I would like to say stay away from Dell's DRAC if possible. I've worked with DRAC III, IV, and V and they all *suck* compared to ILO 1/RILOE/RILOE II/ILO2
Then how the fscking much can they suck? Like a trillion blackholes?
I've use ILO 1/RILOE/RILOE II/ILO2 and they all suck donkey nuts with their complete java FAILURE.
Numerous times the ACTUAL SERVER has had to be power-cycled because the damn remote access card had crapped all over itself.
The problem is not someone talking to cell phone or to someone in person in the airplane. Or in a bus or train or wherever (excluding, well, churches, libraries etc.) The problem is people talking too loudly. I don't know why people often shout at their cell phones (like they wouldn't believe in RF transmission), but they often seem to speak unnecessary loudly at phone. It is possible to speak so quietly to the person sitting next to you that the person on the other side doesn't hear it. It is certainly possible to talk as quietly at cell phone.
It doesn't make sense to ban cell phones just because people sometimes shout at them, that would be like banning neighbor's TV because one can list set the volume too high. To fix the actual problem, people should be more polite and consider others, too. (Of course, that may be unrealistic to ask.)
Where did he say that? Last I heard he used
3.2 in (at least one of) his machines because of
sb-optimal 2.4 vm...
Also, the ac-series is different from Linus' tree; the diff between 2.4.4ac17 and 2.4.5 is
many megs. (Incidentally 2.4.2 OOM rambo killed my diff process when cache filled up (230 megs cache), so I can't give you exact figures) Alan includes much more experimental stuff, but also has some other differences in his tree. Merging in between does happen piece by piece.
ISTR that recent make version should do the right thing with 'make -j3' as well, but now that I search for it I can't find it anywhere. Judging from the >40% performance boost, they pretty much maxed the dual boxes as well.
However, I'd be more conserned with the 2.4.4ac9/2.4.4 difference in the benchmark. I think the Athlon benchmark should've been rerun on 2.4.4ac9. The -ac kernels (and recent -2.4.5pre's I think) incorporate such optimizations and VM changes that I think the benchmark numbers can vary just because of that. Particularly on these kind of SMP tests that stress the scheduler, VM, fs and cache system -- and not just the CPU. Of course, there might not be a significant difference -- propably isn't -- but now we'll never know.
I think it was just a gwbasic proggie...
10 PLAY "edabd...."
etc.
I think I have it somewhere. I also think there's a program to play those gwbasic PLAY "" things on linux...
Cygwin doesn't slow anything down, when you don't use it. Even when you do, it doesn't. Sometimes some programs decide to hog 100% cpu time, which of course brings Windows to its knees - but this is a weakness in the NT scheduler. (Try DOS edit for an example.) Cygwin programs rarely do that, however.
I use zsh instead of bash (I absolutely love it), and I have that "zsh prompt here" in Exploder context menu.
I think I couldn't live without cygwin anymore. Give it a shot before you judge, you might like it (especially if you have unix background). The installallation is a breeze: just go to http://sources.redhat.com/cygwin and click install (I love the installer.)
4NT does file-name completion better then any shell I have seen. Everytime I hit tab, it completes the NEXT filename to match. Shift-Tab does the previous match.
Of course zsh does that!
setopt automenu
(I think bash does as well.) Believe me: zsh (and bash) is much more powerful than 4NT. I use both on NT (zsh under cygwin). When I moved from DOS/4dos to unix I just couldn't believe what zsh was capable of.
Jeff Dike, the author on User-mode Linux has been actively seraching for people to undertake the Win32 port of UML. (UML is a cool hack that makes it possible to run Linux on top of the Unix system call API, eg. as a userlevel process in linux.) I think someone already started porting. They are using cygwin (at least in the beginning).
If they are succesful, one can execute linux binaries on Win32.
Of all (GUI) browsers I use (IE 5.5sp1/NT4SP6, Opera 5/NT4SP6, Latest Mozilla nightly/NT4SP6, Netscape 4.76/NT4SP6, Latest Mozilla nightly / Redhat7.0, Konqueror-2.01/ Redhat7.0, Opera 4 beta3/ Redhat7.0) or have used (IE 5.0, IE 4.0, Netscape 4.5, 4.0, 3.0 (NT/Linux), Opera 4, 3.6/NT)
Netscape 4.76 is ABSOLUTELY the most stable one. On both Linux and NT. It doesn't crash even once a month.
All the other browsers crash or hang (IE) every now and then - many of them don't do that annoyingly often, but more often than 4.76 anyhow.
Mozilla is perhaps the least stable (after Opera 4beta3 for linux) - but is definetely improving.
And the 4.7x series have been stable for me for long time.
4.76 IS slow and memory hungry, but Mozilla is MUCH more so (and Mozilla has other problems like missing shortcuts and mouse disturbing problems).
Guys, run memory test for your computer. Check your processor fan. Don't overclock. And if you using Win9x... just forget it.
I've seen a NT4SP5 (latest at the time) SMP
machine BSOD right after the 49.7 day watermark. That was not the only time I kept it up that long, and BSOD only happened once. That was my heavy-duty desktop machine with lots of compilation and multimedia stuff, so the uptimes were not all that bad.
If I had the money I'd still be interested in an SGI that can create games that make that much money. Who cares if the games suck ass as long as you collect the bucks.
4.0 + nVidia 0.93 works fine for me too. But DPMS didn't work with 4.0 + xfree nvidia driver, and it does not work with nVidia 0.93 either. Also, DGA doesn't seem to work with Vmware. Those are the only things I miss from 3.3.x - performance is _much_ better (2d, nevermind 3d), and stability ain't much worse either. Does anyone know when/how dpms and dga will be supported?
IntelliSense could speed up developing if it actually worked. I can understand it's slow, but it's downright ridiculous I have to wait half a minute after typing "0." while IntelliSense tries to find class members for 0. It rarely finds any... Although it often doesn't find even if it should.
That said, similar tools are available for emacs and although they may not work any more correctly (C++ grammar is a PITA to parse), they are way faster and trimmer. I mean, 20MB browse info file and it takes 5 minutes to build it... come on!
After a few.c's, most of the include files are propably in the files cache anyway. At least my experience has been that compiling almost always nearly maxes out CPU (unless something stupid is happening like MSVC waiting for include files from a network disk.) Memory bandwidth is very likely big issue, though.
For the compiling speed, I was surprised to find how well a 550 Athlon did compiling linux kernel compared to Intel. I don't have any exact numbers (different.config's), but I'd say Athlon won't propably lose much.
Slashdot Mirror
I would like to say stay away from Dell's DRAC if possible. I've worked with DRAC III, IV, and V and they all *suck* compared to ILO 1/RILOE/RILOE II/ILO2
Then how the fscking much can they suck? Like a trillion blackholes?
I've use ILO 1/RILOE/RILOE II/ILO2 and they all suck donkey nuts with their complete java FAILURE. Numerous times the ACTUAL SERVER has had to be power-cycled because the damn remote access card had crapped all over itself.
The problem is not someone talking to cell phone or to someone in person in the airplane. Or in a bus or train or wherever (excluding, well, churches, libraries etc.) The problem is people talking too loudly. I don't know why people often shout at their cell phones (like they wouldn't believe in RF transmission), but they often seem to speak unnecessary loudly at phone. It is possible to speak so quietly to the person sitting next to you that the person on the other side doesn't hear it. It is certainly possible to talk as quietly at cell phone.
It doesn't make sense to ban cell phones just because people sometimes shout at them, that would be like banning neighbor's TV because one can list set the volume too high. To fix the actual problem, people should be more polite and consider others, too. (Of course, that may be unrealistic to ask.)
but it's impossible, since posting the number in this comment would add to the total.
Recently posted: Intel's New Platform Verses AMD's 64-bit Prowess. Similar scope in benchmarks, perhaps better analyzed.
Fast Show anyone?
h anel9.shtml
For those who do not know: http://www.bbc.co.uk/comedy/fastshow/characters/c
Just don't do morse over ssh console session and think it's safe!
1 .p df1 1/08 /ssh_keystroke.html
http://www.cs.berkeley.edu/~daw/papers/ssh-use0
http://linux.oreillynet.com/lpt/a/linux/2001/
You'd have to use something like Ingo Molnar's netconsole patch but I guess it's doable.
http://home.earthlink.net/~descubes/C--
Where did he say that? Last I heard he used 3.2 in (at least one of) his machines because of sb-optimal 2.4 vm...
Also, the ac-series is different from Linus' tree; the diff between 2.4.4ac17 and 2.4.5 is many megs. (Incidentally 2.4.2 OOM rambo killed my diff process when cache filled up (230 megs cache), so I can't give you exact figures) Alan includes much more experimental stuff, but also has some other differences in his tree. Merging in between does happen piece by piece.
ISTR that recent make version should do the right thing with 'make -j3' as well, but now that I search for it I can't find it anywhere. Judging from the >40% performance boost, they pretty much maxed the dual boxes as well.
However, I'd be more conserned with the 2.4.4ac9/2.4.4 difference in the benchmark. I think the Athlon benchmark should've been rerun on 2.4.4ac9. The -ac kernels (and recent -2.4.5pre's I think) incorporate such optimizations and VM changes that I think the benchmark numbers can vary just because of that. Particularly on these kind of SMP tests that stress the scheduler, VM, fs and cache system -- and not just the CPU. Of course, there might not be a significant difference -- propably isn't -- but now we'll never know.
I think it was just a gwbasic proggie...
10 PLAY "edabd...."
etc.
I think I have it somewhere. I also think there's a program to play those gwbasic PLAY "" things on linux...
Come on, Savolainen can't be that hard to type... You could even cut'n'paste these days.
Also there has already been one pcspeaker patch for ages - it would've been nice if that had been mentioned.
Other than that, nice article.
Excuse me, but that's bullshit.
Cygwin doesn't slow anything down, when you don't use it. Even when you do, it doesn't. Sometimes some programs decide to hog 100% cpu time, which of course brings Windows to its knees - but this is a weakness in the NT scheduler. (Try DOS edit for an example.) Cygwin programs rarely do that, however.
I use zsh instead of bash (I absolutely love it), and I have that "zsh prompt here" in Exploder context menu.
I think I couldn't live without cygwin anymore. Give it a shot before you judge, you might like it (especially if you have unix background). The installallation is a breeze: just go to http://sources.redhat.com/cygwin and click install (I love the installer.)
You'll have to reverse the arguments of kill for that to work:
- kill(SIGALRM, getppid());
+ kill(getppid(), SIGALRM);
I think you may get lucky with google or freshmeat as well...
Of course zsh does that!
setopt automenu
(I think bash does as well.) Believe me: zsh (and bash) is much more powerful than 4NT. I use both on NT (zsh under cygwin). When I moved from DOS/4dos to unix I just couldn't believe what zsh was capable of.
I've seen it, but only on windows. And guess what? I happens much more often with IE!
Jeff Dike, the author on User-mode Linux has been actively seraching for people to undertake the Win32 port of UML. (UML is a cool hack that makes it possible to run Linux on top of the Unix system call API, eg. as a userlevel process in linux.) I think someone already started porting. They are using cygwin (at least in the beginning).
If they are succesful, one can execute linux binaries on Win32.
What are you all smoking?
/NT4SP6, Netscape 4.76 /NT4SP6, Latest Mozilla nightly / Redhat7.0, Konqueror-2.01/ Redhat7.0, Opera 4 beta3/ Redhat7.0) or have used (IE 5.0, IE 4.0, Netscape 4.5, 4.0, 3.0 (NT/Linux), Opera 4, 3.6 /NT)
Of all (GUI) browsers I use (IE 5.5sp1/NT4SP6, Opera 5/NT4SP6, Latest Mozilla nightly
Netscape 4.76 is ABSOLUTELY the most stable one. On both Linux and NT. It doesn't crash even once a month.
All the other browsers crash or hang (IE) every now and then - many of them don't do that annoyingly often, but more often than 4.76 anyhow.
Mozilla is perhaps the least stable (after Opera 4beta3 for linux) - but is definetely improving.
And the 4.7x series have been stable for me for long time.
4.76 IS slow and memory hungry, but Mozilla is MUCH more so (and Mozilla has other problems like missing shortcuts and mouse disturbing problems).
Guys, run memory test for your computer. Check your processor fan. Don't overclock. And if you using Win9x... just forget it.
I've seen a NT4SP5 (latest at the time) SMP
machine BSOD right after the 49.7 day watermark. That was not the only time I kept it up that long, and BSOD only happened once. That was my heavy-duty desktop machine with lots of compilation and multimedia stuff, so the uptimes were not all that bad.
If I had the money I'd still be interested in an SGI that can create games that make that much money. Who cares if the games suck ass as long as you collect the bucks.
Gee, I thought programmers created Quake. Well, I guess artifical intelligence is a further than I thought.
4.0 + nVidia 0.93 works fine for me too. But DPMS didn't work with 4.0 + xfree nvidia driver, and it does not work with nVidia 0.93 either. Also, DGA doesn't seem to work with Vmware. Those are the only things I miss from 3.3.x - performance is _much_ better (2d, nevermind 3d), and stability ain't much worse either. Does anyone know when/how dpms and dga will be supported?
IntelliSense could speed up developing if it actually worked. I can understand it's slow, but it's downright ridiculous I have to wait half a minute after typing "0." while IntelliSense tries to find class members for 0. It rarely finds any... Although it often doesn't find even if it should.
That said, similar tools are available for emacs and although they may not work any more correctly (C++ grammar is a PITA to parse), they are way faster and trimmer. I mean, 20MB browse info file and it takes 5 minutes to build it... come on!
After a few .c's, most of the include files are propably in the files cache anyway. At least my experience has been that compiling almost always nearly maxes out CPU (unless something stupid is happening like MSVC waiting for include files from a network disk.) Memory bandwidth is very likely big issue, though.
.config's), but I'd say Athlon won't propably lose much.
For the compiling speed, I was surprised to find how well a 550 Athlon did compiling linux kernel compared to Intel. I don't have any exact numbers (different