What is sheduling? "scheduling" ? Unix cron kicks the crap out of Win32 at any day. More flexiable and you can run it under differant users, etc.
>>>>>>>>>>>
I meant process scheduling. As in choosing which process to run next. Win2K makes special cases for GUI apps. For example, when a process releases a semaphore, it automatically gets a (temporary) 1 point priority boost. However, if the process is in the desktop's foreground, it gets a 2 point boost. If a process wakes up due to I/O being completed, it will get a 1 point boost if the I/O is to the disk, but an 8 point boost if the I/O is to a soundcard. These types of "hacks" violate standard UNIX semantics, but tend to make desktop-type apps have better interactive performance. Linux will never do this because it wants to be fair to all processes.
Win32 GUI is smooth, but with reniace X/X apps can run the same way. I have enlightenment at -10 nice;)
>>>>>>>>
I've had X down at -10 for years. It still isn't as smooth as Windows.
Win32 GUI:
more uniformed, yes
tons faster locally, yes
tons faster remotly, no
more customizable, no
>>>>>>>>>>>
Who cares about remote performance? I'm a desktop user. I SAID that Win2K isn't for everyone. However, for my purposes, it is better than Linux.
There is trade offs. Win32 GUI wins some, X wins some.
>>>>>>>.
Win2K wins all the desktop bits... (except maybe customization, but XP should help that)
*cough* isn't VC++ the compiler at allow `void main()` in C?;)
>>>>>>>>>>
Again, I don't develop heavy duty apps. For my OS design projects, Visual C++ is plenty standards complient. Plus, it has lots of features GCC doesn't, like non-braindead ASM inlining (because GCC wants to keep the backends and frontends separate) and keywords designed to help people writing hackish (ie. kernel type) code.
Speed: VC++ wins on win32 platforms
Standard complaint: gcc wins
Number of lauanges supported: gcc wins
Number of os supported: gcc wins
Number of procs suppored: gcc wins
Cost: gcc wins
>>>>>>>>>
Again, who cares about standards compliance, language support, OS support, or processor support? Maybe you do, but most desktop users don't.
Have you checked out vim with color support? I can go from source code -> compile -> run faster with a keyboard than you can with your mouse.
>>>>>>>.
Does VIM automatically give pop-ups for function prototypes? Some of these C++ classes can be a pain to memorize.
Put it in a ram disk.
>>>>>>>>>>
Oh, great solution. If you'd care to send me the extra RAM, I'll try it. Besides, isn't that what disk caching is for?
GTK is MUCH MUCH cleaner than MFC. Unix does have OpenGL (*cough* didn't these come from Unix (SGI) and not MS?).
>>>>>>>>>
Yes, I said that Win32 (and by extension MFC) sucks. However, DirectX does not. As for OpenGL, I don't see your point. Who cares where it came from, the point is that Windows supports it better than Linux does.
Go ahead and code up your app in DirectX... When you want to port it to MacOS*, BeOS, Unix, etc go ahead and have a fun time deturding all the MS crap from your application. Sure you cold pre-process it to holy hell, but the code base is going to be twice as large and twice as dirty. Use OpenGL and you don't have that problem...
>>>>>>>>>>>
DirectX and OpenGL are not directly comparable. There's lots of stuff that DirectX has that OpenGL does not. Sound, input, and MIDI APIs come to mind. Besides, why would you want to port it to UNIX? Everyone uses Windows anyway;)
Use the right tool for the job.
>>>>>>>>>
Look whose talking! I laid out my requirements, and Win2K is the right tool for the job! This f*cking thread started because you didn't like my.sig, and now you tell ME to use the right tool for the job? I know it sounds harsh, but it's a reality check to all you Linux grognards. You can't just blindly dismiss Windows as sucking and Linux being better. Windows does suck to some extent and in general, Linux sucks much less. However, Linux just sucks in the wrong places for most desktop users, and for those users, Windows is a more technically superior solution.
Actually, I multi-boot Win2K and Linux. I've been using Linux since Slack 3.5. Should teach you something about looking at the.sig (or the screename!) rather than the post. As for why I care, I was just curious. I do lots of graphics type applications and a good compiler can really speed up matrix processing (which lends itself to pipelining quite well).
Oh my god! He likes Windows 2000 better than Linux! He must be trolling! Let me give you a technical summery of why I like Win2K better:
A) Better sheduling. Shorter process quanta and priority boosts for GUI processes result in a much smoother GUI.
B) Better GUI: Not only more uniform, but tons faster than X.
C) Better OpenGL: Linux still can't beat Win2K OpenGL performance at high resolutions.
D) Better compilers: GCC might be nice and complient, but Visual C++ just plain produces faster code. Plus, PE is an easier format to play with for OS design than ELF.
E) Better IDE: KDevelop is almost on par with Visual Studio, but isn't quite there.
F) Better desktop. KDE and GNOME may equal Win2K feature-wise, but I'm sick of waiting so damn long for Konqueror to start up.
G) Better APIs. While Win32 may be pukalicious, DirectX is sweet, and in combination with OpenGL, there is nothing in Linux land (SDL, hah!) that can compete.
But this is just me. People with different needs will like other things better. NTFS is pretty bad as JFSs go, so servers should go elsewhere. The processes semantics are clearly wrong (as GUI apps get special boosts from the kernel) and for those who run in console mode, are detrimental to their work. The CLI sucks ass (even with Cygwin), and it DOES crash more often; two weeks vs. two months. Still, I reboot into Linux everyday, so what do I care? Linux is not the greatest OS ever made. Neither is Win2K. The greatest OS ever is Be... I mean, specific to the individual person. I might not be crazy about Linux, but that doesn't mean I'm trolling.
As far as standards go, this is not that terribly bad. Plus, it finally makes Linux start to resemble an OS, rather than a bunch of code people threw together. However, most of these "new" standards were standard anyway. It has stuff like X11, Xt, libz, ncurses, etc. The main problem with it is that it is too tame. While it standardizes some of the miscellaneous stuff, it totally ignores standards for GUI systems (other than X11, which was standard anway). To be complete, the LSB needs to dictate a GUI standard. It doesn't have to pick either KDE or GNOME, but it could take a subset of the functionality of both and define a source-level API standard that both could implement. Thus, there could still be multiple implementations, but apps wouldn't be tied to a particular one. If this sounds familer, its because it is basically what POSIX tries to do. POSIX was a good idea, and a standard GUI API would be too.
Actually, if you got your head out of your elitist ass, you'd realize that JFSs are useful to EVERYONE, not just 'leet server admins. BeOS users have had a kick-ass JFS for years now helping keep their MP3 collections safe. For these desktop users, speed matters as well as safety. Thus, a benchmark of JFSs IS valid. If XFS and JFS are both equally safe in the "pull-the-plug" test, then why not use the faster one?
Is it just me, or is everyone talking about which compiler can vectorize code better for cutting edge architectures, while GCC is still trying to get good P6 optimizations? Seriously, though, does anyone know if GCC 3.0 is in any way competitive with the new MS and Intel compilers?
I may sound like a troll of sorts or anti Intel, but when it comes to high end scientific engineering does anyone actually use anything outside the realms of Sun, Irix, and Alpha? Although benchmarks claim to show factual information, I've always seen them as a bit biased.
>>>>>>>>>>>
Not everyone working on a scientific application is blessed to be in a huge project with infintely deep pockets. There are tons of college students/projects doing different types of scientific computing, and x86 provides a very good price/performance ratio for these users.
More meaningless blathering about meaningless numbers. This article wasn't TRYING to measure real world performance! Why do you think they used a benchmark that fit entirely in L1 cache? They were simply trying to measure the peak throughput of the floating point units on the Athlon and the P4.
You probably have it misconfigured. I have three machines here running Win2K, and two of them run some pretty heavy duty apps (Photoshop, 3D Studio, Visual C++). Hasn't crashed on me yet. The majority of people on the web (either through posting or doing reviews or whatnot), the "word on the street" if you will, says that Win2K is damn stable. Thus, personal experience (neither mine nor yours) counts for nothing.
As for MS's operating systems, Linux could learn somethings from Win2K (just as Win2K could learn things from Linux) Win2K sheduler gives much better response to GUI apps than Linux's does, its GUI is much smoother, and some of the internals (look in an OS case study) are much more suited to a desktop OS. On the other hand, Win2K could stand to dump the Win32 environment subsystem paradigm, and pick up XFS and UVM while it was at it.
Whenever a potentially interesting tech article comes up on/., all we get is a bunch of pansy-assed posts about licenses or whatnot. This is news for NERDs. Where are the hard-hitting questions? Why hasn't somebody downloaded the damn thing and posted benchmarks comparing the major JFSs? (I'm working on it!) How fast is it? How stable is it? How easy is it to install? How does it work internally? Good grief, you'd think you'd get something meaty on a discussion like this...
Multiple filesystems don't add to fragmentation, since they are all interface compatible. When multiple things have the same interface (whether they be software components or cars) multiplicity is called 'choice'. When they have different interfaces (GNOME, KDE, etc) that is 'fragmentation'.
What does M$ have to do with a stable platform maintained by a competent company?
>>>>>>>>>>>>>
It's called Windows 2000. MS might not be the "nicest" company in the world, but nobody looking at the success of Windows would argue that the are not "competent."
Because Win2K supports my sound hardware better, and because it has better compilers. I still like BeOS more than anything else, I just can't use it right now.
Not really. Before the P6, all x86 chips had execution units that worked directly with x86 instructions. Now, the instruction units on the P4 don't even know what an x86 instruction is, all they handle are P4 micro-ops.
"Modern" CISC is essentially RISC. If you take a look at the architectures of ALL modern x86 CPUs, they are internally RISC, but decode x86 instructions. On Intel, these x86 instructions decode onto multiple "micro-ops" and on the P4, it is these micro-ops that are stored in L1 instruction cache. On the AMD K7, these instructions are called ROPs (RISC-like operations). On both these CPUs, the simple x86 instructions are translated directly into ROPs or micro-ops, while the more complicated ones are translated into multiple ROPs or micro-ops. On both the CPUs, the execution units only execute the micro-ops, not the x86 instructions themselves.
Actually, IA-64 is neither. It is based on something called VLIW (very long instruction word) which packs multiple, simple, independant instructions into one big instruction. Thus instruction can have its parts executed in parallel, which means that the CPU doesn't have to do instruction reordering.
Last I recall, Trolltech and KDE got a huge amount of flak for KDE's distributing non-GPL Qt with the GPL'ed KDE. Where was your idea then? Don't pretend that this stuff doesn't happen, because it does. A lot of GPL software writers are very anti-proprietory software and that causes problems when the two are mixed. The writer of dietlibc, for example, urges people not to port his software to Windows. The OSS community trying to pretend that they are totally innocent and being taken advantage of is just as much bullshit as MS pretending that OSS software will destroy "freedom-loving corporate America." (slightly paraphrased...)
Microsoft is going to try to spin this by saying that they are not trying to prevent people from adopting Open Source, but to protect themselves from licensing problems. I think there was a thread a while ago about what would happen if an MS drone used some GPL'ed code in Windows. They'll say that this is what they're trying to prevent. I'd give them the benifet of the doubt IF it wasn't for the following line: "ii) not using Potentially Viral Software (e.g. tools) to develop Recipient software which includes the Software, in whole or in part." Meaning you can't use GCC to compile the resultant program. Since no existing OSS license automatically places its output under said license, this clause is entirely an attempt to keep people from using GPL'ed development tools (and thus threatening their precious Visual Studio.) MS is making mistakes now. They're freaked about open source and their not reacting in the most subtle way possible. Their tactics would work if they were trying to go up against another big company like Sun, but directly going up against a community of developers is not good image. If they want to take out Linux, they're going to have to fire their stragegists and hire some people who know how to fight guirilla style.
He he. Funny you should say that. What about all the new engines that are coming out? Is the monitoring system part of the engine? Is the turbo-charger part of the engine? What about the fuel monitoring electronics? What about the intercooler? There's lots of grey areas where engines are concerned. Again, is Linux Linux without GNU? Theoretically, Linux is just the kernel, true. BUT, if you take away X and you take away the GNU userspace, can the OS rightly be called "Linux" or is it simply an OS based on the Linux kernel?
Slashdot Mirror
Would you believe me if I did?
What is sheduling? "scheduling" ? Unix cron kicks the crap out of Win32 at any day. More flexiable and you can run it under differant users, etc.
;)
;)
;)
.sig, and now you tell ME to use the right tool for the job? I know it sounds harsh, but it's a reality check to all you Linux grognards. You can't just blindly dismiss Windows as sucking and Linux being better. Windows does suck to some extent and in general, Linux sucks much less. However, Linux just sucks in the wrong places for most desktop users, and for those users, Windows is a more technically superior solution.
>>>>>>>>>>>
I meant process scheduling. As in choosing which process to run next. Win2K makes special cases for GUI apps. For example, when a process releases a semaphore, it automatically gets a (temporary) 1 point priority boost. However, if the process is in the desktop's foreground, it gets a 2 point boost. If a process wakes up due to I/O being completed, it will get a 1 point boost if the I/O is to the disk, but an 8 point boost if the I/O is to a soundcard. These types of "hacks" violate standard UNIX semantics, but tend to make desktop-type apps have better interactive performance. Linux will never do this because it wants to be fair to all processes.
Win32 GUI is smooth, but with reniace X/X apps can run the same way. I have enlightenment at -10 nice
>>>>>>>>
I've had X down at -10 for years. It still isn't as smooth as Windows.
Win32 GUI:
more uniformed, yes
tons faster locally, yes
tons faster remotly, no
more customizable, no
>>>>>>>>>>>
Who cares about remote performance? I'm a desktop user. I SAID that Win2K isn't for everyone. However, for my purposes, it is better than Linux.
There is trade offs. Win32 GUI wins some, X wins some.
>>>>>>>.
Win2K wins all the desktop bits... (except maybe customization, but XP should help that)
*cough* isn't VC++ the compiler at allow `void main()` in C?
>>>>>>>>>>
Again, I don't develop heavy duty apps. For my OS design projects, Visual C++ is plenty standards complient. Plus, it has lots of features GCC doesn't, like non-braindead ASM inlining (because GCC wants to keep the backends and frontends separate) and keywords designed to help people writing hackish (ie. kernel type) code.
Speed: VC++ wins on win32 platforms
Standard complaint: gcc wins
Number of lauanges supported: gcc wins
Number of os supported: gcc wins
Number of procs suppored: gcc wins
Cost: gcc wins
>>>>>>>>>
Again, who cares about standards compliance, language support, OS support, or processor support? Maybe you do, but most desktop users don't.
Have you checked out vim with color support? I can go from source code -> compile -> run faster with a keyboard than you can with your mouse.
>>>>>>>.
Does VIM automatically give pop-ups for function prototypes? Some of these C++ classes can be a pain to memorize.
Put it in a ram disk.
>>>>>>>>>>
Oh, great solution. If you'd care to send me the extra RAM, I'll try it. Besides, isn't that what disk caching is for?
GTK is MUCH MUCH cleaner than MFC. Unix does have OpenGL (*cough* didn't these come from Unix (SGI) and not MS?).
>>>>>>>>>
Yes, I said that Win32 (and by extension MFC) sucks. However, DirectX does not. As for OpenGL, I don't see your point. Who cares where it came from, the point is that Windows supports it better than Linux does.
Go ahead and code up your app in DirectX... When you want to port it to MacOS*, BeOS, Unix, etc go ahead and have a fun time deturding all the MS crap from your application. Sure you cold pre-process it to holy hell, but the code base is going to be twice as large and twice as dirty. Use OpenGL and you don't have that problem...
>>>>>>>>>>>
DirectX and OpenGL are not directly comparable. There's lots of stuff that DirectX has that OpenGL does not. Sound, input, and MIDI APIs come to mind. Besides, why would you want to port it to UNIX? Everyone uses Windows anyway
Use the right tool for the job.
>>>>>>>>>
Look whose talking! I laid out my requirements, and Win2K is the right tool for the job! This f*cking thread started because you didn't like my
Actually, I multi-boot Win2K and Linux. I've been using Linux since Slack 3.5. Should teach you something about looking at the .sig (or the screename!) rather than the post. As for why I care, I was just curious. I do lots of graphics type applications and a good compiler can really speed up matrix processing (which lends itself to pipelining quite well).
Oh my god! He likes Windows 2000 better than Linux! He must be trolling! Let me give you a technical summery of why I like Win2K better:
A) Better sheduling. Shorter process quanta and priority boosts for GUI processes result in a much smoother GUI.
B) Better GUI: Not only more uniform, but tons faster than X.
C) Better OpenGL: Linux still can't beat Win2K OpenGL performance at high resolutions.
D) Better compilers: GCC might be nice and complient, but Visual C++ just plain produces faster code. Plus, PE is an easier format to play with for OS design than ELF.
E) Better IDE: KDevelop is almost on par with Visual Studio, but isn't quite there.
F) Better desktop. KDE and GNOME may equal Win2K feature-wise, but I'm sick of waiting so damn long for Konqueror to start up.
G) Better APIs. While Win32 may be pukalicious, DirectX is sweet, and in combination with OpenGL, there is nothing in Linux land (SDL, hah!) that can compete.
But this is just me. People with different needs will like other things better. NTFS is pretty bad as JFSs go, so servers should go elsewhere. The processes semantics are clearly wrong (as GUI apps get special boosts from the kernel) and for those who run in console mode, are detrimental to their work. The CLI sucks ass (even with Cygwin), and it DOES crash more often; two weeks vs. two months. Still, I reboot into Linux everyday, so what do I care? Linux is not the greatest OS ever made. Neither is Win2K. The greatest OS ever is Be... I mean, specific to the individual person. I might not be crazy about Linux, but that doesn't mean I'm trolling.
As far as standards go, this is not that terribly bad. Plus, it finally makes Linux start to resemble an OS, rather than a bunch of code people threw together. However, most of these "new" standards were standard anyway. It has stuff like X11, Xt, libz, ncurses, etc. The main problem with it is that it is too tame. While it standardizes some of the miscellaneous stuff, it totally ignores standards for GUI systems (other than X11, which was standard anway). To be complete, the LSB needs to dictate a GUI standard. It doesn't have to pick either KDE or GNOME, but it could take a subset of the functionality of both and define a source-level API standard that both could implement. Thus, there could still be multiple implementations, but apps wouldn't be tied to a particular one. If this sounds familer, its because it is basically what POSIX tries to do. POSIX was a good idea, and a standard GUI API would be too.
Actually, if you got your head out of your elitist ass, you'd realize that JFSs are useful to EVERYONE, not just 'leet server admins. BeOS users have had a kick-ass JFS for years now helping keep their MP3 collections safe. For these desktop users, speed matters as well as safety. Thus, a benchmark of JFSs IS valid. If XFS and JFS are both equally safe in the "pull-the-plug" test, then why not use the faster one?
Is it just me, or is everyone talking about which compiler can vectorize code better for cutting edge architectures, while GCC is still trying to get good P6 optimizations? Seriously, though, does anyone know if GCC 3.0 is in any way competitive with the new MS and Intel compilers?
I may sound like a troll of sorts or anti Intel, but when it comes to high end scientific engineering does anyone actually use anything outside the realms of Sun, Irix, and Alpha? Although benchmarks claim to show factual information, I've always seen them as a bit biased.
>>>>>>>>>>>
Not everyone working on a scientific application is blessed to be in a huge project with infintely deep pockets. There are tons of college students/projects doing different types of scientific computing, and x86 provides a very good price/performance ratio for these users.
However, Intel's C compiler is in Beta for Linux. Thus, apps that need vectorizing could simply pony up $500 for a license and compile with that.
More meaningless blathering about meaningless numbers. This article wasn't TRYING to measure real world performance! Why do you think they used a benchmark that fit entirely in L1 cache? They were simply trying to measure the peak throughput of the floating point units on the Athlon and the P4.
As of yet, Intel's compiler is the only optimizing game in town. Even AMD uses Intel's compiler when giving Athlon benchmarks.
You probably have it misconfigured. I have three machines here running Win2K, and two of them run some pretty heavy duty apps (Photoshop, 3D Studio, Visual C++). Hasn't crashed on me yet. The majority of people on the web (either through posting or doing reviews or whatnot), the "word on the street" if you will, says that Win2K is damn stable. Thus, personal experience (neither mine nor yours) counts for nothing.
As for MS's operating systems, Linux could learn somethings from Win2K (just as Win2K could learn things from Linux) Win2K sheduler gives much better response to GUI apps than Linux's does, its GUI is much smoother, and some of the internals (look in an OS case study) are much more suited to a desktop OS. On the other hand, Win2K could stand to dump the Win32 environment subsystem paradigm, and pick up XFS and UVM while it was at it.
Whenever a potentially interesting tech article comes up on /., all we get is a bunch of pansy-assed posts about licenses or whatnot. This is news for NERDs. Where are the hard-hitting questions? Why hasn't somebody downloaded the damn thing and posted benchmarks comparing the major JFSs? (I'm working on it!) How fast is it? How stable is it? How easy is it to install? How does it work internally? Good grief, you'd think you'd get something meaty on a discussion like this...
Actually, XFS is dynamically resizable too. And fast at that.
BTW> Is it just me, or do ugly people and slow systems have something in common...
Multiple filesystems don't add to fragmentation, since they are all interface compatible. When multiple things have the same interface (whether they be software components or cars) multiplicity is called 'choice'. When they have different interfaces (GNOME, KDE, etc) that is 'fragmentation'.
What does M$ have to do with a stable platform maintained by a competent company?
>>>>>>>>>>>>>
It's called Windows 2000. MS might not be the "nicest" company in the world, but nobody looking at the success of Windows would argue that the are not "competent."
Because Win2K supports my sound hardware better, and because it has better compilers. I still like BeOS more than anything else, I just can't use it right now.
130.
Not really. Before the P6, all x86 chips had execution units that worked directly with x86 instructions. Now, the instruction units on the P4 don't even know what an x86 instruction is, all they handle are P4 micro-ops.
Hey, don't the lastest benchmarks show the Itanium kicking Alpha's ass in floating point?
"Modern" CISC is essentially RISC. If you take a look at the architectures of ALL modern x86 CPUs, they are internally RISC, but decode x86 instructions. On Intel, these x86 instructions decode onto multiple "micro-ops" and on the P4, it is these micro-ops that are stored in L1 instruction cache. On the AMD K7, these instructions are called ROPs (RISC-like operations). On both these CPUs, the simple x86 instructions are translated directly into ROPs or micro-ops, while the more complicated ones are translated into multiple ROPs or micro-ops. On both the CPUs, the execution units only execute the micro-ops, not the x86 instructions themselves.
Actually, IA-64 is neither. It is based on something called VLIW (very long instruction word) which packs multiple, simple, independant instructions into one big instruction. Thus instruction can have its parts executed in parallel, which means that the CPU doesn't have to do instruction reordering.
Last I recall, Trolltech and KDE got a huge amount of flak for KDE's distributing non-GPL Qt with the GPL'ed KDE. Where was your idea then? Don't pretend that this stuff doesn't happen, because it does. A lot of GPL software writers are very anti-proprietory software and that causes problems when the two are mixed. The writer of dietlibc, for example, urges people not to port his software to Windows. The OSS community trying to pretend that they are totally innocent and being taken advantage of is just as much bullshit as MS pretending that OSS software will destroy "freedom-loving corporate America." (slightly paraphrased...)
Microsoft is going to try to spin this by saying that they are not trying to prevent people from adopting Open Source, but to protect themselves from licensing problems. I think there was a thread a while ago about what would happen if an MS drone used some GPL'ed code in Windows. They'll say that this is what they're trying to prevent. I'd give them the benifet of the doubt IF it wasn't for the following line: "ii) not using Potentially Viral Software (e.g. tools) to develop Recipient software which includes the Software, in whole or in part." Meaning you can't use GCC to compile the resultant program. Since no existing OSS license automatically places its output under said license, this clause is entirely an attempt to keep people from using GPL'ed development tools (and thus threatening their precious Visual Studio.) MS is making mistakes now. They're freaked about open source and their not reacting in the most subtle way possible. Their tactics would work if they were trying to go up against another big company like Sun, but directly going up against a community of developers is not good image. If they want to take out Linux, they're going to have to fire their stragegists and hire some people who know how to fight guirilla style.
He he. Funny you should say that. What about all the new engines that are coming out? Is the monitoring system part of the engine? Is the turbo-charger part of the engine? What about the fuel monitoring electronics? What about the intercooler? There's lots of grey areas where engines are concerned. Again, is Linux Linux without GNU? Theoretically, Linux is just the kernel, true. BUT, if you take away X and you take away the GNU userspace, can the OS rightly be called "Linux" or is it simply an OS based on the Linux kernel?