Excuse me? Since when were the original Riva drivers are crappy? nVidia's drivers are among the most stable ones available and have been for a long time. I don't see current OSS projects being a pinnacle of stability either. Look at GNOME it is bloated and slow and getting even more bloated and slow with upcoming releases. (Mozilla used as an HTML viewing object? Isn't it really far along and still leaks like hell and uses a bloody huge memory footprint? God help us! Cobra used to implement reusable objects? What the hell is wrong with them! Even the Berlin guys admit that Cobra is slow. At least KDE has the right idea. KOM or something else similar to COM is the way to go for reusable objects. Stuff that needs cobra-like feautures can implement those at a higher level as to avoid bloat in the core. Still, KDE isn't exactly the worlds most efficiant GUI either. As for OSS drivers, what about all the crappy sound drivers that don't support duplexing and whatnot. True, in general OSS drivers are higher stability, but they are rarely as fast. The only graphics driver that runs faster in Linux than in Windows (ie. one that makes Linux look like it should) is the Matrox made G400 driver.
Oh get of you're jihad. Sure that's your opinion, but you can't go around saying that that's the "Right Thing" without proof, evidance, or some shred of anything that this is something other than your opinion. If I got on here and said NO MORE OPEN SOURCE KERNEL MODULES! people would flame my ass so fast it isn't even funny. What gives people the right to do the opposite thing? In my opinion, a driver manufacturer should do what they want to do with their source. If that means that you won't buy their cards, so be it. I doubt they care. Look where demanding open source drivers has got us. We have all this infrastructure without many company made drivers. Linus knows it, thats why binarly only kenel modules have gotten so easy. The XFree guys know it, thats why XFree4.0 supports binary drivers so well. Coming from a windows world, I can tell you that drivers matter. I still think that the driver manufacturer are in a better position to write drivers for their cards. Its just from the dynamics of the thing. How is a person not deeply involved with a project supposed to know about all the intracacies of the hardware? Docs only go so far. And if they did document all these naunces, they might was well write the driver themselves. Take the detonator drivers on the Riva 128 and TNT for example. Performance increased a good 35% (brought a 128 close to a Voodoo2 for QuakeII), just by changing drivers. You rarely see that kind of optimization in Open source stuff.
Well, extensions might not really be the best way to do this. Consider OpenGL. It really wasn't designed for a way to expand the API other than the extension mechanism. This works fine for awhile, but breaks down eventually. Believe it or not, MS actually has it right with the DirectX model of doing things. It has proved to be flexible, versitile, and stable. (Remember, mostof the instability of DirectX comes from dealing with windows. The other part comes from the fact that it trades utter stability for speed and passable stability.) MS puts a new featues, say texture compression, into the API, emulating it at first, and providing a "carrot" for vendors to come in and implement it in hardware. If a vendor wants, they can implement their own scheme, (like with S3TC and the 3DFx one) but eventually everybody standardizes on one thing. This is easy for developers, because most features in the API are emulated if not available on hardware, and eventually a card will come out that implements this features. Also there are not a lot of competing implementations of features. This is good for the API maker because most games will use the API to a much fuller extent, and it is good for the card maker because they can implement API features in hardware and thus have a trump card over competitors. This is in contrast to the OpenGL extension mechanism. Updates for OpenGL come out relativly infrequently, and any new features come out as extensions first. If you take a look at the extensions supported by your card, you can see there are quite a few implementations of extensions, all imcompatible. The ARB dictate kind of nullifies this, but they don't dictate very often. Also there is no "carrot" for the hardware vendors. This difference comes mostly from the original market of the API (OpenGL from workstations, supported the fast preview stuff, knowing most high quality effects would be done in a software render for final rendering. Direct3D on the other hand wants the preview to look as high quality as possible, so implemented a method to easily add hardware acceleration for special effects.) I'm not saying OpenGL has lower visual quality, I'm saying it is easier to extend Direct3D in a device independant manner. Thats why I was so happy when I heard about Farenheit. It was to meld the great OpenGL API with some of the cool features of Direct3D such as easy extendibility, an object-based interface, and good integration with a 2D API (namely DirectDraw). Too bad it died.
Yea, everybody should have the libraries installed, but... A) Its wasteful. Loading the Qt and KDE libraries on my GNOME system eats up about 6 to 8 megs of RAM. Times that by 4, which it will be by the time the 2.0 series comes around, thats 24-32 meg for the GUI libraries! B) It's unnecessary. Programming structures aside, GNOME and KDE are not terribly different. Sure on the surface they look different, but at the core, they do the same thing, and they both do it in mostly the same way. Whats the point of the redundancy. Back when the KDEvsGNOME flame wars were going on, people were complaining about the duplicated programming effort. I could care less about that, what about the duplicated burden on my memory! C) New office apps should not demand a high power system. Bloat is bloat wether it comes from Linux or Windows. Office is bloated and GNOME/KDE together is bloated. The whole point of Linux is elegance and speed (and stability for the sysadmins.) It's silly when Gobe Productive, (which whoops Abiword, and comes damn close to wordperfect) can load in less memory than GNOME. (6-7 megs.) The desktop UI should not put that much of a burden on the system. One of the evangalists for DirectX was saying how MS puts an 80% OS tax on everything. Linux is becoming no different. The OS tax should be down around 10% like it is in BeOS. (Maybe 20% for all the extra stuff in Linux, but still way lower than it is now.)
Metacreations makes some nice products, but to tell the truth their UIs are fairly retarded. Nontheless, I feel this is good for Linux. Assuming these are ported, it makes Linux viable as a graphics tool. (Kinda like SGI, though BeOS already has the "poor man's SGI" moniker.) Some of these apps may go over well, but I doubt stuff like Bryce (does it use 3D hardware yet?) or other 3D apps will go over too well on Linux. In that arena, NT still has the lead in terms of hardware acceleration. Sure linux has DRI, but it is relatvily unsupported (I don't consider the Voodoo3 to be a graphics person's 3D card. 16bit color c'mon! That's so passe!) Linux is aiming at the workstation market in terms of desktop (at least thus far, IMNSHO it is still not ready for the corporate desktop and definatly not for the home user) Thus it has to support workstation level 3D cards. There is one ray of hope, however. Though it pains OSS zealots everywhere, these kind of apps in conjuction with nVidia's propriotory, binary-only 3D drivers for it's Quadro card (GeForce and Quadro are very similar, so if GeForce drivers come out, Quadro should too.) will catapault Linux to a major hold on the workstation market.
DISCLAMER: Some sense of asthetics is necessary to appreciate this post. People using FVWM need not read further. ______________________________________________ Take a look at Corel Wordperfect for Linux. Then take a look at the windows version. Which looks more pleasing? The windows version of course. Take a look at the Corel help file for Linux, then take a look at the one for windows. Which one isn't a garish shade of blue? Linux has advanced significantly in the prettyiness department, and GNOME/w Sawmill is probably the most asthetically pleasing GUI out there. (Nice looking without being garish. Probably the best looking GUI in my taste) It is also farily fast and stable (with XFree4.0 and some tweeks) For those of you who like KDE, KDE is just as good. Now with these good desktops out, why does Corel continue to use the basic X widgets. Not only does it add bloat, it looks ugly. The reason is that they cannot count on everybody having GNOME or KDE installed, or those that do, having enough RAM to support both GTK and Qt if they have a different one from the one Corel uses. This hodgepodge of libraries will grow by two in the coming year. GNOME 2.0 and KDE 2.0 will come out which will increase the memory use even more. Thus, developers will continue to use custom (and inconsistant) widgets, or use the standard X widgets. Neither is an acceptable option for most people. As you've guessed by now, this is a cry for a standard widget set. (Not window manager mind you, just widget set and environment.) With a clearly defined standard so many window managers could easily implant themselves into the system, the resultant desktop would be both functional and flexible. Now the obsessive choice people scream out, "But we want freedom! That's not what Linux is about!" I tell them to ask themselves. What freedom are you asking about? There are only three significantly different GUIs at the core level (not the interface/window manager level, the environment level, like GNOME core) One of them, GNUStep, is hardly developed for. The other is the standard X libraries which get the bulk of the development, and the other is GNOME/KDE which are very different at the interface level, and maybe implementation details, but essentially try to solve the same thing. One uses KOM, the other tries to use Cobra and Bonobo, but both are trying to make reusable objects. Do you care if your reusable objects use one tech vs. the other? Sure they look different, but thats a theme/window manager thing. Now maybe it isn't a problem now, but as applications become larger, developers will be increasingly reluctant to develop for either KDE or GNOME. Take KDevelop for example. I like GNOME, but I like KDevelop too. So I run KDevelop under GNOME and just eat the extra 15-20 meg of memory usage. What does this have to do with meta creations? I'm getting to that. Metacreations represents standard mainstream apps that are being ported to Linux. We have two good GUIs on Linux, but it is doubtful that these apps will use either, for fear of locking people out. So we have some nice apps tethered with ugly GUIs. As more apps are ported, this will hinder Linux's mainstream acceptance. People don't like ugly apps. In the real world, ascthetics counts for a lot, especially among the graphics artists crowd. Why do you think SGIs and Macs look so nice? In addition, among most normal people, asthetics counts. I can't work in a ugly room, I wouldn't buy an ugly desk, and I don't like reading ugly documents. Same for my computer. Wordperfect 8 for linux is highly functional, but I hate using it because it is so ugly. I use Abiword instead, which leaves something to be desired. (Actually, I use BeOS most of the time, but when I do use Linux, its Abiword.) At least it supports GTK nicely. Choice is a great thing, but intelligance is too. A well planned, well thought out, carefully executed GUI can have both flexibility, and functionality. Take the BeOS GUI. The actual stuff associated with the UI is very small. This portion can be switched out quite easily, with the major underlying technologies intact. Sure this means that you're stuck to either COM or COBRA (COBRA sucks btw. COM is fast and elegant, Cobra is flexible and slow, 'nuff said) but I doubt many end users care one way or the other.
This is hilarious. I just read the first 5 posts on this article. All ACs, all score 0, but all good posts. What is wrong with the moderators? Why are these posts at the same level as the grits guy? Read at -1 moderators! In my opinion it signals a big problem that the/. moderation system wasn't designed to handle. Now that/. was a much larger readerbase, not everyone will be a member, and not everyone will bother to make an account. Also with the hidiously increasing number of trolls, moderators are increasingly moving to reading at 1 or above. That leaves out a lot of ACs who have a good point, but either don't want to bother to create an account, or have an unpopular viewpoint and don't want to endanger their account. I suggest that the moderation system be changed to something so you can browse based on what it was moderated to, not just the score. So people you could browse by insightful, or AC, and ignore the troll or flamebait posts. This would also require moderator to more activly moderate down the trolls, but they are more willing to do that than moderate up the ACs.
This is hilarious. I just read the first 5 posts on this article. All ACs, all score 0, but all good posts. What is wrong with the moderators? Why are these posts at the same level as the grits guy? Read at -1 moderators! In my opinion it signals a big problem that the/. moderation system wasn't designed to handle. Now that/. was a much larger readerbase, not everyone will be a member, and not everyone will bother to make an account. Also with the hidiously increasing number of trolls, moderators are increasingly moving to reading at 1 or above. That leaves out a lot of ACs who have a good point, but either don't want to bother to create an account, or have an unpopular viewpoint and don't want to endanger their account. I suggest that the moderation system be changed to something so you can browse based on what it was moderated to, not just the score. So people you could browse by insightful, or AC, and ignore the troll or flamebait posts. This would also require moderator to more activly moderate down the trolls, but they seem more willing to do that than moderate up the ACs.
I'm sorry, I can't give an exact reference and I don't have time to look now. But I remember a recent/. thread where there was references to MS blaming users on doing something wrong being a reason why NT was unstable. Weak evidance,I know, but if you look you can probably find something.
Does anybody bother to read the entire post anymore? I had a few questions. That might deserve an offtopic (fine if you don't want to answer my question) but I doubt it deserves a flamebait. Or were you just reading the first two lines?
Yea, but then you miss a lot of the good posts from ACs who have a good point. A lot of the pro MS stuff gets posted by ACs and/. moderators rarely bother to mark them up. Thats why I keep saying that slashdot needs to retool its moderation system a bit.
These OSs become mainstream when people other than sysadmins and nerds use them. When they are preinstalled at CompUSA is when they become mainstream. It's what people "mainly" use, hence the name.
Well, either way, it's slow. Thats why I keep saying network transparency should be at a higher level, but I wasn't around back when X was designed. And XFree 4.0 still isn't nearly as fast as BeOS even though it has supposedly no performance hit on local machines.
Actually BeOS runs graphics and UI in a seperate server too, and I've never had any speed issues with it (aside form the fact that it makes me cry to go back into windows to do something.)
I'm tired of all you people saying that faster procs are useless because of other bottlenecks. I got news for you, in most non-server environments the proc is still the biggest bottle neck. To tell the truth, I enjoyed much more the 50% boost from 200 to 300 MHz than I did the 50% boost from a 66 to a 100MHz bus. For the most part apps are still computer bound, EXCEPT in server space. Thats why the Xeon is still chugging along at 550MHz. Examples of apps that are compute bound. 1) 3D, games, rendering,modling, you name it. 2) Any kind of realtime graphics. 3) Photoshop type apps depending on wether you use filters more often or just edits to large files. 4) Compiling. 5) Audio editing 6) Real-time video editing. (What, I have to wait 2 minutes to render the changes!) Things not compute bound 1) Serving webpages, files, etc. 2) Working with large photoshop files. 3) Some types of scientific computing where data crunching is high volume, low workload. Yes I've forgotten a few tihngs, but the market for more CPU power is clearly more important than the market for higher bandwidth.
1) If you talking pgcc, it doesn't come standard with most distros. If you think I am going to recompile every app on my system with pgcc, you've lost it. 2) Correct, XFree4.0 is pretty good. But he is talking BeOS level here. Even XFree4.0 is nowhere near the performance you get with BeOS. (Lets put it this way. On the SDL demos, BeOS is about 25% faster that Windows95 without hardware acceleration (SDL for BeOS does not support hardware acceleration.)Now Windows is pretty fast, especially considering that SDL for Windows uses DirectDraw (though I ran the test without a hardware blitter.) Now you go into BDirectWindow an you leave everything else in the dust. 4) Actually, the important Linux GUIs, Qt and GTK+ are not thread safe. Qt is very not thead safe and GTK+ requires a bit of a hack to be thread safe. 5) Well, no-one ever said that Unicies are good at threading. Linux is worse than WindowsNT at threads (in management, switch time, and spawn time) and BeOS is about 10 times faster than NT. You do that math. 6) Yea, and if I use TWM I'll have to burn my eyeballs out. If you want speed you can use DOS, but no sane person would. 7) Yea BeOS is free. Maybe not OpenSource, but free. It is a very open system, has some open source parts and in general the guys at be are pretty good with the source. Now I would rather use a fast, mostly closed system, than a slower OSS system.
Actually, that is a pretty stupid reason to call it GNU/Linux. There is a lot more to Linux than tar and grep. BeOS uses the GNU suite too. Does that make it GNU/BeOS? Hell no. Face it, the GNU stuff is pretty minor these days. Back in the day, maybe tar was worth something. These days tar and cp aren't exactly a major part of the OS, (unlike GNOME, or X, or KDE, or OSS, ad nauseum.)
Re:A "Free" OS from a company that is croaking
on
BeOS For Linux!
·
· Score: 2
Actually a fairly functional BeOS gets down to 2 meg. Quite a bit bigger than QNX, but still not horribaly bloated.
Re:It's definitely not for PPC
on
BeOS For Linux!
·
· Score: 2
There is a difference between working and advanced. A laser sight rifle is really advanced, but a bazooka is a lot more destructive. I was talking about advanced in the sense of newer software design. However, the BeOS is microkernel done right, and is a big improvement on monolithic kernels. I have tried restarting the different servers. Aside from the app server (which can't be restarted becase it has to be there to run the script) most of the other servers can be restarted. In fact, most changes in options are done this way. Change network settings? Change it and the system reboots the net server. Same thing for the media server. And BeOS does not run a gigantic server process, there are quite a few independant ones, and they are all modular. This comes in the form of stability, scalability, and flexibility. A badly implemented server can't take the kernel out, (badly written drivers tend to do this in NT), the server load can be distributed over many procs both on a process level and on a thread level (a BeOS system has over 100 threads going at bootup.) and flexibility comes from the ability to modify or replace those servers on the fly.
Opps. Sorry, Haven't been keeping track of the OSS modules lately. However, the vortex drivers are still in beta.
Re:Where will it go when they're on top?
on
BeOS For Linux!
·
· Score: 2
I'm not talking about the OSS community. I'm talking about all the commercial developers using Linux for their products. You can bet if FreeBSD had as much developer support they would use that instead.
Is it a little disturbing to you that this thing has the Linux icon? Sure I can kind of understand that this is the FreeBe version FOR Linux, but wouldn't the Be icon be more appropriate as its is BEOS for linux? Or is the author trying to take advantage of the fact that an article filed under linux has a much better chance of getting in? (Actually I'm not trolling. You wouldn't believe the alternative OS stuff that gets ignored on/. That neverwinter nights thing really pissed me off (where neverwinter nights for Linux was the only one posted on/. even though the BeOS version had been announced a month before.))
Slashdot Mirror
Excuse me? Since when were the original Riva drivers are crappy? nVidia's drivers are among the most stable ones available and have been for a long time. I don't see current OSS projects being a pinnacle of stability either. Look at GNOME it is bloated and slow and getting even more bloated and slow with upcoming releases. (Mozilla used as an HTML viewing object? Isn't it really far along and still leaks like hell and uses a bloody huge memory footprint? God help us! Cobra used to implement reusable objects? What the hell is wrong with them! Even the Berlin guys admit that Cobra is slow. At least KDE has the right idea. KOM or something else similar to COM is the way to go for reusable objects. Stuff that needs cobra-like feautures can implement those at a higher level as to avoid bloat in the core. Still, KDE isn't exactly the worlds most efficiant GUI either. As for OSS drivers, what about all the crappy sound drivers that don't support duplexing and whatnot. True, in general OSS drivers are higher stability, but they are rarely as fast. The only graphics driver that runs faster in Linux than in Windows (ie. one that makes Linux look like it should) is the Matrox made G400 driver.
Oh get of you're jihad. Sure that's your opinion, but you can't go around saying that that's the "Right Thing" without proof, evidance, or some shred of anything that this is something other than your opinion. If I got on here and said
NO MORE OPEN SOURCE KERNEL MODULES! people would flame my ass so fast it isn't even funny. What gives people the right to do the opposite thing? In my opinion, a driver manufacturer should do what they want to do with their source. If that means that you won't buy their cards, so be it. I doubt they care. Look where demanding open source drivers has got us. We have all this infrastructure without many company made drivers. Linus knows it, thats why binarly only kenel modules have gotten so easy. The XFree guys know it, thats why XFree4.0 supports binary drivers so well. Coming from a windows world, I can tell you that drivers matter. I still think that the driver manufacturer are in a better position to write drivers for their cards. Its just from the dynamics of the thing. How is a person not deeply involved with a project supposed to know about all the intracacies of the hardware? Docs only go so far. And if they did document all these naunces, they might was well write the driver themselves. Take the detonator drivers on the Riva 128 and TNT for example. Performance increased a good 35% (brought a 128 close to a Voodoo2 for QuakeII), just by changing drivers. You rarely see that kind of optimization in Open source stuff.
Well, extensions might not really be the best way to do this. Consider OpenGL. It really wasn't designed for a way to expand the API other than the extension mechanism. This works fine for awhile, but breaks down eventually. Believe it or not, MS actually has it right with the DirectX model of doing things. It has proved to be flexible, versitile, and stable. (Remember, mostof the instability of DirectX comes from dealing with windows. The other part comes from the fact that it trades utter stability for speed and passable stability.) MS puts a new featues, say texture compression, into the API, emulating it at first, and providing a "carrot" for vendors to come in and implement it in hardware. If a vendor wants, they can implement their own scheme, (like with S3TC and the 3DFx one) but eventually everybody standardizes on one thing. This is easy for developers, because most features in the API are emulated if not available on hardware, and eventually a card will come out that implements this features. Also there are not a lot of competing implementations of features. This is good for the API maker because most games will use the API to a much fuller extent, and it is good for the card maker because they can implement API features in hardware and thus have a trump card over competitors. This is in contrast to the OpenGL extension mechanism. Updates for OpenGL come out relativly infrequently, and any new features come out as extensions first. If you take a look at the extensions supported by your card, you can see there are quite a few implementations of extensions, all imcompatible. The ARB dictate kind of nullifies this, but they don't dictate very often. Also there is no "carrot" for the hardware vendors. This difference comes mostly from the original market of the API (OpenGL from workstations, supported the fast preview stuff, knowing most high quality effects would be done in a software render for final rendering. Direct3D on the other hand wants the preview to look as high quality as possible, so implemented a method to easily add hardware acceleration for special effects.) I'm not saying OpenGL has lower visual quality, I'm saying it is easier to extend Direct3D in a device independant manner. Thats why I was so happy when I heard about Farenheit. It was to meld the great OpenGL API with some of the cool features of Direct3D such as easy extendibility, an object-based interface, and good integration with a 2D API (namely DirectDraw). Too bad it died.
Yea, everybody should have the libraries installed, but...
A) Its wasteful. Loading the Qt and KDE libraries on my GNOME system eats up about 6 to 8 megs of RAM. Times that by 4, which it will be by the time the 2.0 series comes around, thats 24-32 meg for the GUI libraries!
B) It's unnecessary. Programming structures aside, GNOME and KDE are not terribly different. Sure on the surface they look different, but at the core, they do the same thing, and they both do it in mostly the same way. Whats the point of the redundancy. Back when the KDEvsGNOME flame wars were going on, people were complaining about the duplicated programming effort. I could care less about that, what about the duplicated burden on my memory!
C) New office apps should not demand a high power system. Bloat is bloat wether it comes from Linux or Windows. Office is bloated and GNOME/KDE together is bloated. The whole point of Linux is elegance and speed (and stability for the sysadmins.) It's silly when Gobe Productive, (which whoops Abiword, and comes damn close to wordperfect) can load in less memory than GNOME. (6-7 megs.) The desktop UI should not put that much of a burden on the system. One of the evangalists for DirectX was saying how MS puts an 80% OS tax on everything. Linux is becoming no different. The OS tax should be down around 10% like it is in BeOS. (Maybe 20% for all the extra stuff in Linux, but still way lower than it is now.)
Actually I did write it in the tiny box. But I don't consider this a document so it doesn't matter.
Metacreations makes some nice products, but to tell the truth their UIs are fairly retarded. Nontheless, I feel this is good for Linux. Assuming these are ported, it makes Linux viable as a graphics tool. (Kinda like SGI, though BeOS already has the "poor man's SGI" moniker.) Some of these apps may go over well, but I doubt stuff like Bryce (does it use 3D hardware yet?) or other 3D apps will go over too well on Linux. In that arena, NT still has the lead in terms of hardware acceleration. Sure linux has DRI, but it is relatvily unsupported (I don't consider the Voodoo3 to be a graphics person's 3D card. 16bit color c'mon! That's so passe!) Linux is aiming at the workstation market in terms of desktop (at least thus far, IMNSHO it is still not ready for the corporate desktop and definatly not for the home user) Thus it has to support workstation level 3D cards. There is one ray of hope, however. Though it pains OSS zealots everywhere, these kind of apps in conjuction with nVidia's propriotory, binary-only 3D drivers for it's Quadro card (GeForce and Quadro are very similar, so if GeForce drivers come out, Quadro should too.) will catapault Linux to a major hold on the workstation market.
DISCLAMER: Some sense of asthetics is necessary to appreciate this post. People using FVWM need not read further. /w Sawmill is probably the most asthetically pleasing GUI out there. (Nice looking without being garish. Probably the best looking GUI in my taste) It is also farily fast and stable (with XFree4.0 and some tweeks) For those of you who like KDE, KDE is just as good. Now with these good desktops out, why does Corel continue to use the basic X widgets. Not only does it add bloat, it looks ugly. The reason is that they cannot count on everybody having GNOME or KDE installed, or those that do, having enough RAM to support both GTK and Qt if they have a different one from the one Corel uses. This hodgepodge of libraries will grow by two in the coming year. GNOME 2.0 and KDE 2.0 will come out which will increase the memory use even more. Thus, developers will continue to use custom (and inconsistant) widgets, or use the standard X widgets. Neither is an acceptable option for most people. As you've guessed by now, this is a cry for a standard widget set. (Not window manager mind you, just widget set and environment.) With a clearly defined standard so many window managers could easily implant themselves into the system, the resultant desktop would be both functional and flexible. Now the obsessive choice people scream out, "But we want freedom! That's not what Linux is about!" I tell them to ask themselves. What freedom are you asking about? There are only three significantly different GUIs at the core level (not the interface/window manager level, the environment level, like GNOME core) One of them, GNUStep, is hardly developed for. The other is the standard X libraries which get the bulk of the development, and the other is GNOME/KDE which are very different at the interface level, and maybe implementation details, but essentially try to solve the same thing. One uses KOM, the other tries to use Cobra and Bonobo, but both are trying to make reusable objects. Do you care if your reusable objects use one tech vs. the other? Sure they look different, but thats a theme/window manager thing. Now maybe it isn't a problem now, but as applications become larger, developers will be increasingly reluctant to develop for either KDE or GNOME. Take KDevelop for example. I like GNOME, but I like KDevelop too. So I run KDevelop under GNOME and just eat the extra 15-20 meg of memory usage. What does this have to do with meta creations? I'm getting to that. Metacreations represents standard mainstream apps that are being ported to Linux. We have two good GUIs on Linux, but it is doubtful that these apps will use either, for fear of locking people out. So we have some nice apps tethered with ugly GUIs. As more apps are ported, this will hinder Linux's mainstream acceptance. People don't like ugly apps. In the real world, ascthetics counts for a lot, especially among the graphics artists crowd. Why do you think SGIs and Macs look so nice? In addition, among most normal people, asthetics counts. I can't work in a ugly room, I wouldn't buy an ugly desk, and I don't like reading ugly documents. Same for my computer. Wordperfect 8 for linux is highly functional, but I hate using it because it is so ugly. I use Abiword instead, which leaves something to be desired. (Actually, I use BeOS most of the time, but when I do use Linux, its Abiword.) At least it supports GTK nicely. Choice is a great thing, but intelligance is too. A well planned, well thought out, carefully executed GUI can have both flexibility, and functionality. Take the BeOS GUI. The actual stuff associated with the UI is very small. This portion can be switched out quite easily, with the major underlying technologies intact. Sure this means that you're stuck to either COM or COBRA (COBRA sucks btw. COM is fast and elegant, Cobra is flexible and slow, 'nuff said) but I doubt many end users care one way or the other.
______________________________________________
Take a look at Corel Wordperfect for Linux. Then take a look at the windows version. Which looks more pleasing? The windows version of course. Take a look at the Corel help file for Linux, then take a look at the one for windows. Which one isn't a garish shade of blue? Linux has advanced significantly in the prettyiness department, and GNOME
This is hilarious. I just read the first 5 posts on this article. All ACs, all score 0, but all good posts. What is wrong with the moderators? Why are these posts at the same level as the grits guy? Read at -1 moderators! In my opinion it signals a big problem that the /. moderation system wasn't designed to handle. Now that /. was a much larger readerbase, not everyone will be a member, and not everyone will bother to make an account. Also with the hidiously increasing number of trolls, moderators are increasingly moving to reading at 1 or above. That leaves out a lot of ACs who have a good point, but either don't want to bother to create an account, or have an unpopular viewpoint and don't want to endanger their account. I suggest that the moderation system be changed to something so you can browse based on what it was moderated to, not just the score. So people you could browse by insightful, or AC, and ignore the troll or flamebait posts. This would also require moderator to more activly moderate down the trolls, but they are more willing to do that than moderate up the ACs.
This is hilarious. I just read the first 5 posts on this article. All ACs, all score 0, but all good posts. What is wrong with the moderators? Why are these posts at the same level as the grits guy? Read at -1 moderators! In my opinion it signals a big problem that the /. moderation system wasn't designed to handle. Now that /. was a much larger readerbase, not everyone will be a member, and not everyone will bother to make an account. Also with the hidiously increasing number of trolls, moderators are increasingly moving to reading at 1 or above. That leaves out a lot of ACs who have a good point, but either don't want to bother to create an account, or have an unpopular viewpoint and don't want to endanger their account. I suggest that the moderation system be changed to something so you can browse based on what it was moderated to, not just the score. So people you could browse by insightful, or AC, and ignore the troll or flamebait posts. This would also require moderator to more activly moderate down the trolls, but they seem more willing to do that than moderate up the ACs.
I'm sorry, I can't give an exact reference and I don't have time to look now. But I remember a recent /. thread where there was references to MS blaming users on doing something wrong being a reason why NT was unstable. Weak evidance,I know, but if you look you can probably find something.
Does anybody bother to read the entire post anymore? I had a few questions. That might deserve an offtopic (fine if you don't want to answer my question) but I doubt it deserves a flamebait. Or were you just reading the first two lines?
Yea, but then you miss a lot of the good posts from ACs who have a good point. A lot of the pro MS stuff gets posted by ACs and /. moderators rarely bother to mark them up. Thats why I keep saying that slashdot needs to retool its moderation system a bit.
How is this any different from Microsoft who blames all of Window's problems on the user fuqing up?
These OSs become mainstream when people other than sysadmins and nerds use them. When they are preinstalled at CompUSA is when they become mainstream. It's what people "mainly" use, hence the name.
Well, either way, it's slow. Thats why I keep saying network transparency should be at a higher level, but I wasn't around back when X was designed. And XFree 4.0 still isn't nearly as fast as BeOS even though it has supposedly no performance hit on local machines.
Actually BeOS runs graphics and UI in a seperate server too, and I've never had any speed issues with it (aside form the fact that it makes me cry to go back into windows to do something.)
Actually, BeOS uses gcc was well. RMS isn't on Be's case.
I'm tired of all you people saying that faster procs are useless because of other bottlenecks. I got news for you, in most non-server environments the proc is still the biggest bottle neck. To tell the truth, I enjoyed much more the 50% boost from 200 to 300 MHz than I did the 50% boost from a 66 to a 100MHz bus. For the most part apps are still computer bound, EXCEPT in server space. Thats why the Xeon is still chugging along at 550MHz. Examples of apps that are compute bound.
1) 3D, games, rendering,modling, you name it.
2) Any kind of realtime graphics.
3) Photoshop type apps depending on wether you use filters more often or just edits to large files.
4) Compiling.
5) Audio editing
6) Real-time video editing. (What, I have to wait 2 minutes to render the changes!)
Things not compute bound
1) Serving webpages, files, etc.
2) Working with large photoshop files.
3) Some types of scientific computing where data crunching is high volume, low workload.
Yes I've forgotten a few tihngs, but the market for more CPU power is clearly more important than the market for higher bandwidth.
1) If you talking pgcc, it doesn't come standard with most distros. If you think I am going to recompile every app on my system with pgcc, you've lost it.
2) Correct, XFree4.0 is pretty good. But he is talking BeOS level here. Even XFree4.0 is nowhere near the performance you get with BeOS. (Lets put it this way. On the SDL demos, BeOS is about 25% faster that Windows95 without hardware acceleration (SDL for BeOS does not support hardware acceleration.)Now Windows is pretty fast, especially considering that SDL for Windows uses DirectDraw (though I ran the test without a hardware blitter.) Now you go into BDirectWindow an you leave everything else in the dust.
4) Actually, the important Linux GUIs, Qt and GTK+ are not thread safe. Qt is very not thead safe and GTK+ requires a bit of a hack to be thread safe.
5) Well, no-one ever said that Unicies are good at threading. Linux is worse than WindowsNT at threads (in management, switch time, and spawn time) and BeOS is about 10 times faster than NT. You do that math.
6) Yea, and if I use TWM I'll have to burn my eyeballs out. If you want speed you can use DOS, but no sane person would.
7) Yea BeOS is free. Maybe not OpenSource, but free. It is a very open system, has some open source parts and in general the guys at be are pretty good with the source. Now I would rather use a fast, mostly closed system, than a slower OSS system.
Actually, that is a pretty stupid reason to call it GNU/Linux. There is a lot more to Linux than tar and grep. BeOS uses the GNU suite too. Does that make it GNU/BeOS? Hell no. Face it, the GNU stuff is pretty minor these days. Back in the day, maybe tar was worth something. These days tar and cp aren't exactly a major part of the OS, (unlike GNOME, or X, or KDE, or OSS, ad nauseum.)
Actually a fairly functional BeOS gets down to 2 meg. Quite a bit bigger than QNX, but still not horribaly bloated.
There is a difference between working and advanced. A laser sight rifle is really advanced, but a bazooka is a lot more destructive. I was talking about advanced in the sense of newer software design. However, the BeOS is microkernel done right, and is a big improvement on monolithic kernels. I have tried restarting the different servers. Aside from the app server (which can't be restarted becase it has to be there to run the script) most of the other servers can be restarted. In fact, most changes in options are done this way. Change network settings? Change it and the system reboots the net server. Same thing for the media server. And BeOS does not run a gigantic server process, there are quite a few independant ones, and they are all modular. This comes in the form of stability, scalability, and flexibility. A badly implemented server can't take the kernel out, (badly written drivers tend to do this in NT), the server load can be distributed over many procs both on a process level and on a thread level (a BeOS system has over 100 threads going at bootup.) and flexibility comes from the ability to modify or replace those servers on the fly.
Opps. Sorry, Haven't been keeping track of the OSS modules lately. However, the vortex drivers are still in beta.
I'm not talking about the OSS community. I'm talking about all the commercial developers using Linux for their products. You can bet if FreeBSD had as much developer support they would use that instead.
Is it a little disturbing to you that this thing has the Linux icon? Sure I can kind of understand that this is the FreeBe version FOR Linux, but wouldn't the Be icon be more appropriate as its is BEOS for linux? Or is the author trying to take advantage of the fact that an article filed under linux has a much better chance of getting in? (Actually I'm not trolling. You wouldn't believe the alternative OS stuff that gets ignored on /. That neverwinter nights thing really pissed me off (where neverwinter nights for Linux was the only one posted on /. even though the BeOS version had been announced a month before.))