It doesn't matter whether StarOffice is an adequate replacement for Microsoft Office or not, you can be certain that your users will complain. It's not really surprising either: imagine someone was forced to replace their Lincoln Towncar with a Toyota Camry. The Camry may be a reasonable replacement, but the person forced to switch won't like it and will become nostalgic about their wonderful old car.
So, try to figure out how to motivate people to use the new software. Maybe you can arrange for people to share in the financial rewards of the switch (a small raise for all the MS Office users, financed from the license savings). If people see and share in the financial benefits, that might motivate them. On the other hand, if they are forced by decree to use something they consider inferior, it's going to be a disaster.
Also consider introducing it gradually over the next year, requiring to use StarOffice for some peripheral business processes and getting people used to it without forcing them to switch cold.
I would probably go with StarOffice (as opposed to the free suites), though. That's not because StarOffice is necessarily better, it's because you can point out that this is a commercial program, developed and supported by a large software company. You probably don't want to fight the "switch from Office" and "switch to open source software" battles at the same time. Once your users accept StarOffice, you can then still switch to OpenOffice.
We have the sclerotic and bureaucratic republic, an empire run by some evil guys dressed in black, and a bunch of rebellious royals. I'm with Brin: Star Trek offers a more inspiring vision of the future.
Having extensions usually doesn't make a compiler non-compliant (GNU compilers usually also have a "strict ANSI mode"). Using extensions makes a program non-compliant.
I find the notion of a C++ program that "compiles best" under some compiler pretty odd. It may be true, but that's really only half the story.
The full story is something like "Program X uses a bunch of non-standard compiler extensions that only Compiler Y implements" or "Program X uses some really obscure standards-compliant feature that only Compiler Y implements correctly" or "Program X relies on the implementation-specific memory layout implemented by Compiler Y".
Since GNU C++ is quite good when it comes to standards compliance, I suspect that if a program "compiles best" under Watcom C++, it must be using some non-standard or unportable features. Does anybody know for sure?
The increased pace of security patches from MS may indicate that they're finally serious about security.
Statistically, a steady-state of a large number of patches means that there are many more problems waiting to be discovered. Only once the problems start drying up do we have some indication that they have made a dent.
I have grown to appreciate the way Windows XP patches itself.
I have both an XP and a Debian machine, and the way XP updates itself is horrible. Not only has it managed to completely hose my machine once already within a few months, requiring a complete reinstall, it also only updates the operating system and not the application. (Reinstalls still haven't gotten easier either--they wipe out almost everything.)
The Debian updater has worked like a charm for, oh, probably a year and a half now. It updates not only the OS but all applications. And since each package is identified, documented, separately maintained, and versioned, if you want to revert, it's easy.
XP auto updating is a gamble. And when something goes wrong you are basically stuck. And with XP, things do go wrong. As usual, Microsoft's copy of other people's ideas is inferior.
Distributed human processing is hardly new--it's the way science and engineering work. Trouble is: getting a bunch of human beings to solve ill-defined problems is just not the same as having a bunch of computers do a well-specified and verifiable task over and over again.
Let's say someone has a brilliant idea for waste water treatment. How is ThinkCycle going to test that that idea works? It can't just run a computation or ask a bunch of random people to verify the idea. You need to build a pilot and try it out. Well, the mechanisms for evaluating what ideas work and what ideas don't already exist, and they are already distributed: publications, peer review, libraries, conferences, symposia, citation statistics, recommendations, talks, etc.
The mechanisms by which you do cooperative problem solving already exist, and they have existed for hundreds of years. They are the mechanism by which we collaborate in science, technology, engineering, and the economy. And there has been very active research in supporting them with computers, through groupware, electronic communications, and many other means.
As for the site itself, it looks to me like a fairly regular groupware site. It's nice that someone set up a groupware site to discuss these topics. I find it somehwat ironic, though, that a site which writes "Open Source Design" on its banner has so many DOC and PPT files.
H.323 and associated protocols for video conferencing and collaboration have been standardized for a while. They are kind of messy, but there were Windows implementations like NetMeeting, Linux implementations like Open H.323, and commercial implementations like CU-SeeMe (for Windows and Mac). These things could even talk to one another and to GnomeMeeting.
Fast forward to 2002. Microsoft still kind of ships Netmeeting with Windows XP Home, but there are no shortcuts, their documentation discourages you from using it (it also blue-screened my XP machine when I tried running it). Instead, they want you to use Microsoft Messenger, which only seems to want to talk through Microsoft's servers. Yahoo! give you video conferencing, but only through Yahoo! messenger and only on Windows. CU-SeeMe doesn't seem to exist anymore. In fact, I couldn't find any Windows or OSX H.323 implementations.
Instead, now the next thing seems to be SIP (Session Initiation Protocol, which is curiously what Vovida is based on. Well, it's kind of like HTTP, and that's nice compared to H.323's ASN protocols. MSN Messenger seems to be using it. There is Linphone, which is SIP based and works on Linux.
But... how do we do cross platform video conferencing now? Microsoft Messenger may speak SIP, but as far as I can tell, it doesn't let me do machine to machine calls. Even if it did, GnomeMeeting doesn't seem to support SIP (yet?) and Linphone doesn't do video. And MacOSX, as far as I can tell, is almost completely out in the cold; at least, I couldn't find any commercial video conferencing software for it. The closest is the OpenH.323 sample applications, running under X11 on MacOSX. That's not exactly what you can ask average Mac users to use.
So, if I want to do cross-platform video conferencing between Linux, Windows, and/or Macintosh, what software and protocols should I use?
Except its not. For almost all cards (NVIDIA and its kernel driver nonwithstanding) the whole driver is in userspace, acessing the hardware via user-mapped I/O ports.
That's what I said: the video driver (i.e., on Linux, the thing that does the mapping and switching) is in the kernel. The drawing and acceleration is in the display server, and for certain special applications, the application itself gets memory mapped into its address space.
This isn't the ideal situation, because for the absolute best performance, you need some stuff in the kernel (which DRI does, but DRI support is rather limited and only for 3D).
That's a crappy design. Its more flexible, but its faster to have the toolkit server-side. That's why Qt (and GTK+) on X is slower than Qt on Windows.
Qt is slower on X11 than on Windows because Qt ignores most of the server-side facilities that X offers. The "crappy design" there is Qt, not X11, and it mostly means that the authors of Qt just didn't want to bother doing a high-quality X11 implementation: Windows apparently matters more to them. Furthermore, on Windows, the "toolkit" isn't server side either: the display server runs in the kernel, and Qt runs in user space.
In fact, with the right toolkit, X11 is often faster than GDI. The reason is that X11 was designed to go through a bottleneck. GDI was designed assuming direct library calls and had to get retrofitted to work in a protected mode environment. Furthermore, X11 naturally takes advantage of multiple processors and graphics processors.
Before, when basically everything was simple blits or pixel plotting, the latency of individual operations was critical.
I'm not sure what you mean by "before". Windows was written that way. X11 had a client/server architecture from the start and has always worked well with it. Windows is the latecomer, and it still isn't very good at it.
These days, with OpenGL serving as the support for the GUI (see Longhorn and Jaguar) clients have to package up command anyway (vertex buffers, display lists, etc) and each operation takes comparitvely longer than a single PutPixel(). Thus, the latency of the communication isn't as much of a factor anymore.
As I was saying, X11 got it right from the start because X11 didn't assume that any program can just bash pixels in the frame buffer. Windows is playing catch-up.
I'm a bit confused about what Linux interfaces they are planning on supporting. I mean, from the point of view of most application software, Solaris and Linux don't look all that different already. They both are POSIX compliant and they both support a lot of the BSD stuff. Most of the porting problems are byte order issues and idiosyncratic header differences.
So, what is Sun actually going to do? Are they going to make their headers 100% identical to Linux? What other interfaces are they going to support? Is there going to be a/proc and a devfs?
The biggest problem with X, no matter how wonderful it is, technically speaking, is that it does not enforce GUI semantics.
X11 is the equivalent of GDI or Quartz; it doesn't have to enforce GUI semantics. If you want to enforce a "coherent" desktop on top of it, you can impose whatever draconian measures you like. KDE looks quite coherent and standardized to me, for example.
It's a myth, in any case, that Windows or MacOS are any more coherent than, say, KDE. Take a look here for an extensive critique. And you think that the appearance or window management behavior can't be changed on Windows? Think again: Stardock, Litestep, Microsoft PowerToys.
but isn't the fact that video drivers are implemented in userland an architectural problem to begin with?
The video drivers are in the kernel. The drawing and acceleration is in the display server. The toolkit is in the application. It's fast and it's robust. It's what NeXTStep and MacOSX do as well. Where is the "architectural" problem?
Plus, the resources mechanism is absolutely byzantine and needs to be be razed,
Neither Gnome nor KDE use the X11 resource mechanism. They use something much more like Windows. That's actually a shame because the X11 resource mechanism is better.
as well the complex distinctions between server and client (wait, who's the server, who's the client, who has the toolkit?, who's running the window manager? what the fsck is going on?).
Windows, MacOS, and NeXTStep make the same distinction as X11: they have a low-level graphics and windowing component that runs in a display server, and they have a high level toolkit part that runs in a display client.
Altogether, it looks to me like you have a rather outdated notion of what Windows, MacOS, and X11 are. Windows and MacOS have pretty much become like X11 architecturally; they simply lack the well-defined and efficient X11 protocol to support that architecture. On the other hand, X11 toolkits (for better or worse) have become much more like Windows and MacOS toolkits. All three of them have gotten direct rendering and 3D acceleration.
I don't really understand those issues. If a Linux user does the same thing that a Windows user does, install a ready-made distribution, they don't ever have to deal with any of this. Modern Linux distributions configure the server and OpenGL, figure out the monitor, install a desktop, and install TrueType fonts. Even the DVD players work out of the box (you have to get a third party CSS decoder plug-in yourself for encrypted DVDs--thank the MPAA).
If you choose to compile and install things by hand, of course, you have to edit XF86Config files. But even there things have gotten much easier (e.g., in XF86v4, you don't need to worry about modelines anymore).
In joe-user mode, I want a standard look&feel across all my apps
X11 is the equivalent of GDI, Quartz, or DisplayPostscript. If you want to try to enforce a consistent user environment on top of it for your new OS, you can do that as much as you can do it on any other platform.
Conversely, Quartz and GDI fail to enforce a consistent UI either. Toolkits like Borland, FLTK, wxWindows/Universal, Qt, Swing, and OpenStep run on top of both of them and they give you applications that look and behave differently from native apps.
In different words, I think this is a red herring. If you build a new windowing system, for a while, things will be consistent because your own applications will be the only ones. Once it becomes popular, the consistency vanishes.
The East Coast is prone to earthquakes as well. Even New Jersey and New York. Have a look here. There are regions on the East Coast that are even more susceptible, and unlike the West Coast, the East Coast is almost completely unprepared for even a moderate earthquake.
please understand that the previous sentence does NOT mean that I will be using XFree86 as the engine (shudder), just that Cosmoe will leverage XFree86's existing drivers to drive Cosmoe's graphics engine.
What is supposed to be wrong with the X11 graphics engine? Why do people keep complaining about it?
X11 does hardware-accelerated 2D and 3D graphics with transparency, anti-aliased fonts and graphics, ClearType-like rendering, and server-side geometric transforms.
You can get a good X11 server into less than 1Mbyte with almost no off-screen memory, or you can give it hundreds of megabytes for caching, buffering, fonts, and textures.
X11 provides a uniform and powerful API for all sorts of input devices.
It can't be the client/server architecture that people complain about because neither Windows nor OSX do direct graphics I/O for their UI either. Neither can it be the footprint, because if you look at Windows or OSX GUI apps and system resources, they are as big as X11 on similar hardware, or bigger.
Graphic design can't be the problem either: X11 imposes no constraints on the toolkits, GUIs, or desktops you build on top of it, and X11 toolkits have emulated Windows, OS/2, and MacOS/OSX so closely that it is really hard to tell the difference.
So, what concretely is supposed to be wrong with X11? Why this visceral dislike? Why do people keep starting projects to replace X11 with unaccelerated display servers?
Of course some people will call that stupid, but look how popular it is!
How would anybody know? End users don't get asked whether they want Flash. it is forced upon them. End users usually can't even turn off the blink tag or GIF animations.
In fact, I would like Slashdot more if it had more pictures to help navigate, and some other font besides Times New Roman.
Slashdot has whatever font you choose as your default text font. Take a look in your browser's preference settings some time.
You forgot to mention that the U.S. is also the leading force for freedom and democracy in the world, and the most prosperous and moral civilization in human history.
Oh my god, not by a long shot. The US is clearly not the most prosperous civilization; Switzerland and Canada, among others, have a higher standard of living. US citizens may collectively command the largest amount of aggregate wealth right now, but that would be a rather odd-ball measure of prosperity.
The "most moral" by what measure? In terms of teenage pregnancies, legal system, percentage of population in prison, economic disparities, poverty, and race relations, the US is not particularly "moral". Neither is the vastly disproportionate use of resources or the per capita cost Americans impose on the environment. So, what sense of "most moral" were you thinking of?
In terms of supporting freedom and democracy in the world (and at home), that seems pretty iffy, too. Take a look at the book "Blowback", which chronicles some of the US foreign adventures. Arguably, the Marshall plan was good, but even there, the US was not single-handedly responsible for freedom and democracy in Europe after WWII.
The US isn't a bad country. In fact, on balance, it's doing pretty well. But your assessment seems way out of proportion with reality.
The primary attacks smart cards are designed to protect against are eavesdropping and replay. They can do that because they can run zero knowledge and public key protocols. That's a whole lot better than the magnetic strip on your credit card and is unaffected by this attack.
Protection against physical tampering is secondary. It's nice, but even if it didn't exist at all, smart cards would still be very useful. This particular attack seems so tricky that it may not even be worth doing anything about.
People in the third world see our affluence and freedoms on the new media and they compare it with their own lot. What are they supposed to do? For the most part, they don't have any democratic institutions, so they can't just vote their leaders out of office. If they overthrow their governments, it endangers their lives and risks the ire of international corporations and financial markets, making their economic situation even worse. If the acquiesce, nothing will change. Anger or depression seem the two most natural consequences.
It seems to me the ball is in our court: the only way I see global poverty is going to be addressed is if we help these other countries and if we voluntarily give up some of our power. Maybe we should let third world nations unilaterally dump textiles on our markets without forcing reciprocal, "fair" trading arrangements. Maybe we should let them "violate" our copyrights and patents, just like we violated European copyrights and patents during the first few hundred years of our history.
but the point is that the layout is still similar to everything that's come before.
Yes, and that's no accident: the size, layout, and feel of controls on current cars is driven by physics and engineering. You know roughly how big a manual transmission or steering wheel ought to be. There were a few designs that worked and were cost effective; anything else was really expensive or downright impossible.
Electronics already made things worse to some degree. But GUIs remove all those constraints and designers can do things in arbitrary ways. Users of a GUI have no guidance based on physical constraints what controls to expect or how to operate the car. You can't put 10000 physical buttons into a car, but you can put 10000 GUI buttons into a UI.
When cars were first being made, why didn't they use a "horse reigns"-like steering system, since that's what everybody was using before and thus must've been "intuitive"? Or why not a rudder lever like on a smaller boat, rather than the steering wheel from larger boats?
You wouldn't be able to turn the wheels at low speed without some kind of gearing; that's why rudders or reigns don't work well. The design of the steering wheel is driven by physics and ergonomics. Try coming up with something that works as well.
Did you just hop in the car and know how to do everything? Of course not, because it's not intuitive
Of course, I did, and you probably did, too. Someone tells you how to turn the thing on and get it into "drive", and after that, it's trivial: one pedal makes it go, the other makes it stop, and the steering wheel turns it.
Almost all driving instruction and learning is safety-related, not about how to move the car.
I think the vast majority of their audience will be viewing this site using IE on 2K, XP or OSX. Those users (90+ percent of users on servers I manage) will be presented with an extremely cool website.
Even with IE on XP, I found the site unintuitive and hard to navigate.
I ordered a Mini [miniusa.com] at our local BMW dealership a few months ago
I find the Mini site just as difficult to navigate and user-unfriendly. Yes, it's flashy, but try finding the specs: you have to sit through several animation sequences, the fonts don't work well on high resolution screens, and it opens windows all over the place.
Their web sites and their cars are strangely similar: very flashy and "cool", but something simpler and less flashy is probably more usable.
They mention that in 1953, the BWM 502 had 26 control and indicator functions. In the late 90's, the 7-series had over 70 functions, with as many indicators, and over 35 control elements (buttons, etc.) Something *had* to be done to reduce the complexity of the cockpit.
Adding a menu system probably doesn't reduce the complexity, distraction, or cognitive load. Maybe it would be better if the designers and marketers figured out what is essential and how to reduce the number of functions back down to 35. That would probably also make the car less distracting and safer.
It seems odd that the strangelet would generate only two earthquakes, one on entry one on exit. Why wouldn't it interact while traveling through the earth and encountering different materials?
Also, if the only evidence is two earthquakes separated by a few seconds, wouldn't we expect some of those merely by chance? What additional evidence is there for strangelets?
It's not clear that the problem making wireless more attractive for the last mile is the equipment cost. Everything helps, I suppose, but installation and ISP transmitter costs (antenna rental, Internet hookup) are likely to be more costly. See sprintbbd.com for one venture that has been put on hold.
Slashdot Mirror
So, try to figure out how to motivate people to use the new software. Maybe you can arrange for people to share in the financial rewards of the switch (a small raise for all the MS Office users, financed from the license savings). If people see and share in the financial benefits, that might motivate them. On the other hand, if they are forced by decree to use something they consider inferior, it's going to be a disaster.
Also consider introducing it gradually over the next year, requiring to use StarOffice for some peripheral business processes and getting people used to it without forcing them to switch cold.
I would probably go with StarOffice (as opposed to the free suites), though. That's not because StarOffice is necessarily better, it's because you can point out that this is a commercial program, developed and supported by a large software company. You probably don't want to fight the "switch from Office" and "switch to open source software" battles at the same time. Once your users accept StarOffice, you can then still switch to OpenOffice.
We have the sclerotic and bureaucratic republic, an empire run by some evil guys dressed in black, and a bunch of rebellious royals. I'm with Brin: Star Trek offers a more inspiring vision of the future.
Having extensions usually doesn't make a compiler non-compliant (GNU compilers usually also have a "strict ANSI mode"). Using extensions makes a program non-compliant.
The full story is something like "Program X uses a bunch of non-standard compiler extensions that only Compiler Y implements" or "Program X uses some really obscure standards-compliant feature that only Compiler Y implements correctly" or "Program X relies on the implementation-specific memory layout implemented by Compiler Y".
Since GNU C++ is quite good when it comes to standards compliance, I suspect that if a program "compiles best" under Watcom C++, it must be using some non-standard or unportable features. Does anybody know for sure?
Statistically, a steady-state of a large number of patches means that there are many more problems waiting to be discovered. Only once the problems start drying up do we have some indication that they have made a dent.
I have both an XP and a Debian machine, and the way XP updates itself is horrible. Not only has it managed to completely hose my machine once already within a few months, requiring a complete reinstall, it also only updates the operating system and not the application. (Reinstalls still haven't gotten easier either--they wipe out almost everything.)
The Debian updater has worked like a charm for, oh, probably a year and a half now. It updates not only the OS but all applications. And since each package is identified, documented, separately maintained, and versioned, if you want to revert, it's easy.
XP auto updating is a gamble. And when something goes wrong you are basically stuck. And with XP, things do go wrong. As usual, Microsoft's copy of other people's ideas is inferior.
Let's say someone has a brilliant idea for waste water treatment. How is ThinkCycle going to test that that idea works? It can't just run a computation or ask a bunch of random people to verify the idea. You need to build a pilot and try it out. Well, the mechanisms for evaluating what ideas work and what ideas don't already exist, and they are already distributed: publications, peer review, libraries, conferences, symposia, citation statistics, recommendations, talks, etc. The mechanisms by which you do cooperative problem solving already exist, and they have existed for hundreds of years. They are the mechanism by which we collaborate in science, technology, engineering, and the economy. And there has been very active research in supporting them with computers, through groupware, electronic communications, and many other means.
As for the site itself, it looks to me like a fairly regular groupware site. It's nice that someone set up a groupware site to discuss these topics. I find it somehwat ironic, though, that a site which writes "Open Source Design" on its banner has so many DOC and PPT files.
Sorry for being so completely out of it, but are "Reloaded" and "Revolution" two separate sequels?
Fast forward to 2002. Microsoft still kind of ships Netmeeting with Windows XP Home, but there are no shortcuts, their documentation discourages you from using it (it also blue-screened my XP machine when I tried running it). Instead, they want you to use Microsoft Messenger, which only seems to want to talk through Microsoft's servers. Yahoo! give you video conferencing, but only through Yahoo! messenger and only on Windows. CU-SeeMe doesn't seem to exist anymore. In fact, I couldn't find any Windows or OSX H.323 implementations.
Instead, now the next thing seems to be SIP (Session Initiation Protocol, which is curiously what Vovida is based on. Well, it's kind of like HTTP, and that's nice compared to H.323's ASN protocols. MSN Messenger seems to be using it. There is Linphone, which is SIP based and works on Linux.
But... how do we do cross platform video conferencing now? Microsoft Messenger may speak SIP, but as far as I can tell, it doesn't let me do machine to machine calls. Even if it did, GnomeMeeting doesn't seem to support SIP (yet?) and Linphone doesn't do video. And MacOSX, as far as I can tell, is almost completely out in the cold; at least, I couldn't find any commercial video conferencing software for it. The closest is the OpenH.323 sample applications, running under X11 on MacOSX. That's not exactly what you can ask average Mac users to use.
So, if I want to do cross-platform video conferencing between Linux, Windows, and/or Macintosh, what software and protocols should I use?
That's what I said: the video driver (i.e., on Linux, the thing that does the mapping and switching) is in the kernel. The drawing and acceleration is in the display server, and for certain special applications, the application itself gets memory mapped into its address space.
This isn't the ideal situation, because for the absolute best performance, you need some stuff in the kernel (which DRI does, but DRI support is rather limited and only for 3D).
That's a crappy design. Its more flexible, but its faster to have the toolkit server-side. That's why Qt (and GTK+) on X is slower than Qt on Windows.
Qt is slower on X11 than on Windows because Qt ignores most of the server-side facilities that X offers. The "crappy design" there is Qt, not X11, and it mostly means that the authors of Qt just didn't want to bother doing a high-quality X11 implementation: Windows apparently matters more to them. Furthermore, on Windows, the "toolkit" isn't server side either: the display server runs in the kernel, and Qt runs in user space.
In fact, with the right toolkit, X11 is often faster than GDI. The reason is that X11 was designed to go through a bottleneck. GDI was designed assuming direct library calls and had to get retrofitted to work in a protected mode environment. Furthermore, X11 naturally takes advantage of multiple processors and graphics processors.
Before, when basically everything was simple blits or pixel plotting, the latency of individual operations was critical.
I'm not sure what you mean by "before". Windows was written that way. X11 had a client/server architecture from the start and has always worked well with it. Windows is the latecomer, and it still isn't very good at it.
These days, with OpenGL serving as the support for the GUI (see Longhorn and Jaguar) clients have to package up command anyway (vertex buffers, display lists, etc) and each operation takes comparitvely longer than a single PutPixel(). Thus, the latency of the communication isn't as much of a factor anymore.
As I was saying, X11 got it right from the start because X11 didn't assume that any program can just bash pixels in the frame buffer. Windows is playing catch-up.
So, what is Sun actually going to do? Are they going to make their headers 100% identical to Linux? What other interfaces are they going to support? Is there going to be a /proc and a devfs?
X11 is the equivalent of GDI or Quartz; it doesn't have to enforce GUI semantics. If you want to enforce a "coherent" desktop on top of it, you can impose whatever draconian measures you like. KDE looks quite coherent and standardized to me, for example.
It's a myth, in any case, that Windows or MacOS are any more coherent than, say, KDE. Take a look here for an extensive critique. And you think that the appearance or window management behavior can't be changed on Windows? Think again: Stardock, Litestep, Microsoft PowerToys.
but isn't the fact that video drivers are implemented in userland an architectural problem to begin with?
The video drivers are in the kernel. The drawing and acceleration is in the display server. The toolkit is in the application. It's fast and it's robust. It's what NeXTStep and MacOSX do as well. Where is the "architectural" problem?
Plus, the resources mechanism is absolutely byzantine and needs to be be razed,
Neither Gnome nor KDE use the X11 resource mechanism. They use something much more like Windows. That's actually a shame because the X11 resource mechanism is better.
as well the complex distinctions between server and client (wait, who's the server, who's the client, who has the toolkit?, who's running the window manager? what the fsck is going on?).
Windows, MacOS, and NeXTStep make the same distinction as X11: they have a low-level graphics and windowing component that runs in a display server, and they have a high level toolkit part that runs in a display client.
Altogether, it looks to me like you have a rather outdated notion of what Windows, MacOS, and X11 are. Windows and MacOS have pretty much become like X11 architecturally; they simply lack the well-defined and efficient X11 protocol to support that architecture. On the other hand, X11 toolkits (for better or worse) have become much more like Windows and MacOS toolkits. All three of them have gotten direct rendering and 3D acceleration.
If you choose to compile and install things by hand, of course, you have to edit XF86Config files. But even there things have gotten much easier (e.g., in XF86v4, you don't need to worry about modelines anymore).
X11 is the equivalent of GDI, Quartz, or DisplayPostscript. If you want to try to enforce a consistent user environment on top of it for your new OS, you can do that as much as you can do it on any other platform.
Conversely, Quartz and GDI fail to enforce a consistent UI either. Toolkits like Borland, FLTK, wxWindows/Universal, Qt, Swing, and OpenStep run on top of both of them and they give you applications that look and behave differently from native apps.
In different words, I think this is a red herring. If you build a new windowing system, for a while, things will be consistent because your own applications will be the only ones. Once it becomes popular, the consistency vanishes.
The East Coast is prone to earthquakes as well. Even New Jersey and New York. Have a look here. There are regions on the East Coast that are even more susceptible, and unlike the West Coast, the East Coast is almost completely unprepared for even a moderate earthquake.
What is supposed to be wrong with the X11 graphics engine? Why do people keep complaining about it?
X11 does hardware-accelerated 2D and 3D graphics with transparency, anti-aliased fonts and graphics, ClearType-like rendering, and server-side geometric transforms.
You can get a good X11 server into less than 1Mbyte with almost no off-screen memory, or you can give it hundreds of megabytes for caching, buffering, fonts, and textures.
X11 provides a uniform and powerful API for all sorts of input devices.
It can't be the client/server architecture that people complain about because neither Windows nor OSX do direct graphics I/O for their UI either. Neither can it be the footprint, because if you look at Windows or OSX GUI apps and system resources, they are as big as X11 on similar hardware, or bigger.
Graphic design can't be the problem either: X11 imposes no constraints on the toolkits, GUIs, or desktops you build on top of it, and X11 toolkits have emulated Windows, OS/2, and MacOS/OSX so closely that it is really hard to tell the difference.
So, what concretely is supposed to be wrong with X11? Why this visceral dislike? Why do people keep starting projects to replace X11 with unaccelerated display servers?
How would anybody know? End users don't get asked whether they want Flash. it is forced upon them. End users usually can't even turn off the blink tag or GIF animations.
In fact, I would like Slashdot more if it had more pictures to help navigate, and some other font besides Times New Roman.
Slashdot has whatever font you choose as your default text font. Take a look in your browser's preference settings some time.
some sites are only Flash.
Some sites are also only blank pages. Same thing.
Oh my god, not by a long shot. The US is clearly not the most prosperous civilization; Switzerland and Canada, among others, have a higher standard of living. US citizens may collectively command the largest amount of aggregate wealth right now, but that would be a rather odd-ball measure of prosperity.
The "most moral" by what measure? In terms of teenage pregnancies, legal system, percentage of population in prison, economic disparities, poverty, and race relations, the US is not particularly "moral". Neither is the vastly disproportionate use of resources or the per capita cost Americans impose on the environment. So, what sense of "most moral" were you thinking of?
In terms of supporting freedom and democracy in the world (and at home), that seems pretty iffy, too. Take a look at the book "Blowback", which chronicles some of the US foreign adventures. Arguably, the Marshall plan was good, but even there, the US was not single-handedly responsible for freedom and democracy in Europe after WWII.
The US isn't a bad country. In fact, on balance, it's doing pretty well. But your assessment seems way out of proportion with reality.
Protection against physical tampering is secondary. It's nice, but even if it didn't exist at all, smart cards would still be very useful. This particular attack seems so tricky that it may not even be worth doing anything about.
It seems to me the ball is in our court: the only way I see global poverty is going to be addressed is if we help these other countries and if we voluntarily give up some of our power. Maybe we should let third world nations unilaterally dump textiles on our markets without forcing reciprocal, "fair" trading arrangements. Maybe we should let them "violate" our copyrights and patents, just like we violated European copyrights and patents during the first few hundred years of our history.
Or do you have any better ideas?
Yes, and that's no accident: the size, layout, and feel of controls on current cars is driven by physics and engineering. You know roughly how big a manual transmission or steering wheel ought to be. There were a few designs that worked and were cost effective; anything else was really expensive or downright impossible.
Electronics already made things worse to some degree. But GUIs remove all those constraints and designers can do things in arbitrary ways. Users of a GUI have no guidance based on physical constraints what controls to expect or how to operate the car. You can't put 10000 physical buttons into a car, but you can put 10000 GUI buttons into a UI.
When cars were first being made, why didn't they use a "horse reigns"-like steering system, since that's what everybody was using before and thus must've been "intuitive"? Or why not a rudder lever like on a smaller boat, rather than the steering wheel from larger boats?
You wouldn't be able to turn the wheels at low speed without some kind of gearing; that's why rudders or reigns don't work well. The design of the steering wheel is driven by physics and ergonomics. Try coming up with something that works as well.
Did you just hop in the car and know how to do everything? Of course not, because it's not intuitive
Of course, I did, and you probably did, too. Someone tells you how to turn the thing on and get it into "drive", and after that, it's trivial: one pedal makes it go, the other makes it stop, and the steering wheel turns it.
Almost all driving instruction and learning is safety-related, not about how to move the car.
Even with IE on XP, I found the site unintuitive and hard to navigate.
I ordered a Mini [miniusa.com] at our local BMW dealership a few months ago
I find the Mini site just as difficult to navigate and user-unfriendly. Yes, it's flashy, but try finding the specs: you have to sit through several animation sequences, the fonts don't work well on high resolution screens, and it opens windows all over the place.
Their web sites and their cars are strangely similar: very flashy and "cool", but something simpler and less flashy is probably more usable.
Adding a menu system probably doesn't reduce the complexity, distraction, or cognitive load. Maybe it would be better if the designers and marketers figured out what is essential and how to reduce the number of functions back down to 35. That would probably also make the car less distracting and safer.
Also, if the only evidence is two earthquakes separated by a few seconds, wouldn't we expect some of those merely by chance? What additional evidence is there for strangelets?
It's not clear that the problem making wireless more attractive for the last mile is the equipment cost. Everything helps, I suppose, but installation and ISP transmitter costs (antenna rental, Internet hookup) are likely to be more costly. See sprintbbd.com for one venture that has been put on hold.