
biggest problem I had when I built two systems in 2002.
Even as far back as Slackware 2.0, I had a system which would crash within
minutes under NT 4.0, and work fine under Linux and DOS. Once I replaced
the RAM on that system, both seemed to work fine.
It's odd, but there are very definitely different characteristic usages of
the CPU for Linux and Windows; I wish I knew exactly what the difference
was though :)
Magic Nose Goblin
> Had to convert my machine to WinXP Pro for a couple of weeks. As I speak,
> Gentoo is compiling underneath me. Doing some stuff with VS.Net.
>
> I had innumerable instability problems, the majority of which if not all
> were solved by reducing the optimisation settings in the BIOS. This meant
> it ran rather slowly. Too many things did not work , even with reinstalls.
> The printer would suddenly just stop printing - I think it was connected
> with a USB problem which only seems to exist on WinXP. Resetting the
> machine caused it to work again. It would work fine for an hour, then stop
> again. No reason could be found.
Interesting...my father is reporting the same thing on HIS new computer.
Magic Nose Goblin
Smart people don't bother wasting time figuring out why poorly-designed
spaghetti-ware crashes... they just move on to properly designed software
like Linux.
Jim Richardson
Hash: SHA1
have to point something out here, the purpose of the thermal grease is
to fill in the gaps, caused by the small unevenness of the cpu top, and
heatsink. It is *not* to have a complete layer of grease, as the thermal
grease is not as good a heat conductor as a direct mating of the HS and
CPU, but it's is far better than air. If you have thin spots in the
grease coverage, that's fine, provided they are thin, (or bare of grease
entirely) because the HS and CPU actually touch there.
Of you wanted to spend the time (and money) you could machine the
surface of the HS, *really* flat, polished, and then polish the top of
the CPU flat, and just hold the two together, that would offer the
lowest thermal resistance. But it's usually not cost effective, and
there's some risk to the CPU in the process. (Not to mention having to
clean up the polishing compound without contaminating the
pins/connections. )
> As for the USB problems:
>
> Flash the BIOS with the latest BIOS from the manufacturer's website
> (Asus?). Download and install the motherboard drivers from the website
> while you're at it. You're looking for new VIA X-in-1 drivers for the
> motherboard, which will update the USB firmware, Northbridge and a few
> other bits and pieces.
>
>
> As to why you don't see this with Linux?
>
> Linux does tend to run CPUs without giving them quite as extensive a
> workout. *shrugs*.
A nice way of saying that Linux, does more, with less CPU power.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)
iD8DBQE/x70vd90bcYOAWPYRAnWaAJ45HwlCrJa6LfUHJRjY4xCQ8YVKqwCgmFW9
WcCV5FBzaY/KNTIEYKBuaXE=
=0LmD
-----END PGP SIGNATURE-----
--
Jim Richardson http://www.eskimo.com/~warlock
Step by step, day by day, machine by machine, the penguins march forward.
paul cooke
theres an update for XP to fix some USB problems that have appeared for XP
with SP1...
<http://support.microsoft.com/?kbid=822603>
sounds like your problems might have been caused by the machine having gone
into suspend mode...
SP1 also updated a lot of the USB code.
--
COMPUTER POWER TO THE PEOPLE! DOWN WITH CYBERCRUD!
Jim Richardson
Hash: SHA1
Well, one big difference, was how Linux and non-NT windows handled cpu
idle, as in, Linux did, and Windows, didn't.
Other than that. If you have lots of RAM, Linux will happily use it for
cache without any intervention on the part of the user, I don't know how
well/aggresively the NT series caches in RAM.
Agreed, there's a lot of room there for differences, and it's quite
possible for hardware to be bad, and affect either system differently.
But in the case at hand, it sure seems more likely that some interaction
within the XP system is fucking things up. It *could* be hardware, but
I'd bet differently based on the current info available. It will be
interesting if we can find the problem.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)
iD8DBQE/x8zUd90bcYOAWPYRApBxAKDEPEIfWnEAmY4dNWxIAxUjfarGjwCgxHxu
5NLwhyWJUsRLqprK++ABb1E=
=40Hx
-----END PGP SIGNATURE-----
--
Jim Richardson http://www.eskimo.com/~warlock
E pluribus Linux
Erik Funkenbusch
>> Hmmm... so let's see... you have obvious hardware problems of some kind
>> with your system (you shouldn't need to reduce the optimization settings
>> in the BIOS), and you're complaining about XP crashing a lot?
>
> Well, that would be fine if the machine didn't run Linux perfectly before
> and afterwards on the same optimisation settings. I did wonder seriously if
> the USB Ports were broken somehow but now they work fine again. If it
> wasn't for that I'd agree the hardware was at fault.
That doesn't really mean a lot. Often, the linux drivers don't take full
advantage of a particular chipset (either because the manufacturer won't
release details or because the driver author doesn't want to do the kinds
of nasty hacks that chipset driver writers often do to make them work
fully).
> Advice on reducing the optimisations is (I think) on the BSOD. It offers
> this as a fixing suggestion. They are now all back on again.
Generally this is related to memory timings and the like. Linux often sets
the memory timings conservitively. This tends to reduce performance, but
is more stable. How the memory controller is programmed is going to be
quite different between OS's, and can hide or expose subtle flaws in the
hardware or drivers.
>> OK; I'm going to stick my neck out on this one: You built this machine
>> yourself, and it has an AMD processor in it.
>
> 75%. It does have an Athlon XP in it. It was built for me by a local
> supplier, who I've probably bought 20-30 machines off over the years (not
> all for me personally obviously !). It's about a year old I think, has run
> Gentoo for most of its life, without a problem. The H/W is a VIA KT266/A
> Mobo, Nvidia Geforce II GFX and XP1900, most of the other stuff is on the
> mobo. It is not a particularly complex piece of hardware.
Hmm.. the KT266A chipset is quite picky, and VIA's 4-in-1 drivers are a
crapshoot. You often have to rollback to a previous release of the drivers
when you can find a stable one. This is 100% a via problem.
> Probably about a third of the crashes were USBUHCI.SYS related which I
> presume is the XP equivalent of the similarly named kernel module. I
> couldn't find anyone else with a similar problem, the newsgroups are full
> of "Linux is for gayboys" posts. It's like my vsshell (sic?) problem.
> No-one has any suggestion and there's no clue from the error or the log
> precisely why it doesn't work.
Again, chipset related. I'd place my money on the 4-in-1 drivers. You
have to dink with them quite a bit to get it stable in some cases. VIA
needs to be shot for the poor quality of those drivers.
> Even in the helpful newsgroups people had no answers. MSDE worked okay, but
> did not use the correct name of the database, so every time I wanted to use
> it I had to type the computer name and database name into the dialog box
> and start it by hand. I asked in the msde newsgroup how I could set this
> default so it just starts (I know that it will autostart). I got a few
> basic answers - which I'd already checked - but no real help.
if you define an ODBC dsn then you can select the default database name.
Otherwise, you ahve to include the database name in the connection request.
SQL Serve (and MSDE is just a limited version of SQL Server) doesn't have a
way to define a default database.
Paul Robson
> theres an update for XP to fix some USB problems that have appeared for XP
> with SP1...
>
> <http://support.microsoft.com/?kbid=822603>
>
> sounds like your problems might have been caused by the machine having
> gone into suspend mode...
>
> SP1 also updated a lot of the USB code.
I've got SP1 but not sure about that one. I think it was installed ; if it
was required by windowsupdate.com it would have been installed. Assuming
Windows would install it ; a couple required IE6. I had IE6 :)
Paul Robson
>> OK; I'm going to stick my neck out on this one: You built this machine
>>> yourself, and it has an AMD processor in it.
>>
>> 75%. It does have an Athlon XP in it. It was built for me by a local
>> supplier, who I've probably bought 20-30 machines off over the years (not
>> all for me personally obviously !). It's about a year old I think, has
>> run Gentoo for most of its life, without a problem. The H/W is a VIA
>> KT266/A Mobo, Nvidia Geforce II GFX and XP1900, most of the other stuff
>> is on the mobo. It is not a particularly complex piece of hardware.
>
> Hmm.. the KT266A chipset is quite picky, and VIA's 4-in-1 drivers are a
> crapshoot. You often have to rollback to a previous release of the
> drivers when you can find a stable one. This is 100% a via problem.
VIAs support seems to be quite good, and no-one else seems to have this
problem.
>> Probably about a third of the crashes were USBUHCI.SYS related which I
>> presume is the XP equivalent of the similarly named kernel module. I
>> couldn't find anyone else with a similar problem, the newsgroups are full
>> of "Linux is for gayboys" posts. It's like my vsshell (sic?) problem.
>> No-one has any suggestion and there's no clue from the error or the log
>> precisely why it doesn't work.
>
> Again, chipset related. I'd place my money on the 4-in-1 drivers. You
> have to dink with them quite a bit to get it stable in some cases. VIA
> needs to be shot for the poor quality of those drivers.
>
You're just blaming anyone but Microsoft. Whose fault other than theirs can
the ZIP drive problem be ? The vsupdate and vs(shell?) problem ?
>> Even in the helpful newsgroups people had no answers. MSDE worked okay,
>> but did not use the correct name of the database, so every time I wanted
>> to use it I had to type the computer name and database name into the
>> dialog box and start it by hand. I asked in the msde newsgroup how I
>> could set this default so it just starts (I know that it will autostart).
>> I got a few basic answers - which I'd already checked - but no real help.
>
> if you define an ODBC dsn then you can select the default database name.
> Otherwise, you ahve to include the database name in the connection
> request. SQL Serve (and MSDE is just a limited version of SQL Server)
> doesn't have a way to define a default database.
Yes it can, because I've installed it before and it definitely can be set up
so that when you boot Windows ServerManager is running with a default
database. It's just no-one seems to know how.
Erik Funkenbusch
> Erik Funkenbusch wrote:
>
>>> OK; I'm going to stick my neck out on this one: You built this machine
>>>> yourself, and it has an AMD processor in it.
>>>
>>> 75%. It does have an Athlon XP in it. It was built for me by a local
>>> supplier, who I've probably bought 20-30 machines off over the years (not
>>> all for me personally obviously !). It's about a year old I think, has
>>> run Gentoo for most of its life, without a problem. The H/W is a VIA
>>> KT266/A Mobo, Nvidia Geforce II GFX and XP1900, most of the other stuff
>>> is on the mobo. It is not a particularly complex piece of hardware.
>>
>> Hmm.. the KT266A chipset is quite picky, and VIA's 4-in-1 drivers are a
>> crapshoot. You often have to rollback to a previous release of the
>> drivers when you can find a stable one. This is 100% a via problem.
>
> VIAs support seems to be quite good, and no-one else seems to have this
> problem.
Haven't done a search on google regarding via and unstable, have you?
>>> Probably about a third of the crashes were USBUHCI.SYS related which I
>>> presume is the XP equivalent of the similarly named kernel module. I
>>> couldn't find anyone else with a similar problem, the newsgroups are full
>>> of "Linux is for gayboys" posts. It's like my vsshell (sic?) problem.
>>> No-one has any suggestion and there's no clue from the error or the log
>>> precisely why it doesn't work.
>>
>> Again, chipset related. I'd place my money on the 4-in-1 drivers. You
>> have to dink with them quite a bit to get it stable in some cases. VIA
>> needs to be shot for the poor quality of those drivers.
>
> You're just blaming anyone but Microsoft. Whose fault other than theirs can
> the ZIP drive problem be ? The vsupdate and vs(shell?) problem ?
Depends. Is an IDE, SCSI, or Paralell Zip drive? The paralell drivers
that iomega ships haven't been updated in years and are quite flaky.
>> if you define an ODBC dsn then you can select the default database name.
>> Otherwise, you ahve to include the database name in the connection
>> request. SQL Serve (and MSDE is just a limited version of SQL Server)
>> doesn't have a way to define a default database.
>
> Yes it can, because I've installed it before and it definitely can be set up
> so that when you boot Windows ServerManager is running with a default
> database. It's just no-one seems to know how.
Server Manager? You mean Enterprise Manager?
Linønut
>> Hmm.. the KT266A chipset is quite picky, and VIA's 4-in-1 drivers are a
>> crapshoot. You often have to rollback to a previous release of the
>> drivers when you can find a stable one. This is 100% a via problem.
>
> VIAs support seems to be quite good, and no-one else seems to have this
> problem.
My mobo with Via Apollo KT133 (/var/log/dmesg) works very well, and it's
a cheap-ass mobo.
> You're just blaming anyone but Microsoft.
This is a reflex with Erik.
--
No, I won't fix your Windows computer!
Linønut
> Haven't done a search on google regarding via and unstable, have you?
I got 45 results for funkenbusch+unstable. Including this little gem:
http://lists.debian.org/debian-user/2001/debian-user-200108/msg03566.html
> Depends. Is an IDE, SCSI, or Paralell Zip drive? The paralell drivers
> that iomega ships haven't been updated in years and are quite flaky.
Not updated in years? A very fishy claim.
Sinister Midget
> Fearing a spontaneous XP reboot, Erik Funkenbusch mumbled this incantation:
>
>> Haven't done a search on google regarding via and unstable, have you?
>
> I got 45 results for funkenbusch+unstable. Including this little gem:
Darn. I was gonna do a whole thread in his honor:
windows unstable 422,000
microsoft unstable 181,000
me instable 1,140,000
95 unstable 603,000
98 unstable 511,000
2000 unstable 2,040,000
xp unstable 118,000
via unstable 901,000
chocolate unstable 35,200
nuclear unstable 273,000
Windoze wins again! It even beats out "nuclear" which should provide
some clues to us all!
--
Yesterday it worked.
Today it is not working.
Windows is like that.
Jim Richardson
Hash: SHA1
On Sat, 29 Nov 2003 09:43:46 -0600,
Linųnut <linųn...@bone.com> wrote:
> Fearing a spontaneous XP reboot, Erik Funkenbusch mumbled this incantation:
>
>> Haven't done a search on google regarding via and unstable, have you?
>
> I got 45 results for funkenbusch+unstable. Including this little gem:
>
> http://lists.debian.org/debian-user/2001/debian-user-200108/msg03566.html
>
amusingly enough, after getting two responces showing the "problem" and
the solution, Erik never replied...
>> Depends. Is an IDE, SCSI, or Paralell Zip drive? The paralell drivers
>> that iomega ships haven't been updated in years and are quite flaky.
>
> Not updated in years? A very fishy claim.
>
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)
iD8DBQE/yPisd90bcYOAWPYRAr5wAJ9KHsZeUaLyrdUl1cExCmnjOH/pggCfdf4I
pJ4G5894XPGMDKdjvCTIgAs=
=iCd5
-----END PGP SIGNATURE-----
--
Jim Richardson http://www.eskimo.com/~warlock
Any nitwit can understand computers. Many do.
Milo T.
> have to point something out here, the purpose of the thermal grease is
> to fill in the gaps, caused by the small unevenness of the cpu top, and
> heatsink. It is *not* to have a complete layer of grease, as the thermal
> grease is not as good a heat conductor as a direct mating of the HS and
> CPU, but it's is far better than air. If you have thin spots in the
> grease coverage, that's fine, provided they are thin, (or bare of grease
> entirely) because the HS and CPU actually touch there.
>
> Of you wanted to spend the time (and money) you could machine the
> surface of the HS, *really* flat, polished, and then polish the top of
> the CPU flat, and just hold the two together, that would offer the
> lowest thermal resistance. But it's usually not cost effective, and
> there's some risk to the CPU in the process. (Not to mention having to
> clean up the polishing compound without contaminating the
> pins/connections. )
Actually, I've had problems with heat pipe systems overheating unless I
smear an entire layer of goop over *everything*. Of course, they had a
"protective copper shim" on top of the CPU die to stop the die from
cracking against the heat sink. *shrugs*
Also note that the grease, being non-conductive, can spread the heat over a
wider area than the heatsink.
>> As for the USB problems:
>>
>> Flash the BIOS with the latest BIOS from the manufacturer's website
>> (Asus?). Download and install the motherboard drivers from the website
>> while you're at it. You're looking for new VIA X-in-1 drivers for the
>> motherboard, which will update the USB firmware, Northbridge and a few
>> other bits and pieces.
>>
>>
>> As to why you don't see this with Linux?
>>
>> Linux does tend to run CPUs without giving them quite as extensive a
>> workout. *shrugs*.
>
> A nice way of saying that Linux, does more, with less CPU power.
Not necessarily. It may well do the same or less with less CPU power.
Unless you've made exact measurements of the same system running the same
software with the same hardware and the same APIs. Which, of course, is
impossible.
Jim Richardson
Hash: SHA1
On Sat, 29 Nov 2003 20:40:48 GMT,
Milo T. <fanta...@malaprop.net> wrote:
> On Fri, 28 Nov 2003 13:25:03 -0800, Jim Richardson wrote:
>> have to point something out here, the purpose of the thermal grease is
>> to fill in the gaps, caused by the small unevenness of the cpu top, and
>> heatsink. It is *not* to have a complete layer of grease, as the thermal
>> grease is not as good a heat conductor as a direct mating of the HS and
>> CPU, but it's is far better than air. If you have thin spots in the
>> grease coverage, that's fine, provided they are thin, (or bare of grease
>> entirely) because the HS and CPU actually touch there.
>>
>> Of you wanted to spend the time (and money) you could machine the
>> surface of the HS, *really* flat, polished, and then polish the top of
>> the CPU flat, and just hold the two together, that would offer the
>> lowest thermal resistance. But it's usually not cost effective, and
>> there's some risk to the CPU in the process. (Not to mention having to
>> clean up the polishing compound without contaminating the
>> pins/connections. )
>
> Actually, I've had problems with heat pipe systems overheating unless I
> smear an entire layer of goop over *everything*. Of course, they had a
> "protective copper shim" on top of the CPU die to stop the die from
> cracking against the heat sink. *shrugs*
>
if you are talking about smearing the grease along the pipes to help
conduction from the cpu section, to the heatsink, depending on the
design of the heat pipe, this may actually *reduce* the effective heat
transfer.
If the heatpipe is solid, relying on conduction along a solid (or
tubular) metal bar, then all well and good, you've increased the volume
by adding the goop. *But*, if instead it's a "fluid transvection system"
meaning, the pipes are filled with a fluid that evapourates from the hot
side, and condenses on the cool heatsink side, then flows back to the
hotside, ad infinitum, then by *improving* the radiative cooling of the
pipe itself, you can contain the heat within the enclosure until such a
time as the deltaT between the pipe, and the heatsink allows the
"transvection" to proceed, Unfortunately, that may be after the CPU has
slagged itself.
Now, I don't know if any CPU heatpipe systems are built with that type
of heatpipe, it's usually used in higher temp systems, often with liquid
sodium as a fluid, but I suspect that plain water with some simple
additives would work well for the temps involved in the CPU systems.
For myself, I follow the manufacturers instructions in such matters,
unless I *know* they are fucked up.
> Also note that the grease, being non-conductive, can spread the heat over a
> wider area than the heatsink.
>
Actually, many of the thermal compounds used now, have a high copper or
silver content, so they aren't as non-conductive (electrically, which is
what I assumed you meant) as you might think.
>
>>> As for the USB problems:
>>>
>>> Flash the BIOS with the latest BIOS from the manufacturer's website
>>> (Asus?). Download and install the motherboard drivers from the
>>> website while you're at it. You're looking for new VIA X-in-1
>>> drivers for the motherboard, which will update the USB firmware,
>>> Northbridge and a few other bits and pieces.
>>>
>>>
>>> As to why you don't see this with Linux?
>>>
>>> Linux does tend to run CPUs without giving them quite as extensive a
>>> workout. *shrugs*.
>>
>> A nice way of saying that Linux, does more, with less CPU power.
>
> Not necessarily. It may well do the same or less with less CPU power.
> Unless you've made exact measurements of the same system running the
> same software with the same hardware and the same APIs. Which, of
> course, is impossible.
Doesn't track.
Either Linux is less CPU hungry in the case above, or it's doing less.
Since it's doing what the poster wanted, and presumably, so was Windows,
then Linux is either more efficient wrt CPU/MEM usage, or it's better
at doing what the poster wanted, without straining the CPU and/or
Memory.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)
iD8DBQE/yRcwd90bcYOAWPYRAmsaAKCXXHziYVPfJlWD4abYZ6bjNdx+kQCcDMkJ
hyXJ4+E7qejUK8M/KfF1t/U=
=hkEn
-----END PGP SIGNATURE-----
--
Jim Richardson http://www.eskimo.com/~warlock
I wasn't born Republican, Democrat, or yesterday.
Jeff Relf
" Linux is just fast and efficient ,
so it sheds the dross "
But why is Moz so Bloated ?
Even the stripped down firebird .7+
takes longer to load than IE6 .
Plus : Although I like firebird ,
I have to use IE6 when I do FTP uploads .
( I don't use an FTP client because IE6 is so simple ,
just one click and I'm ready to do FTP uploads via
my favorite method : Drag and drop )
John Bailo
> Even the stripped down firebird .7+
> takes longer to load than IE6 .
See, you just confirmed my suspicions...
I said the only reason that you were
interested in Firebird was so that you
could troll and lie about it in Cola.
And so, once again, I evaulate the Boolean
value named Relf, and the value comes back --- false.
Jeff Relf
was so that you could troll and lie about it in Cola "
This is my very first day in cola .
And I like Firebird ... Thanks for the suggestion .
Rick
> Hi John Bailo , You say :
> " Linux is just fast and efficient ,
> so it sheds the dross "
>
> But why is Moz so Bloated ?
Why can't you write sentences like a normal human?
>
> Even the stripped down firebird .7+
> takes longer to load than IE6 .
You are using the tyical window$ lie again. You fail to mention that a
large part of IE loads when window$ loads.
>
> Plus : Although I like firebird ,
> I have to use IE6 when I do FTP uploads .
Since I don't use firbird, I don't know if it has ftp capabilities or not.
But then I don't use Mozilla to ftp, I use an ftp client.
>
> ( I don't use an FTP client because IE6 is so simple ,
> just one click and I'm ready to do FTP uploads via
> my favorite method : Drag and drop )
Yeah. drag and drop is so much easier than click, click.
--
Rick
Mark Gary
I don't think you understand how windows works. IE will fire up quickly,
because all the DLL's are already preloaded during bootup. That is
because IE represents about 60 percent (perhaps more) of the over all
Windows operating system (well the GUI part anyway), and strangely enough,
is also represents why windows is so insecure.
Mozilla (which does incidently fire up quite quickly) needs to load up all
its DLL files seperatly.
Mozila can be as bloated or not bloated as you see fit. Its open source,
and therefore is configerable by anybody.
--
Mark
Registered Linux User #329755 - http://counter.li.org
If you must email me then address to :
uk.co.demon.mwgary.nospam@mark (just reverse it and remove nospam)
Kelsey Bjarnason
> Hi John Bailo , You say :
> " Linux is just fast and efficient ,
> so it sheds the dross "
>
> But why is Moz so Bloated ?
>
> Even the stripped down firebird .7+
> takes longer to load than IE6 .
You, of course, already know the answer to this and are simply trolling.
For the benefit, however, of those who might actually buy this as meaning
something, the answer is simple.
IE consists of many components, most of which are loaded by the OS at boot
time, since it uses them itself to render various display elements and so
on. The actual "browser" part is tiny.
With Moz, the whole application gets loaded when you start it, not just
one tiny little portion of it.
Thus it is almost impossible to compare actual loadtimes of the two, as
you can't very well get a real load time value for IE.
It's about like asking why one application loads so much faster than
another, when the first one is kept on a hard disk and the other gets
loaded off your old CD drive. It has nothing to do with the application,
nor with the OS's efficiency; it's simply a question of the speed of the
media the application is kept in.
Tom Shelton
> On Sun, 30 Nov 2003 00:42:07 -0800, Jeff Relf wrote:
>
>> Hi John Bailo , You say :
>> " Linux is just fast and efficient ,
>> so it sheds the dross "
>>
>> But why is Moz so Bloated ?
>>
>> Even the stripped down firebird .7+
>> takes longer to load than IE6 .
>>
>> Plus : Although I like firebird ,
>> I have to use IE6 when I do FTP uploads .
>>
>> ( I don't use an FTP client because IE6 is so simple ,
>> just one click and I'm ready to do FTP uploads via
>> my favorite method : Drag and drop )
>
> I don't think you understand how windows works. IE will fire up quickly,
> because all the DLL's are already preloaded during bootup. That is
> because IE represents about 60 percent (perhaps more) of the over all
> Windows operating system (well the GUI part anyway), and strangely enough,
> is also represents why windows is so insecure.
>
> Mozilla (which does incidently fire up quite quickly) needs to load up all
> its DLL files seperatly.
>
> Mozila can be as bloated or not bloated as you see fit. Its open source,
> and therefore is configerable by anybody.
>
Please show a ref to where you got this info. A quick google on the
subject seems to indicate otherwise. In fact, I found many comments
that suggested that IE starts faster then Moz even when running under
wine - which if true certainly indicates that there is in fact no
preloading. Preloading seems very suspicious to me - especially on NT
systems, where process are given their own address space and there is
very little sharing.
The problem is more likely that the Moz developers really don't
understand the windows dynamic loading mechanism and are not rebasing
their dll's properly (if at all) - causing windows to constantly relocate
them on startup.
--
Tom Shelton
Mark Gary
I over simplified. Sorry.
Where NT4 is concerned, the rot set in after service pack 4, and IE4, when
many parts of the OS were replaced for socalled enhancement. Most
markedly the old windows explorer file manager got replaced (which
includes the task bar), and Winat got replaced with task scheduler, and
also the introduction of active desktop.
The situating now is this. The desktop, the file manager, task bar,
start menu, browser and other components are all IE, and get updated
whenever a new version appears. Infact the browser part is a fiction, its
only windows explorer in a different guise, and in fact is not needed, as
you can browse the web quite easily from windows explorer, its just people
want an icon thats labeled "Internet explorer" because they get confused
otherwise. whithout the DLLs firing up at boot time, none of those parts
would work at all.
> The problem is more likely that the Moz developers really don't
> understand the windows dynamic loading mechanism and are not rebasing
> their dll's properly (if at all) - causing windows to constantly
> relocate them on startup.
Most likely correct. But its also many components loading up at once, but
once loaded, each part will fire up quickly.
Milo T.
GCC doesn't have rebase or bind tools... :)
Tom Shelton
> On Sun, 30 Nov 2003 18:40:39 +0000, Tom Shelton wrote:
>
>> On 2003-11-30, Mark Gary <ma...@nospam.co.uk> wrote:
>>> On Sun, 30 Nov 2003 00:42:07 -0800, Jeff Relf wrote:
>>>
>>>> Hi John Bailo , You say :
>>>> " Linux is just fast and efficient ,
>>>> so it sheds the dross "
>>>>
>>>> But why is Moz so Bloated ?
>>>
<snip>
>> Please show a ref to where you got this info. A quick google on the
>> subject seems to indicate otherwise. In fact, I found many comments
>> that suggested that IE starts faster then Moz even when running under
>> wine - which if true certainly indicates that there is in fact no
>> preloading. Preloading seems very suspicious to me - especially on NT
>> systems, where process are given their own address space and there is
>> very little sharing.
>
> I over simplified. Sorry.
>
> Where NT4 is concerned, the rot set in after service pack 4, and IE4, when
> many parts of the OS were replaced for socalled enhancement. Most
> markedly the old windows explorer file manager got replaced (which
> includes the task bar), and Winat got replaced with task scheduler, and
> also the introduction of active desktop.
>
Hmmm, that new Task Scheduler in my opinion is a vast improvement over
the old at service. In fact, I have used it quite extensively and it is
a much more powerfull and flexible system - and it's not like the at
command disappeared. As for active desktop - I have never once turned
it on :)
> The situating now is this. The desktop, the file manager, task bar,
> start menu, browser and other components are all IE, and get updated
> whenever a new version appears. Infact the browser part is a fiction, its
> only windows explorer in a different guise, and in fact is not needed, as
> you can browse the web quite easily from windows explorer, its just people
> want an icon thats labeled "Internet explorer" because they get confused
> otherwise. whithout the DLLs firing up at boot time, none of those parts
> would work at all.
>
>
Yes, you can browse the web from windows explorer - but that's because
it instantiates a instance of IE. You realize that IE is really a COM
object that lives in shdocvw.dll? All iexplorer.exe is a gui around
shdocvw.dll.
And, if what I've read is true - that IE starts faster on linux under
wine then Moz - it would appear to me that preloading has nothing to do
with it. As, for the rest - I'm going to have to research this
because it doesn't jive with the way NT handles process management.
--
Tom Shelton
Tom Shelton
> On Sun, 30 Nov 2003 18:40:39 GMT, Tom Shelton wrote:
>
>> On 2003-11-30, Mark Gary <ma...@nospam.co.uk> wrote:
>>> On Sun, 30 Nov 2003 00:42:07 -0800, Jeff Relf wrote:
>>>
>>>> Hi John Bailo , You say :
>>>> " Linux is just fast and efficient ,
>>>> so it sheds the dross "
>>>>
>>>> But why is Moz so Bloated ?
>>>>
<snip>
>> The problem is more likely that the Moz developers really don't
>> understand the windows dynamic loading mechanism and are not rebasing
>> their dll's properly (if at all) - causing windows to constantly relocate
>> them on startup.
>
> GCC doesn't have rebase or bind tools... :)
Well that would explain a lot now wouldn't it :) Do you know anything
about this IE preload thing... Seems a little suspicious to me. I see
it so often claimed - but I see no real evidence for it. I'm not saying
it isn't true or anything - but it just doesn't seem right.
To be completly honest, I'm not sure what the fuss is over... So, maybe
IE starts faster - that doesn't mean it is necessarily better. It is kind of
amusing though to see people get all up in arms when IE maybe better at
something then Mozilla ;)
--
Tom Shelton
Jim Richardson
Hash: SHA1
On Mon, 01 Dec 2003 04:37:25 GMT,
Tom Shelton <t...@mtogden.com> wrote:
<snip>
> Hmmm, that new Task Scheduler in my opinion is a vast improvement over
> the old at service. In fact, I have used it quite extensively and it is
> a much more powerfull and flexible system - and it's not like the at
> command disappeared. As for active desktop - I have never once turned
> it on :)
>
I found out an amusing limitation of AD a few days/weeks ago.
Seems that once you have "named" a machine on AD, that machine name,
can't be changed without reinstalling the whole freakin' OS...
Unreal.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)
iD8DBQE/ytPzd90bcYOAWPYRAiMkAJ9oSRLEUoHxAoB7G4fRvYGH7sG6MwCfRxCA
PXkxP6kfdeUXhChv29iSkN4=
=TFUX
-----END PGP SIGNATURE-----
--
Jim Richardson http://www.eskimo.com/~warlock
Microsoft gives you Windows... Linux gives you the whole house
Tom Shelton
>
> On Mon, 01 Dec 2003 04:37:25 GMT,
> Tom Shelton <t...@mtogden.com> wrote:
>
><snip>
>
>> Hmmm, that new Task Scheduler in my opinion is a vast improvement over
>> the old at service. In fact, I have used it quite extensively and it is
>> a much more powerfull and flexible system - and it's not like the at
>> command disappeared. As for active desktop - I have never once turned
>> it on :)
>>
>
> I found out an amusing limitation of AD a few days/weeks ago.
>
> Seems that once you have "named" a machine on AD, that machine name,
> can't be changed without reinstalling the whole freakin' OS...
>
> Unreal.
>
>
>
Do you mean Active Directory or Active Desktop? I've never had occasion
to use the first, and I've never wanted to use the second :)
--
Tom Shelton
Jeff Relf
" The problem is more likely that the Moz developers
really don't understand
the windows dynamic loading mechanism
causing windows to constantly relocate them on startup "
Not rebasing their dll's ... Interesting , Thanks .
Michael Vester
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> On Mon, 01 Dec 2003 04:37:25 GMT,
> Tom Shelton <t...@mtogden.com> wrote:
>
> <snip>
>
> > Hmmm, that new Task Scheduler in my opinion is a vast improvement
> > over
> > the old at service. In fact, I have used it quite extensively and
> > it is a much more powerfull and flexible system - and it's not like
> > the at
> > command disappeared. As for active desktop - I have never once
> > turned it on :)
> >
>
> I found out an amusing limitation of AD a few days/weeks ago.
>
> Seems that once you have "named" a machine on AD, that machine name,
> can't be changed without reinstalling the whole freakin' OS...
>
> Unreal.
>
They still haven't fixed that? I remember that a name change in NT
domain means a reinstall.
--
11:15pm up 6 days, 7:12, 1 user, load average: 1.06, 1.08, 1.04
To email me, change .com to .ca Linux Counter Registration #126647
John Bailo
>
> I found out an amusing limitation of AD a few days/weeks ago.
>
> Seems that once you have "named" a machine on AD, that machine name,
> can't be changed without reinstalling the whole freakin' OS...
I seem to remember that was a characteristic
of a WinNT machine set up as a PDC.
Guess they just moved that code out and productized
it as AD.
Jeff Relf
" GCC doesn't have rebase or bind tools... :) "
Why does Firebird even need DLLs ?
It's not like it's actually sharing anything .
Jeff Relf
You realize that IE is really a COM object
that lives in shdocvw.dll ?
All iexplorer.exe is a gui around shdocvw.dll .
it would appear to me that preloading
has nothing to do with it "
That sounds right to me .
Which brings me back to my original question
" Why is Moz so bloated ? "
My guess is that it's borrowing too much unused code .
Milo T.
> On 2003-11-30, Milo T. <fanta...@malaprop.net> wrote:
>> On Sun, 30 Nov 2003 18:40:39 GMT, Tom Shelton wrote:
>>
>>> On 2003-11-30, Mark Gary <ma...@nospam.co.uk> wrote:
>>>> On Sun, 30 Nov 2003 00:42:07 -0800, Jeff Relf wrote:
>>>>
>>>>> Hi John Bailo , You say :
>>>>> " Linux is just fast and efficient ,
>>>>> so it sheds the dross "
>>>>>
>>>>> But why is Moz so Bloated ?
>>>>>
>
> <snip>
>
>>> The problem is more likely that the Moz developers really don't
>>> understand the windows dynamic loading mechanism and are not rebasing
>>> their dll's properly (if at all) - causing windows to constantly relocate
>>> them on startup.
>>
>> GCC doesn't have rebase or bind tools... :)
>
> Well that would explain a lot now wouldn't it :) Do you know anything
> about this IE preload thing... Seems a little suspicious to me. I see
> it so often claimed - but I see no real evidence for it. I'm not saying
> it isn't true or anything - but it just doesn't seem right.
Well, shdocvw.dll is used by the shell, so theoretically it is already in
memory, which should trim off some of the DLL load time. But DLL load time
is nearly inconsequential compared to DLL entrypoint relocation time -
which is the REAL big killer for Mozilla.
I've run my own tests and have been able to get Firebird to load in the
same amount of time as IE on my system. No way of timing it of course, but
there you go.
Also, if an application is installed using MSI (possibly installed in other
ways as well?), it should automatically get boosted in speed over time as
DLL loads get cached. But again, that's still useless noise in the data if
they're not using fast DLL loading.
Open Office suffers from the same problems - but it's nearly impossible to
fix because they have so many different DLLs and components loading, so
without prior knowledge of the internals (and the willingness to spend a
few weeks learning the app), an outsider can't optimize the DLL load.
> To be completly honest, I'm not sure what the fuss is over... So, maybe
> IE starts faster - that doesn't mean it is necessarily better. It is kind of
> amusing though to see people get all up in arms when IE maybe better at
> something then Mozilla ;)
It's this whole "Microsoft uses secret APIs" thing that people go nuts
over... I mean, either Microsoft writes really incredibly sophisticated
fast code, better than anyone else possibly could and only lets their app
developers use it, or they're incompetent buffoons who write slow, bloated
code. Sometimes, the same people make both claims.
It's rather amusing really.
Milo T.
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> On Mon, 01 Dec 2003 04:37:25 GMT,
> Tom Shelton <t...@mtogden.com> wrote:
>
> <snip>
>
>> Hmmm, that new Task Scheduler in my opinion is a vast improvement over
>> the old at service. In fact, I have used it quite extensively and it is
>> a much more powerfull and flexible system - and it's not like the at
>> command disappeared. As for active desktop - I have never once turned
>> it on :)
>>
>
> I found out an amusing limitation of AD a few days/weeks ago.
>
> Seems that once you have "named" a machine on AD, that machine name,
> can't be changed without reinstalling the whole freakin' OS...
Rebooting I can believe - in fact I've seen that mentioned here, though I
couldn't really talk about it because I've never used AD. Reinstalling I
can't. References?
Milo T.
They're handy for other reasons. You can split your app up along functional
lines, and update only the pieces you need to if you need to release a
patch.
John Bailo
> that IE starts faster on linux under wine then Moz -
That is interesting.
I was experimenting with a VNC ( term serv for Linux ) and
ran Moz from a remote Xwindows session. It ran (it seemed)
about 5 times faster than when run from GNOME on the Linux
machine that was hosting the VNC session !
Is it the WM support that causes Moz the drag ?(?)
All i know is that Moz running fast can't be beat.
It's the best browser in the world.
John Bailo
> It's this whole "Microsoft uses secret APIs" thing that people go nuts
It's no secret. It's very well known inside and outside The Company.
> developers use it, or they're incompetent buffoons who write slow, bloated
You forget what I've told you time and time again.
For m$ to make money they have to 'create functionality'.
If Word can do what Excel does and both can do what Access does
since they are all just shells around the core COM components
(as Jeff Relf pointed out about IE ) then what would be left
to be part of the Office Suite?
You could just have a document witha lot of embeddable formats and
tools...instead of bloated MS Office apps.
Oh, wait, that IAYF...
Jeff Relf
" You can split your app up along functional lines ,
and update only the pieces you need to
If Firebird were small enough ,
it could just patch with a new EXE rather than a DLL .
Milo T.
> On Sat, 29 Nov 2003 20:40:48 GMT,
> Milo T. <fanta...@malaprop.net> wrote:
>> Actually, I've had problems with heat pipe systems overheating unless I
>> smear an entire layer of goop over *everything*. Of course, they had a
>> "protective copper shim" on top of the CPU die to stop the die from
>> cracking against the heat sink. *shrugs*
>>
>
> if you are talking about smearing the grease along the pipes to help
> conduction from the cpu section, to the heatsink, depending on the
> design of the heat pipe, this may actually *reduce* the effective heat
> transfer.
Nah. The heatsink feeds into the heat pipes, which feed out to a fan for
convection cooling.
+-----------------------+
| Heat Pipe Section |
+-----------------------+
+-----------------------+
| Heat Sink Surface |
+-----------------------+
+-----------------------+
| Copper Shim |
+-----------------------+
+-+ +---------+
+-+-+--+---------+------+
| CPU Die |
+-----------------------+
Note the raised surfaces which are the die itself, protected by the copper
shim. The thermal grease fills in the gaps which do not touch the surface
of the heat sink.
[snip]
>> Also note that the grease, being non-conductive, can spread the heat over a
>> wider area than the heatsink.
>>
>
> Actually, many of the thermal compounds used now, have a high copper or
> silver content, so they aren't as non-conductive (electrically, which is
> what I assumed you meant) as you might think.
I'd hope they were non-conductive, given that they're touching the top of
the die. Those dies appear to be open to the air and not sealed. At the
very least, I'd be worrying about shorts.
>> Not necessarily. It may well do the same or less with less CPU power.
>> Unless you've made exact measurements of the same system running the
>> same software with the same hardware and the same APIs. Which, of
>> course, is impossible.
>
> Doesn't track.
> Either Linux is less CPU hungry in the case above, or it's doing less.
> Since it's doing what the poster wanted, and presumably, so was Windows,
> then Linux is either more efficient wrt CPU/MEM usage, or it's better
> at doing what the poster wanted, without straining the CPU and/or
> Memory.
Two different tasks, two different OSes, two different architectures, using
two different sets of parameters for the bus, PCI bus, drivers, etc. Linux
could either be doing less, or perhaps uses higher tolerance settings for
the chipset.
John Bailo
There is also Mozilla and Thunderbird. They might share code.
But if you're that curious -- download the source and see for yourself.
That's the beatuty of OSS -- you can always know for sure.
As opposed to making random guesses about what goes on
inside the Silicon Curtain.
Tom Shelton
> On Mon, 01 Dec 2003 04:43:09 GMT, Tom Shelton wrote:
>
>> On 2003-11-30, Milo T. <fanta...@malaprop.net> wrote:
>>> On Sun, 30 Nov 2003 18:40:39 GMT, Tom Shelton wrote:
>>>
>>>> On 2003-11-30, Mark Gary <ma...@nospam.co.uk> wrote:
>>>>> On Sun, 30 Nov 2003 00:42:07 -0800, Jeff Relf wrote:
>>>>>
>>>>>> Hi John Bailo , You say :
>>>>>> " Linux is just fast and efficient ,
>>>>>> so it sheds the dross "
>>>>>>
>>>>>> But why is Moz so Bloated ?
>>>>>>
>>
>> <snip>
>>
>>>> The problem is more likely that the Moz developers really don't
>>>> understand the windows dynamic loading mechanism and are not rebasing
>>>> their dll's properly (if at all) - causing windows to constantly relocate
>>>> them on startup.
>>>
>>> GCC doesn't have rebase or bind tools... :)
>>
>> Well that would explain a lot now wouldn't it :) Do you know anything
>> about this IE preload thing... Seems a little suspicious to me. I see
>> it so often claimed - but I see no real evidence for it. I'm not saying
>> it isn't true or anything - but it just doesn't seem right.
>
> Well, shdocvw.dll is used by the shell, so theoretically it is already in
> memory, which should trim off some of the DLL load time. But DLL load time
> is nearly inconsequential compared to DLL entrypoint relocation time -
> which is the REAL big killer for Mozilla.
>
Right... That does make sense. And is more in line what I was
thinking - preloading is not the issue.
> I've run my own tests and have been able to get Firebird to load in the
> same amount of time as IE on my system. No way of timing it of course, but
> there you go.
>
Ok. That sounds very reasonable.
> Also, if an application is installed using MSI (possibly installed in other
> ways as well?), it should automatically get boosted in speed over time as
> DLL loads get cached. But again, that's still useless noise in the data if
> they're not using fast DLL loading.
>
I didn't know that... So, the dll mapping gets cached? That's kind of
cool.
> Open Office suffers from the same problems - but it's nearly impossible to
> fix because they have so many different DLLs and components loading, so
> without prior knowledge of the internals (and the willingness to spend a
> few weeks learning the app), an outsider can't optimize the DLL load.
>
I haven't run OO on Windows for a while, I can't even remember what
version it was, but I do remember it was a real pig at start up.
Personally, I think OO is nice on Linux - but it has some real performance
issues when compared to Office2k/XP/2K3 (all of which I have liked very
well) - mostly in the repaint area. I get a lot of artifacts from
dialogs and such with OO. But, overal 1.1 is a significant improvement
over 1.0.
--
Tom Shelton
John Bailo
> Hi Milo T. , You speak of DLL patches :
> " You can split your app up along functional lines ,
> and update only the pieces you need to
> if you need to release a patch a "
But because these DLLs are shared they might then break other functions (!)
That is the reason for .NET
Take any functionality which may be common and put them into a set of
assemblies. Then insulate code with interfaces to prevent changes from
crashing applications.
Good idea -- I'm glad mono is porting it to Linux.
Jeff Relf
If I were to code my own
HTTP , FTP , POP3 , SMTP , NNTP client ...
I'd use Thunderbird and firebird as guides .
I'd start over , using none of the old code .
I'd also use Microsoft's OLE and COM libraries .
Milo T.
John Bailo
> I'd also use Microsoft's OLE and COM libraries .
OLE - Old Libraries and Executables
COM - Crappy Olde Microsoft
Jeff Relf
" OLE - Old Libraries and Executables "
To hell with you and your cat .
Jim Richardson
Hash: SHA1
On Mon, 01 Dec 2003 06:56:02 GMT,
Yes, I know.
My point was, that your post seemed to indicate that you had thermal
grease over the pipes also, I was pointing out that that can, depending
on the design of the heatsink, actually *reduce* the cooling
effectiveness of the heat pipe system.
> [snip]
>
>>> Also note that the grease, being non-conductive, can spread the heat over a
>>> wider area than the heatsink.
>>>
>>
>> Actually, many of the thermal compounds used now, have a high copper or
>> silver content, so they aren't as non-conductive (electrically, which is
>> what I assumed you meant) as you might think.
>
> I'd hope they were non-conductive, given that they're touching the top of
> the die. Those dies appear to be open to the air and not sealed. At the
> very least, I'd be worrying about shorts.
>
I doubt severely that they are open. They are encapsulated for a number
of reasons, moisture control for one. IIRC, Si is hygroscopic under
heat, as it cools, it pulls moisture out of the air around it.
>>> Not necessarily. It may well do the same or less with less CPU power.
>>> Unless you've made exact measurements of the same system running the
>>> same software with the same hardware and the same APIs. Which, of
>>> course, is impossible.
>>
>> Doesn't track.
>> Either Linux is less CPU hungry in the case above, or it's doing less.
>> Since it's doing what the poster wanted, and presumably, so was Windows,
>> then Linux is either more efficient wrt CPU/MEM usage, or it's better
>> at doing what the poster wanted, without straining the CPU and/or
>> Memory.
>
> Two different tasks, two different OSes, two different architectures, using
> two different sets of parameters for the bus, PCI bus, drivers, etc. Linux
> could either be doing less, or perhaps uses higher tolerance settings for
> the chipset.
So it either uses less, or pushes the hardware further :)
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)
iD8DBQE/ywB9d90bcYOAWPYRAnBvAJ4zpMWrN6UgWbwOHd3Ra3Dydu4SOgCfRcGZ
aNNYUyU7uWWZc6Mxv1ja52A=
=6W1A
-----END PGP SIGNATURE-----
--
Jim Richardson http://www.eskimo.com/~warlock
I have an understanding with my local police--I have them outgunned, but
they have me outnumbered.
Jim Richardson
Hash: SHA1
aren't shared object files (.so) basically the same thing as dlls?
mozilla-firebird on Debian has about 85 of them, seems pretty split up
to me :)
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)
iD8DBQE/yv9Id90bcYOAWPYRAiNAAKDlhf3OXNIYjTIsWZH1OE99Fa3RIQCg3BWi
cOWeUqRVbHugwtJ+AxdE3xY=
=XR1v
-----END PGP SIGNATURE-----
--
Jim Richardson http://www.eskimo.com/~warlock
All life is a conjugation of the verb "to eat"
Jim Richardson
Hash: SHA1
On Mon, 01 Dec 2003 06:13:17 GMT,
Tom Shelton <t...@mtogden.com> wrote:
> On 2003-12-01, Jim Richardson <war...@eskimo.com> wrote:
>>
>> On Mon, 01 Dec 2003 04:37:25 GMT,
>> Tom Shelton <t...@mtogden.com> wrote:
>>
>><snip>
>>
>>> Hmmm, that new Task Scheduler in my opinion is a vast improvement over
>>> the old at service. In fact, I have used it quite extensively and it is
>>> a much more powerfull and flexible system - and it's not like the at
>>> command disappeared. As for active desktop - I have never once turned
>>> it on :)
>>>
>>
>> I found out an amusing limitation of AD a few days/weeks ago.
>>
>> Seems that once you have "named" a machine on AD, that machine name,
>> can't be changed without reinstalling the whole freakin' OS...
>>
>> Unreal.
>>
>>
>>
>
> Do you mean Active Directory or Active Desktop? I've never had occasion
> to use the first, and I've never wanted to use the second :)
>
Active directory, It's an exchange install on a W2K server system, for a
client. I don't have to deal with that system fortunately. But I do have
to protect it, and feed it the smtp stream. :) no one trusts MICROS~1
outside the firewall there.
Can't say I blame them.
Fortunately, they have an embedded Linux firewall (Dlink) Two Linux
web/email/other servers as a fallover system, with wide physical and
network seperation, and an internal Linux box acting as backup and
IDS/Honeynet, and if the fileserver burps again, I'll replace it with
Linux running samba.
It's good to be paid, to play with servers.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)
iD8DBQE/yvyrd90bcYOAWPYRAm/LAJ4vihz3hM7gcwh6EyEevM/OTcbOJwCffqVN
dLTTuBXLcT+d/qLor12Dx3Q=
=nyMv
-----END PGP SIGNATURE-----
--
Jim Richardson http://www.eskimo.com/~warlock
"If you choke a smurf, what color does it turn?"
Jim Richardson
Hash: SHA1
Just a comment from a friend of mine, who digs MICROS~1 and uses it for
everything. He even likes Active directory, but was a bit sheepish about
this.
But just for you, I googled for
active directory domain change reinstall (no quotes) got a bunch of
hits, (not surprising for such a broad search) and #1 was to
<http://www.networkcomputing.com/1303/1303centerfoldtext.html>
highlighting the problem.
But this one
<http://www.analogduck.com/blog/modules/news/article.php?storyid=295>
was quite interesting. Verifying the problem with W2K, but saying that
W2K3 has added tools to make it possible, if a bit hairy.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)
iD8DBQE/yv6Sd90bcYOAWPYRAgo6AJ4jNeDwdKysLI+D1yHi71scbXPX9ACghHeP
xAbIkP40Pruo7LfSGrYWl40=
=FaBg
-----END PGP SIGNATURE-----
--
Jim Richardson http://www.eskimo.com/~warlock
E pluribus Linux
John Bailo
Foret about all that.
I have a pressing need to take a CLIT test.
I'm willing to pay a qualified instructor top dollar
if she's heigh weight proportional and under 5'5".
Erik Funkenbusch
>>> I found out an amusing limitation of AD a few days/weeks ago.
>>>
>>> Seems that once you have "named" a machine on AD, that machine name,
>>> can't be changed without reinstalling the whole freakin' OS...
>>
>> Rebooting I can believe - in fact I've seen that mentioned here,
>> though I couldn't really talk about it because I've never used AD.
>> Reinstalling I can't. References?
>
> Just a comment from a friend of mine, who digs MICROS~1 and uses it for
> everything. He even likes Active directory, but was a bit sheepish about
> this.
>
> But just for you, I googled for
>
> active directory domain change reinstall (no quotes) got a bunch of
> hits, (not surprising for such a broad search) and #1 was to
>
> <http://www.networkcomputing.com/1303/1303centerfoldtext.html>
>
> highlighting the problem.
That's not the problem you described. The link talks about how AD doesn't
let you chaing the root domain name without reinstalling. There is no need
to reinstall a machine just because you've changed it's machine name.
> But this one
>
> <http://www.analogduck.com/blog/modules/news/article.php?storyid=295>
> was quite interesting. Verifying the problem with W2K, but saying that
> W2K3 has added tools to make it possible, if a bit hairy.
Also not the problem you describe.
Jim Richardson
Hash: SHA1
More a matter of me not being too hip on the terminology MICROS~1
uses for AD.
Still, it's pretty lame, to have to reinstall an OS, over such a
relatively minor change...
Aparantly you *can* change it, in the same context that you *can* eat
ground glass and survive, if you're lucky.
Mind you, I haven't installed LDAP on Linux (yet) ironically enough, I
have a server pair I will be doing that to in a few weeks. I'll let you
know how it goes. Can't compare it to AD, since I don't use MICROS~1 if
I can possibly avoid it. But it'll be interesting none-the-less.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)
iD8DBQE/y3Zed90bcYOAWPYRAlf8AKC5cwIKYfasHanqRWyuM7f67KeeDgCgpb4N
XfNeuJAoTbUrHxx2PWtuw1c=
=Z7Wd
-----END PGP SIGNATURE-----
--
Jim Richardson http://www.eskimo.com/~warlock
Reality continues to ruin my life.
-- Calvin
Bob Hauck
> GCC doesn't have rebase or bind tools... :)
There's a separate tool available for ELF executables like Linux uses.
The tool is called "prelink". I don't know if there is an equivalent
for PE files.
--
-| Bob Hauck
-| To Whom You Are Speaking
-| http://www.haucks.org/
John Bailo
> On Sun, 30 Nov 2003 21:29:43 GMT, Milo T. <fanta...@malaprop.net> wrote:
>
>> GCC doesn't have rebase or bind tools... :)
>
> There's a separate tool available for ELF executables like Linux uses.
> The tool is called "prelink". I don't know if there is an equivalent
> for PE files.
For all newbies reading this thread.
Moz is not bloated.
MS Office is.
Okay, go back to your fun.
Linux Sucks
Rick <ri...@none.com> wrote:
>On Sun, 30 Nov 2003 00:42:07 -0800, Jeff Relf wrote:
>
>> Hi John Bailo , You say :
>> " Linux is just fast and efficient ,
>> so it sheds the dross "
>>
>> But why is Moz so Bloated ?
>
>Why can't you write sentences like a normal human?
>
Why can't you accept people for what they do, because it is their
choice?
Stupid Linux losers, want people to post to their taste too.
Linux Sucks
Mark Gary <ma...@nospam.co.uk> wrote:
>On Sun, 30 Nov 2003 00:42:07 -0800, Jeff Relf wrote:
>
>> Hi John Bailo , You say :
>> " Linux is just fast and efficient ,
>> so it sheds the dross "
>>
>> But why is Moz so Bloated ?
>>
>> Even the stripped down firebird .7+
>> takes longer to load than IE6 .
>>
>> Plus : Although I like firebird ,
>> I have to use IE6 when I do FTP uploads .
>>
>> ( I don't use an FTP client because IE6 is so simple ,
>> just one click and I'm ready to do FTP uploads via
>> my favorite method : Drag and drop )
>
>I don't think you understand how windows works.
It seems that you are the one who doesn't understand how Windows works.
>IE will fire up quickly,
>because all the DLL's are already preloaded during bootup. That is
Bullshit.
Only essential system dlls are loaded, you IDIOT.
>because IE represents about 60 percent (perhaps more) of the over all
>Windows operating system (well the GUI part anyway), and strangely enough,
>is also represents why windows is so insecure.
According to Linux losers, but you keep forgetting no one outside cola
or slash dot takes you morons seriously.
>Mozilla (which does incidently fire up quite quickly) needs to load up all
>its DLL files seperatly.
Bullshit again.
>Mozila can be as bloated or not bloated as you see fit. Its open source,
>and therefore is configerable by anybody.
What a technically incompetent idiot you are.
No wonder you use Linux.
Jeff Relf
" Stupid Linux losers ,
Nothing says " Moron " like :
" Would you please use only My style when you write "
Pete Goodwin
> Great -- now trot off to the Gate$ mansion
> to tell Bill and Steve and
> to get your fees and the new autographed
> pictures of Michael Jackson waiting for you
> in the Pete Townsend memorial shrine.
Nah, I'll let you do that, you're the one obsessed with them.
--
Pete Goodwin, using XanaNews 1.15.8.2 on Windows 2000.
"Nothing is permanent except change."
The Ghost In The Machine
<s...@sletom002.com>
wrote
on Sat, 29 Nov 2003 13:33:13 -0600
<slrnbsht...@home.harry.net>:
> Linųnut blubbered effusively on Sat, 29 Nov 2003 at 15:43 GMT:
>
>> Fearing a spontaneous XP reboot, Erik Funkenbusch mumbled this incantation:
>>
>>> Haven't done a search on google regarding via and unstable, have you?
>>
>> I got 45 results for funkenbusch+unstable. Including this little gem:
>
> Darn. I was gonna do a whole thread in his honor:
>
> windows unstable 422,000
One wonders how many of these hits are for Microsoft Windows, and
how many are for, say, "unstable windows" during an earthquake.
Of course, I suspect far more of the former than the latter... :-)
> microsoft unstable 181,000
> me instable 1,140,000
> 95 unstable 603,000
> 98 unstable 511,000
> 2000 unstable 2,040,000
> xp unstable 118,000
> via unstable 901,000
> chocolate unstable 35,200
An interesting concept, that.
> nuclear unstable 273,000
>
> Windoze wins again! It even beats out "nuclear" which should provide
> some clues to us all!
>
Java has a clause in its license that precludes its use for
nuclear control systems. I wonder if Microsoft has provided
for such.... :-)
3. RESTRICTIONS. Software is confidential and
copyrighted. Title to Software and all associated
intellectual property rights is retained by Sun and/or
its licensors. Unless enforcement is prohibited
by applicable law, you may not modify, decompile,
or reverse engineer Software. You acknowledge
that Licensed Software is not designed or intended
for use in the design, construction, operation or
maintenance of any nuclear facility. Sun Microsystems,
Inc. disclaims any express or implied warranty of
fitness for such uses. No right, title or interest
in or to any trademark, service mark, logo or trade
name of Sun or its licensors is granted under this
Agreement. Additional restrictions for developers
and/or publishers licenses are set forth in the
Supplemental License Terms.
Personally, I wonder why one can't use Java for the design or
simulation of a nuclear facility. Construction and operation
might be issues in light of the garbage collection; Java's not
QNX or other such real-time operating systems, after all.
Maintenance? An interesting question.
But that's the license; go fig... :-)
--
#191, ewi...@earthlink.net
It's still legal to go .sigless.
Sinister Midget
> In comp.os.linux.advocacy, Sinister Midget
><s...@sletom002.com>
> wrote
> on Sat, 29 Nov 2003 13:33:13 -0600
><slrnbsht...@home.harry.net>:
>> nuclear unstable 273,000
>>
>> Windoze wins again! It even beats out "nuclear" which should provide
>> some clues to us all!
>>
>
> Java has a clause in its license that precludes its use for
> nuclear control systems. I wonder if Microsoft has provided
> for such.... :-)
More likely they have a FUD page that claims it's illegal to use
anything other than WinDoze in a nuclear facility, complete with
pictures of the sticker that shows how to tell if your machine has been
illegally changed from using the "competitively-priced" MICROS~1
software, supported by a very socially responsible corporation that has
several ten$ of billion$ to stand behind its products, to something
that's illegal because its name isn't on the sticker, especially that
free stuff that was put together by a rag-tag bunch of drug-induced
teenagers in their parents' basements.
Micro-Soft: Standing behind _you_!*
* Hope you're greased.
--
It is easier to fix Linux than it is to live with XP.
jeremyn
Many dlls that IE is dependant on are loaded during system startup. But
the figures aren't 60%, and IE _is not_ the sole reason that Windows is
insecure. Seeing as most parts of IE are integrated into Windows they
pretty much are essential.