This sounds like the kind of thing a journalist would make up on April 1st. Or it's the kind of kludge a somewhat irresponsible sysadmin might put in place as a joke. It is not a serious or useful approach to security, however.
Still, it would be enormously funny if one of the largest E-mail providers would actually do such a thing, as well as the consequences. "Medireview" indeed. Apparently, Yahoo! programmers don't even know about/\beval\b/. It's under "perldoc perlre".
Numerically, the number of Linux installations is comparable to that of MacOS, but the majority of those by far are server installations.
And how would you know? We can count Macintosh and Linux Internet servers and we can count Macintosh desktop systems (from Apple sales). But there is no way to count Linux desktop users. In fact, even many Linux servers are used for running desktop applications anyway, using Windows and Macintosh as displays. Linux is widely used at universities around the world for workstations. The Linux desktop has been adopted as the standard by Sun and HP. RedFlag will probably be used by millions of desktop users in China alone. Desktop oriented Linux distributions have a large marketshare in many countries. And Gnome and KDE are both easy-to-use, robust, modern desktops by any measure.
I just find this antagonism of the Macintosh community towards things Linux on the desktop fascinating. What have you got to prove? Linux isn't the enemy. The more Linux gets adopted on the desktop, the better for Apple because, unlike Microsoft, the Linux community doesn't hide behind proprietary APIs or patents.
You know, regarding the "moderation", I own several Macs myself and I like them. But the zealotry of Mac users is really annoying at times. Come on, guys, face the facts. Linux is a major desktop platform, with user numbers comparable to those of MacOS. And much of that wonderful software that makes MacOSX such a nice platform comes from that community.
it's *extremely* presumptious to casually refer to Linux along with Windoze and MacOS as one of "the three major platforms"
There are millions of Linux users, probably more in absolute numbers than MacOSX, if not Macintosh as a whole. An Linux is very widely used among unversity students, who play lots of games and like getting them free, too.
- so far as I know, less than one percent of all commercial video games are launched on linux.
So what? The fact that people don't buy a lot of games (or software) for Linux doesn't make it a "minor platform", it only makes it a "minor platform" for commercial game developers. In fact, many commercial games are based on ideas from old, free games built at universities on top of UNIX.
As long as they don't try to enforce it, it doesn't matter. They may well realize that the patent is bogus and has plenty of prior art. Lots of companies have lots of bogus patents that someone in their organization applied for, unaware of prior art.
Of course, unlike trademarks, the risk that they will try to enforce it remains throught the life of the patent. However, if it really worries you, you can have the patent reexamined or get a declaratory judgement.
I don't think the people of American now or ever have been interested in being relevant to international affairs. Nor do I think we want to be "predominant." We seem to be simply interested in getting rich,
Americans want to get cheap oil, cheap third world labor, easy access to world capital, easy access to foreign markets, etc. Many other nations don't think that those are particularly good ideas. It is this burning desire to be materially rich that causes Americans to try to influence politics around the globe. And without that kind of influence, Americans would be a lot poorer.
I'm sure most Americans would be happy if Europe would solve the Middle East crisis
That's a good example. Compared to other conflicts, the Israeli/Palestinian conflict is a small one. Europeans, left to their own devices, would periodically condemn the actions of both sides, send humanitarian aid to both sides, and otherwise not meddle too much. Let them work it out themselves.
It's only a "crisis" because Americans are, on the one hand, politically deeply committed to Israel, and on the other hand deathly afraid of losing Middle East oil imports from Arab states because they aren't willing to reduce energy usage.
In fact, inflatable structures could be great for all sorts of space applications. In the absence of gravity or wind, you really don't need much in the way of structural support. Inflatable structures can give you a huge volume with very little weight. And, yes, they can be made safe against puncture by space debris--probably safer than rigid structures.
Unfortunately, even though an inflatable module was considered for the ISS, it was not built. Pretty much all our space engineering seems to be done in terms of big, heavy, metal structures.
Interest seems to be picking up, though. There has been a workshop at ESA recently.
I found my mother had no problems using Linux, but it was difficult for her to get things like DSL service, non-technical books at various levels, and peripherals.
So, she now has a Mac running OSX. It's roughly as stable as Linux. It's about as easy to use as Gnome or KDE (not worse but not better either), and a lot nicer than Windows. If there is one thing that's worse it's that my mother finds a lot less software for the Mac that she likes than for Linux.
On the other hand, she can now go out and buy a piece of hardware or software, asking for something that is "Mac OSX compatible" and she can get books that are aimed at non-technical users. Also, the Apple brand name stands for pretty consistently decent hardware, whereas with PCs, finding good hardware is a gamble even if you buy a brand name.
So, consider getting your mother a Macintosh. Technically, it is really no better and no worse than Linux, but Apple's market presence and the support infrastructure around it makes it useful for non-technical users. As long as they remain mostly UNIX/Linux-compatible and don't do something really stupid in their relationship with the open source community, I think they are a decent choice.
If they were all good, why would it matter? However, I think there there is a best browser Mozilla or Galeon (which use the same rendering engine). It is by far the most standards compliant one.
2. Prompting for a filesystem scan.
You can easily fix this by adding "-y" to/etc/init.d/checkfs.sh. Traditionally, that was considered bad because if some inode was broken, someone would go in and hack the file system manually. These days, that's illusory. If fsck doesn't fix it properly, you need to restore from backup. So, I agree that this is bad, and it's easy to fix.
3. Printing needs to be easier to configure.
There are a variety of printer configuration programs that help you set up printers. Desktops should include something better. The main problem I see with printing is that it still wedgess.
Note that both Windows and MacOS printing and printer setup are also very rough around the edges. The only case that works smoothly most of the time seems to be installing a printer locally that either came with a disk or is completely standard.
4. Make it easy for the user to find out how to do things
Yes, I agree 100%.
5. Cleaner redraws.
At fault seem to be the Gtk+, Mozilla, and Qt toolkits. Mozilla and Qt were apparently written from the outset with a cross-platform mindset, where X11 redraw logic wasn't their primary consideration, and Gtk+ was apparently written trying to "insulate" developers from some tricky but important X11 functionality. X11 might benefit from adding some additional, small features (clear-after-delay, backing-store-during-move, etc.) to help with cleaner updates. However, if the toolkits aren't going to use them, what's the point?
You can get completely clean updates by setting backing store. On modern hardware, that is perhaps acceptable (it isn't on small machines). That should probably be an option.
6. Die stray processes, die!
Linux desktops should include a "process killer" application, accessible through a secure attention key, like Windows. Unlike Windows, it should have more intelligence about showing you processes likely at fault. Also, servers (print server, etc.), should be properly "nannied" so that they get restarted if they are killed, but that they also get suspended or killed automatically if they misbehave. That's quite common for server installations.
7. Easy way of sharing files.
Yes, Linux desktops should include a GUI for this. Traditionally, people consider this a sys admin task, and the sys admin GUIs are pretty good.
A fairly simple way of dealing with this would be to standardize on "public_html", "public_ftp", and "public_nfs" subdirectories in the home directory, with nothing to enable or disable.
8. Sound support.
This is a symptom of a deeper problem: dynamically loadable driver support in Linux sucks. Everybody I know ends up having to recompile the kernel, or having someone to recompile the kernel for them, if they want things like sound, APM, etc. to work.
9. No common editor which supports "soft wrapping."
Some of the GUI editors that come with desktops do this. However, it's not clear that it's a good thing.
10. No easy way to configure X - especially change resolution on the fly.
You can change X resolutions on the fly: have a look at "xvidtune". Also, many games change the resolution on the fly, and back again when they are done. So, all that is really missing is a better GUI.
A hash collision in a ReiserFS directory (where two filenames hash out to the same value) causes the older file to BE OVERWRITTEN without so much as a warning.
This is not necessarily a bug if the probability of that happening in real world scenarios is negligible. After all, you risk data loss from many sources.
Unfortunately, programmers often seem a bit unreasonable about probabilities. They complain about a (say) 1:10^20 chance of losing a file, while at the same time writing the whole file system in C, which basically guarantees a several-fold increase in the probability of undetected software faults compared to alternatives. In fact, the fix for such a remote possibility may not only kill performance, it may actually increase the overall probability of a fault that causes data loss--because the extra code may have bugs.
So, no, this doesn't bother me. I suspect that if Reiser knows about it and he isn't fixing it, he probably thought about it and decided the probability is too remote. If you disagree, I would like to see a more detailed analysis from you.
The US came into its current position of power through historical accident. Its rich resources, secure food supply, and geographic isolation allowed the US to become an influential power in WWII. Afterwards, it filled the power vacuum that was left by the self-destruction of Europe, and it managed to attract huge numbers of skilled immigrants from the rubble of Europe, which helped the US achieve technological and scientific predominance.
There is no reason to believe that this is inevitably a long-term state. The US is a mid-size country (by population), and food, geographic isolation, and natural resources are becoming less and less important. And other countries are becoming as attractive as the US for skilled international workers.
If the US continues to have a leadership role, it will be because it earns it. But that means that US politicians have to give up on their assumption that US predominance is a right that Americans are born with. Isolationist policies like those we have seen over the last few years will likely simply make the US less and less relevant to international affairs.
OS5 will include native SSL 3.0/TLS 1.0, support for systemwide encryption,
Applications that need that already have it built-in. This makes no difference to users, at least in the short term.
a doubling of screen resolution,
There are already plenty of PalmOS devices with doubled and higher screen resolution, and applications support them. It's convenient for developers that there is a single API now, but that makes no difference to users.
better support for web browsing, native support for 802.11b,
If you want those features, they are already available and they work. Again, no difference to users.
and the ability to finally use multitasking and multithreading applications.
That is not one of the features Palm lists on their web site.
I maintain: PalmOS 5 is a non-event as far as users are concerned, and mostly as far as developers are concerned as well. All indications are that for most users, it will work no differently from PalmOS 4.
Palm devices will continue to be a great way to store and manage you personal data on the road.
Did I ever say they were not? All I'm saying is that you don't need a PalmOS 5 device for that--stick with PalmOS 4 devices--they are cheaper and they work for those applications.
It really is. Palladium will interfere with Windows software development, it will restrict the availability of Windows software development tools, it will destroy the availability of shareware, and it will be a complete nuisance to developers. It will also be a big distraction to Microsoft software development. You couldn't design something more likely to make Windows even less usable and drive people to open source operating systems and other platforms if you tried.
What Palladium will not succeed at is kill off the competition. If Intel were foolish enough to make code signing mandatory, there is plenty of non-Intel hardware that won't have these mechanisms built in, and there will continue to be because without such hardware, out world would come to screeching halt.
And what Palladium won't be either is a magic bullet for security problems. Those are still human problems, and they still need to be fixed one at a time.
What Palladium isn't either is novel. These ideas have been kicking around for a long time and nobody has been foolish enough to implement them. Microsoft is continuing their habit for taking old, discarded ideas and shoving them into Windows; Windows is quickly becoming the dustbin of history for discarded ideas in computer science.
You can already run the emulator if you really care.
But why would you? PalmOS 5 on ARM will run applications in interpreted 68k code; only the kernel will run in native ARM code (and maybe a few assembly routines you hand-code yourself). That means it will likely not run faster. It also doesn't look like there is going to be a lot of new functionality. Altogether, I don't see much reason to hold my breath for these devices. In terms of nifty hardware, the Sony's are hard too beat (although they are not the most reliable).
When is Apple going to come out with machines based on new processors? The benchmarks are beginning to look pretty mediocre (and that seems to agree with the real performance I see as well).
They are easily distinguishable even by the blind, by size, texture, thickness, and weight. The color derives from the metals used, and that choice is driven largely by other considerations.
The whole multi-window application thing bothers me on X. On Windows or Mac, a dialog for an app stays in from of the app.
This is not a problem with X11, it's a design choice of the GUI or desktop software. X11 provides the mechanisms to tie dialog boxes to applications, but it's up to your application to decide to tie them and up to your window manager to do it or not to do it. And I think it's a matter of taste. Personally, I find the Windows behavior extremely annoying, as do most traditional X11 users (but, then, I think that dialog boxes are almost always the wrong UI element to use anyway).
Just yesterday, Mozilla 1.0.0 hosed X 4.2.0 on ATI (Radeon) hardware. It was font-related
I haven't seen this happen, but of course, any system as large as XFree86 will have bugs. But when such problems happen, they don't crash the operating system. Something like this happens with some regularity to me on Windows with IE, and it requires a reboot, not just killing the browser. And, of course, you have several other choices for font servers if xfs is giving you trouble.
My overall point is: X11 is a good layer to build a GUI on: it is fast, it is small, it is very modular, and it is generally quite reliable. People may like or dislike Gnome or KDE, but don't blame X11 for their bulk.
Compared to the rest of Windows, the NT kernel seems reasonably well engineered. The problem I think is that the end product is a combination of features that marketing thinks really need to go in there for their feature check lists, and pet ideas of the developers/researchers.
UNIX and Linux are different. UNIX (at least Research UNIX) was constrained by its paradigms: it was vigorously policed by its developers. For Linux, something doesn't make it into the kernel unless it really scratches an itch that a lot of people have--the feedback is immediate and direct: no interest, no developers.
Microsoft software development doesn't operate in a competitive market of ideas (let alone a competitive market), it doesn't have a paradigm to focus it, and it doesn't even have resource constraints to focus it. It's nice that they make the software engineering work out, but the end result still is mediocre at best.
First of all, you are missing the point: Gnome and KDE are resource intensive. Everybody agrees there. But X11 is not. X11 scales from tiny handhelds and wrist watches to high-end engineering workstations. If you want a desktop with a smaller footprint than Gnome or KDE, you can build it on top of X11--it gives you the choice.
But your comparison is ridiculous for another reason. Windows NT on a Pentium 90 box is as different from Windows XP on a modern machine as CDE on an old IBM workstation is from Gnome/KDE on today's Linux boxes. The fact that Microsoft marketing calls both of them "Windows" and that they share APIs doesn't change this basic fact.
Gnome/KDE are designed for current machines. That's why they use a lot of resources. The same is true for current generation Windows. And when we actually look, lo and behold, the Windows (and MacOS) desktop environments are just as big and just as resource intensive as currently popular X11-based desktop environments.
However, unlike Windows, Linux gives still gives you the choice of running the older desktops, even if you are using the latest versions of the kernel, OS, and X11 server. And that's really great. In fact, I see no reason really to run Gnome/KDE. Even if you want an "integrated desktop environment", there are more lightweight choices (I think XFCE is pretty neat).
The mantra is: Optimize for the common case
You put this in bold face, so you must think it's important. What are you actually trying to say?
Thank you for pointing that out, since it further supports my point: XP GUI desktop apps are generally no more lightweight than equivalent X11 GUI desktop apps.
Re:his X11 claims are completely bogus
on
A Linux User Goes Back
·
· Score: 3, Insightful
Compared to, say, Windows or MacOSX.
On my Linux machine (running Mozilla and a window manager), the X server process is 11Mbytes big (all numbers are RSS because that's what matters). That includes the frame buffer, I/O ranges, off-screen buffers, etc. The MacOSX window server on my Mac is 28Mbytes big. MS Windows won't tell you the answer as easily, but if you total up all the GDI-related DLLs and memory, it's big.
Applications don't fare much better. Even with Microsoft's DLL-hiding tricks, Windows applications are big. Quicken starts up a 28Mbyte process at boot time just to make itself appear to load fast, and Microsoft applications do similar things. A MacOSX terminal window application is 5.5Mbytes, X11's xvt is 1Mbyte, and xterm (with a full Tektronix emulator) is 2.2Mbytes. Using a more space efficient toolkit, you could get that down to under 100kbytes (embedded systems do this). MacOSX's simple mail client is 6.3Mbytes (with no mail loaded), something comparable like spruce or althea is 3Mbytes.
Now, unlike those other systems, you can configure X11 to be much smaller by reducing the amount of off-screen buffering it provides and other options. Remember: people used to run X11 on the state-of-the-art workstations of 15 years ago, which means machines that have less power and less memory than a Palm handheld today. X11 does scale down nicely, and even in its common configuration, which allows it to use lots of memory, it is small compared to the size of the desktop software itself.
People see that Gnome, KDE, and Mozilla are sluggish, hang on occasion, drop them into the command line, and are big and they blame X11. This is wrong.
XFree86 is very fast. XFree86 is far more stable than either the Windows or Macintosh GUI. XFree86, as well as MIT X11, also are tiny and can be configured to run in around 1.5Mbytes and will live happily on a 66MHz handheld. Try that with any of the other window systems. X11 font installation doesn't have to suck either--if it does, it's because the desktop you are using lacks the right utilities.
XFree86 is also the most stable window system I have used. I use XFree86 with ATI, NVIDIA, 3Dfx, and some frame buffers. It certainly doesn't "crash Linux" (it's just a user process), and I can't remember when I got the last crash (with some newly released, proprietary NVIDIA driver at that). In contrast, I have seen my share of blue screens on both OSX and Windows, even though I use them less and even though they came preconfigured. When a desktop running X11 appears to crash, usually what happened is that the desktop manager of the desktop you are running died and the desktop doesn't handle that case (this happens with some regularity under Windows and MacOS, but they just quietly restart it).
Run X11 with twm, icewm, or blackbox, and use only applications with toolkit actually written for X11, and you'll see that it is really fast. Look at X11 on a handheld and you'll see that it can be really small and fast.
X11's supposed X11 overhead really is overhead that comes from toolkits and applications that were not really written for X11. Most of the major X11 toolkits aren't X11 toolkits at all, they are cross-platform toolkits with an MS Windows orientation: Qt, Mozilla, FLTK, and wxWindows, were designed as cross-platform toolkits and Gtk+ might as well be. People wrote those with a local frame-buffer API in mind, and it's not surprising that their performance under X11, whose APIs are very different, is less than optimal. Gnome and KDE are also not using the X11 IPC mechanisms or the X11 resource mechanisms, instead substituting their own inefficient and less functional versions. It's not surprising that applications are slow if they need to talk to an object broker; if, instead, they communicate with each other through X11 properties, they are zippier and they work correctly over the network.
Of course, the user doesn't care why the Linux GUI is big or why it appears to crash, and these issues do need to get addressed. But they need to get addressed where they are being caused. Replacing X11 with a frame buffer system will not fix anything, it will just waste many man-years. If you want a faster and more reliable Linux desktop, either Qt, KDE, Gtk+, and Gnome need to shape up, or you need to use another desktop. While it still uses a fairly inefficient toolkit, XFCE is already a great improvement over those other desktops in terms of performance.
1) frustration with graphics in general (both performance and fonts)
I run X11 on NVidia, ATI, 3Dfx, and some handhelds. It is stable like a rock, small, lightning fast, and it doesn't crash, either itself or Linux.
KDE, Mozilla, and Gnome can be slow, and some misbehaved applications that don't use mouse grabs properly can make X11 appear to "crash" (it's really working fine, you just need to kill the application--happens under OSX and Windows as well).
Those are not X11's problems, they are problems with the toolkits that those systems use. Switching to a frame-buffer based system is not going to fix those problems with the applications.
Slashdot Mirror
Still, it would be enormously funny if one of the largest E-mail providers would actually do such a thing, as well as the consequences. "Medireview" indeed. Apparently, Yahoo! programmers don't even know about /\beval\b/. It's under "perldoc perlre".
And how would you know? We can count Macintosh and Linux Internet servers and we can count Macintosh desktop systems (from Apple sales). But there is no way to count Linux desktop users. In fact, even many Linux servers are used for running desktop applications anyway, using Windows and Macintosh as displays. Linux is widely used at universities around the world for workstations. The Linux desktop has been adopted as the standard by Sun and HP. RedFlag will probably be used by millions of desktop users in China alone. Desktop oriented Linux distributions have a large marketshare in many countries. And Gnome and KDE are both easy-to-use, robust, modern desktops by any measure.
I just find this antagonism of the Macintosh community towards things Linux on the desktop fascinating. What have you got to prove? Linux isn't the enemy. The more Linux gets adopted on the desktop, the better for Apple because, unlike Microsoft, the Linux community doesn't hide behind proprietary APIs or patents.
You know, regarding the "moderation", I own several Macs myself and I like them. But the zealotry of Mac users is really annoying at times. Come on, guys, face the facts. Linux is a major desktop platform, with user numbers comparable to those of MacOS. And much of that wonderful software that makes MacOSX such a nice platform comes from that community.
There are millions of Linux users, probably more in absolute numbers than MacOSX, if not Macintosh as a whole. An Linux is very widely used among unversity students, who play lots of games and like getting them free, too.
- so far as I know, less than one percent of all commercial video games are launched on linux.
So what? The fact that people don't buy a lot of games (or software) for Linux doesn't make it a "minor platform", it only makes it a "minor platform" for commercial game developers. In fact, many commercial games are based on ideas from old, free games built at universities on top of UNIX.
Of course, unlike trademarks, the risk that they will try to enforce it remains throught the life of the patent. However, if it really worries you, you can have the patent reexamined or get a declaratory judgement.
Americans want to get cheap oil, cheap third world labor, easy access to world capital, easy access to foreign markets, etc. Many other nations don't think that those are particularly good ideas. It is this burning desire to be materially rich that causes Americans to try to influence politics around the globe. And without that kind of influence, Americans would be a lot poorer.
I'm sure most Americans would be happy if Europe would solve the Middle East crisis
That's a good example. Compared to other conflicts, the Israeli/Palestinian conflict is a small one. Europeans, left to their own devices, would periodically condemn the actions of both sides, send humanitarian aid to both sides, and otherwise not meddle too much. Let them work it out themselves.
It's only a "crisis" because Americans are, on the one hand, politically deeply committed to Israel, and on the other hand deathly afraid of losing Middle East oil imports from Arab states because they aren't willing to reduce energy usage.
Unfortunately, even though an inflatable module was considered for the ISS, it was not built. Pretty much all our space engineering seems to be done in terms of big, heavy, metal structures.
Interest seems to be picking up, though. There has been a workshop at ESA recently.
So, she now has a Mac running OSX. It's roughly as stable as Linux. It's about as easy to use as Gnome or KDE (not worse but not better either), and a lot nicer than Windows. If there is one thing that's worse it's that my mother finds a lot less software for the Mac that she likes than for Linux.
On the other hand, she can now go out and buy a piece of hardware or software, asking for something that is "Mac OSX compatible" and she can get books that are aimed at non-technical users. Also, the Apple brand name stands for pretty consistently decent hardware, whereas with PCs, finding good hardware is a gamble even if you buy a brand name.
So, consider getting your mother a Macintosh. Technically, it is really no better and no worse than Linux, but Apple's market presence and the support infrastructure around it makes it useful for non-technical users. As long as they remain mostly UNIX/Linux-compatible and don't do something really stupid in their relationship with the open source community, I think they are a decent choice.
If they were all good, why would it matter? However, I think there there is a best browser Mozilla or Galeon (which use the same rendering engine). It is by far the most standards compliant one.
2. Prompting for a filesystem scan.
You can easily fix this by adding "-y" to /etc/init.d/checkfs.sh. Traditionally, that was considered bad because if some inode was broken, someone would go in and hack the file system manually. These days, that's illusory. If fsck doesn't fix it properly, you need to restore from backup. So, I agree that this is bad, and it's easy to fix.
3. Printing needs to be easier to configure.
There are a variety of printer configuration programs that help you set up printers. Desktops should include something better. The main problem I see with printing is that it still wedgess.
Note that both Windows and MacOS printing and printer setup are also very rough around the edges. The only case that works smoothly most of the time seems to be installing a printer locally that either came with a disk or is completely standard.
4. Make it easy for the user to find out how to do things
Yes, I agree 100%.
5. Cleaner redraws.
At fault seem to be the Gtk+, Mozilla, and Qt toolkits. Mozilla and Qt were apparently written from the outset with a cross-platform mindset, where X11 redraw logic wasn't their primary consideration, and Gtk+ was apparently written trying to "insulate" developers from some tricky but important X11 functionality. X11 might benefit from adding some additional, small features (clear-after-delay, backing-store-during-move, etc.) to help with cleaner updates. However, if the toolkits aren't going to use them, what's the point?
You can get completely clean updates by setting backing store. On modern hardware, that is perhaps acceptable (it isn't on small machines). That should probably be an option.
6. Die stray processes, die!
Linux desktops should include a "process killer" application, accessible through a secure attention key, like Windows. Unlike Windows, it should have more intelligence about showing you processes likely at fault. Also, servers (print server, etc.), should be properly "nannied" so that they get restarted if they are killed, but that they also get suspended or killed automatically if they misbehave. That's quite common for server installations.
7. Easy way of sharing files.
Yes, Linux desktops should include a GUI for this. Traditionally, people consider this a sys admin task, and the sys admin GUIs are pretty good.
A fairly simple way of dealing with this would be to standardize on "public_html", "public_ftp", and "public_nfs" subdirectories in the home directory, with nothing to enable or disable.
8. Sound support.
This is a symptom of a deeper problem: dynamically loadable driver support in Linux sucks. Everybody I know ends up having to recompile the kernel, or having someone to recompile the kernel for them, if they want things like sound, APM, etc. to work.
9. No common editor which supports "soft wrapping."
Some of the GUI editors that come with desktops do this. However, it's not clear that it's a good thing.
10. No easy way to configure X - especially change resolution on the fly.
You can change X resolutions on the fly: have a look at "xvidtune". Also, many games change the resolution on the fly, and back again when they are done. So, all that is really missing is a better GUI.
This is not necessarily a bug if the probability of that happening in real world scenarios is negligible. After all, you risk data loss from many sources.
Unfortunately, programmers often seem a bit unreasonable about probabilities. They complain about a (say) 1:10^20 chance of losing a file, while at the same time writing the whole file system in C, which basically guarantees a several-fold increase in the probability of undetected software faults compared to alternatives. In fact, the fix for such a remote possibility may not only kill performance, it may actually increase the overall probability of a fault that causes data loss--because the extra code may have bugs.
So, no, this doesn't bother me. I suspect that if Reiser knows about it and he isn't fixing it, he probably thought about it and decided the probability is too remote. If you disagree, I would like to see a more detailed analysis from you.
There is no reason to believe that this is inevitably a long-term state. The US is a mid-size country (by population), and food, geographic isolation, and natural resources are becoming less and less important. And other countries are becoming as attractive as the US for skilled international workers.
If the US continues to have a leadership role, it will be because it earns it. But that means that US politicians have to give up on their assumption that US predominance is a right that Americans are born with. Isolationist policies like those we have seen over the last few years will likely simply make the US less and less relevant to international affairs.
Applications that need that already have it built-in. This makes no difference to users, at least in the short term.
a doubling of screen resolution,
There are already plenty of PalmOS devices with doubled and higher screen resolution, and applications support them. It's convenient for developers that there is a single API now, but that makes no difference to users.
better support for web browsing, native support for 802.11b,
If you want those features, they are already available and they work. Again, no difference to users.
and the ability to finally use multitasking and multithreading applications.
That is not one of the features Palm lists on their web site.
I maintain: PalmOS 5 is a non-event as far as users are concerned, and mostly as far as developers are concerned as well. All indications are that for most users, it will work no differently from PalmOS 4.
Palm devices will continue to be a great way to store and manage you personal data on the road.
Did I ever say they were not? All I'm saying is that you don't need a PalmOS 5 device for that--stick with PalmOS 4 devices--they are cheaper and they work for those applications.
That's what I said.
so 68K programs that run on OS 5 devices will be quite a bit faster.
That seems unlikely. Look at how much time applications on other platforms spend in the kernel: it's not much.
What Palladium will not succeed at is kill off the competition. If Intel were foolish enough to make code signing mandatory, there is plenty of non-Intel hardware that won't have these mechanisms built in, and there will continue to be because without such hardware, out world would come to screeching halt.
And what Palladium won't be either is a magic bullet for security problems. Those are still human problems, and they still need to be fixed one at a time.
What Palladium isn't either is novel. These ideas have been kicking around for a long time and nobody has been foolish enough to implement them. Microsoft is continuing their habit for taking old, discarded ideas and shoving them into Windows; Windows is quickly becoming the dustbin of history for discarded ideas in computer science.
You can already run the emulator if you really care. But why would you? PalmOS 5 on ARM will run applications in interpreted 68k code; only the kernel will run in native ARM code (and maybe a few assembly routines you hand-code yourself). That means it will likely not run faster. It also doesn't look like there is going to be a lot of new functionality. Altogether, I don't see much reason to hold my breath for these devices. In terms of nifty hardware, the Sony's are hard too beat (although they are not the most reliable).
When is Apple going to come out with machines based on new processors? The benchmarks are beginning to look pretty mediocre (and that seems to agree with the real performance I see as well).
And for an OS without DRM:
Debian
They are easily distinguishable even by the blind, by size, texture, thickness, and weight. The color derives from the metals used, and that choice is driven largely by other considerations.
This is not a problem with X11, it's a design choice of the GUI or desktop software. X11 provides the mechanisms to tie dialog boxes to applications, but it's up to your application to decide to tie them and up to your window manager to do it or not to do it. And I think it's a matter of taste. Personally, I find the Windows behavior extremely annoying, as do most traditional X11 users (but, then, I think that dialog boxes are almost always the wrong UI element to use anyway).
Just yesterday, Mozilla 1.0.0 hosed X 4.2.0 on ATI (Radeon) hardware. It was font-related
I haven't seen this happen, but of course, any system as large as XFree86 will have bugs. But when such problems happen, they don't crash the operating system. Something like this happens with some regularity to me on Windows with IE, and it requires a reboot, not just killing the browser. And, of course, you have several other choices for font servers if xfs is giving you trouble.
My overall point is: X11 is a good layer to build a GUI on: it is fast, it is small, it is very modular, and it is generally quite reliable. People may like or dislike Gnome or KDE, but don't blame X11 for their bulk.
UNIX and Linux are different. UNIX (at least Research UNIX) was constrained by its paradigms: it was vigorously policed by its developers. For Linux, something doesn't make it into the kernel unless it really scratches an itch that a lot of people have--the feedback is immediate and direct: no interest, no developers.
Microsoft software development doesn't operate in a competitive market of ideas (let alone a competitive market), it doesn't have a paradigm to focus it, and it doesn't even have resource constraints to focus it. It's nice that they make the software engineering work out, but the end result still is mediocre at best.
But your comparison is ridiculous for another reason. Windows NT on a Pentium 90 box is as different from Windows XP on a modern machine as CDE on an old IBM workstation is from Gnome/KDE on today's Linux boxes. The fact that Microsoft marketing calls both of them "Windows" and that they share APIs doesn't change this basic fact.
Gnome/KDE are designed for current machines. That's why they use a lot of resources. The same is true for current generation Windows. And when we actually look, lo and behold, the Windows (and MacOS) desktop environments are just as big and just as resource intensive as currently popular X11-based desktop environments.
However, unlike Windows, Linux gives still gives you the choice of running the older desktops, even if you are using the latest versions of the kernel, OS, and X11 server. And that's really great. In fact, I see no reason really to run Gnome/KDE. Even if you want an "integrated desktop environment", there are more lightweight choices (I think XFCE is pretty neat).
The mantra is: Optimize for the common case
You put this in bold face, so you must think it's important. What are you actually trying to say?
Thank you for pointing that out, since it further supports my point: XP GUI desktop apps are generally no more lightweight than equivalent X11 GUI desktop apps.
On my Linux machine (running Mozilla and a window manager), the X server process is 11Mbytes big (all numbers are RSS because that's what matters). That includes the frame buffer, I/O ranges, off-screen buffers, etc. The MacOSX window server on my Mac is 28Mbytes big. MS Windows won't tell you the answer as easily, but if you total up all the GDI-related DLLs and memory, it's big.
Applications don't fare much better. Even with Microsoft's DLL-hiding tricks, Windows applications are big. Quicken starts up a 28Mbyte process at boot time just to make itself appear to load fast, and Microsoft applications do similar things. A MacOSX terminal window application is 5.5Mbytes, X11's xvt is 1Mbyte, and xterm (with a full Tektronix emulator) is 2.2Mbytes. Using a more space efficient toolkit, you could get that down to under 100kbytes (embedded systems do this). MacOSX's simple mail client is 6.3Mbytes (with no mail loaded), something comparable like spruce or althea is 3Mbytes.
Now, unlike those other systems, you can configure X11 to be much smaller by reducing the amount of off-screen buffering it provides and other options. Remember: people used to run X11 on the state-of-the-art workstations of 15 years ago, which means machines that have less power and less memory than a Palm handheld today. X11 does scale down nicely, and even in its common configuration, which allows it to use lots of memory, it is small compared to the size of the desktop software itself.
XFree86 is very fast. XFree86 is far more stable than either the Windows or Macintosh GUI. XFree86, as well as MIT X11, also are tiny and can be configured to run in around 1.5Mbytes and will live happily on a 66MHz handheld. Try that with any of the other window systems. X11 font installation doesn't have to suck either--if it does, it's because the desktop you are using lacks the right utilities.
XFree86 is also the most stable window system I have used. I use XFree86 with ATI, NVIDIA, 3Dfx, and some frame buffers. It certainly doesn't "crash Linux" (it's just a user process), and I can't remember when I got the last crash (with some newly released, proprietary NVIDIA driver at that). In contrast, I have seen my share of blue screens on both OSX and Windows, even though I use them less and even though they came preconfigured. When a desktop running X11 appears to crash, usually what happened is that the desktop manager of the desktop you are running died and the desktop doesn't handle that case (this happens with some regularity under Windows and MacOS, but they just quietly restart it).
Run X11 with twm, icewm, or blackbox, and use only applications with toolkit actually written for X11, and you'll see that it is really fast. Look at X11 on a handheld and you'll see that it can be really small and fast.
X11's supposed X11 overhead really is overhead that comes from toolkits and applications that were not really written for X11. Most of the major X11 toolkits aren't X11 toolkits at all, they are cross-platform toolkits with an MS Windows orientation: Qt, Mozilla, FLTK, and wxWindows, were designed as cross-platform toolkits and Gtk+ might as well be. People wrote those with a local frame-buffer API in mind, and it's not surprising that their performance under X11, whose APIs are very different, is less than optimal. Gnome and KDE are also not using the X11 IPC mechanisms or the X11 resource mechanisms, instead substituting their own inefficient and less functional versions. It's not surprising that applications are slow if they need to talk to an object broker; if, instead, they communicate with each other through X11 properties, they are zippier and they work correctly over the network.
Of course, the user doesn't care why the Linux GUI is big or why it appears to crash, and these issues do need to get addressed. But they need to get addressed where they are being caused. Replacing X11 with a frame buffer system will not fix anything, it will just waste many man-years. If you want a faster and more reliable Linux desktop, either Qt, KDE, Gtk+, and Gnome need to shape up, or you need to use another desktop. While it still uses a fairly inefficient toolkit, XFCE is already a great improvement over those other desktops in terms of performance.
I run X11 on NVidia, ATI, 3Dfx, and some handhelds. It is stable like a rock, small, lightning fast, and it doesn't crash, either itself or Linux.
KDE, Mozilla, and Gnome can be slow, and some misbehaved applications that don't use mouse grabs properly can make X11 appear to "crash" (it's really working fine, you just need to kill the application--happens under OSX and Windows as well).
Those are not X11's problems, they are problems with the toolkits that those systems use. Switching to a frame-buffer based system is not going to fix those problems with the applications.