> The VLIW architecture is beautiful in many ways,
That is true. It is similar to RISC, but it makes more efficient use of execution units and code size. However, in one aspect VLIW is even worth than RISC: the predicted efficiency is only realised if the ISA matches the CPU architecture. Otherwise, a translation layer is required, and the main advantage against CISC evaporates.
Sun had big troubles with than on SPARC. The ISA worked well for the SPARC processors in the early nineties, but it is difficult to execute efficiently on a modern CPU.
I am sure that Intel ran into that problem when designing new revisions of the Itanium, and eventually it is doomed. The marked place just doesn't permit to change the ISA with every CPU architecture, binary compatibility is to important for that.
So: ISA does not matter for efficiency, unless you want to recompile your apps for every new CPU. And most people don't want to do that for a meager performance advantage.
> My own home network contains multiple Windows, OS X, and Linux boxes.
Under the assumption, that most people don't want to set up their own private computer lab, this can only mean that at least two of these OSes suck. In the best of worlds, all three OSes would run happily side by side.
Of course, commercial interest is against that. While Linux (in theory) can coexist with other systems, MacOS runs into licenses issues, and Windows plays every dirty trick from to book to not coexist with another system.
> So when people come to me with problems or for advice, I don't preach from the Gospel According to Steve or the Revelation to St. Bill (or the Epistles of Linus). I listen to what their needs are, and I suggest whatever offers the best solution for them.
Get a Mac, a Wintel PC, a Linux AMD64 system, a fileserver and a firewall? I am not sure they would be happy with that answer:-)
I stay with my judgement that computers as of today suck, and it is absolutely amazing that people to up with it. Ok, considering the alternative of not using a computer, it may be understandable...
Just give it a try, it is actually quite a nice browser. Sure, while it does block pop-ups very reliably, it cannot block ads on the page itself. Apart from that it works really well.
Certainly the title bar and the controls are nice and efficient. They got a lot of things right there: the controls are intuitive, relevant and mostly complete, which is more than you can say about Firefox. They also take up less precious desktop space (it is not just eye candy).
Apart from that, I think it is "good enough" for most users. I would certainly recommend it over IE, even if it falls short of delivering "the safest browsing experience".
Actually, it is. The standard was updated in 1998 to avoid confusion (Standard IEC 60027-2). Giga is 10^9, and it is constant, which means it does not change just because you use it for hard disks or memory.
If you mean 2^30, then you have to say gigabinary, abbreviated as gibi or Gi. Having different name for different things can avoid an awful lot of confusion, so it would very much recommend using them.
And now please put the following events into the correct order: America goes metric, hell freezes over, people use Gibi correctly.
Yes, you can append a . to the name. http://www.google.com./ will only ever get you google, or nothing at all.
This "trick" is a lot older than mozilla, it applies to all DNS lookups. It also prevents the name from matching a machine on the local network. Mozilla also seems to recognise the dot, and it avoids the "guessing" step.
I think it is a shame that this "episode" was so destructive. Many similar open source projects have benefited quite a lot from a company that is "productising" the development versions. Think wine, think Qemu, think Mozilla: they all have commercial partners that apply the last polish and charge money for that.
The same could have been true here, if MXS had been a bit more open and honest. After all, the GPL does not prevent you from charging money.
As it is right now, PearPC seems pretty dead. By now, Qemu clearly has the better PPC CPU emulation, it only takes a clever delevoper to make it boot Mac OS X. Which would be awesome.
> Look, ID says that science should leave room for evidence of a plan in the way things happen.
The is always room for evidence, but in this case there is no evidence. Look at designed things: cars, airplanes, computers. They show precisely repeated features (cylinders, fuselage windows, bolts), and simple geometric features (lines, circles etc). That would be evidence of design. Conspicuously absent in nature, if you ask me.
On the other hand, you find oddities in nature that are very difficult to explain without evolution. Take the wing of a bird, the leg of a dinosaur and the human hand, they all have the same basic topological features (not dimensions, of course). Since they have a completely different purpose, this could not have happened by design.
Or what about the wheel? The single most important design idea. It is completely absent by nature, because it cannot evolve (easily) by evolution. Just imagine wheel and bodies evolving along different lines:-)
> As a scientist, I'd like to think that if a supernatural explanation fits the evidence better than the alternatives, and enables us to make accurate predictions about future events (and is thus able to be invalidated by those predictions being incorrect), then it would eventually pass into the scientific mainstream.
Of course. But you certainly have noticed that most supernatural explanations have no predictive power whatsoever. How would you expect a "created lifeform" to look like? Do humans fit this description? It is undecidable, because there is no such thing as "evidence for creation".
The big problem is that evolution is not a law of nature, it is more of a concept. And "intelligent design" is just a polite way of declaring intellectual defeat. So both are very difficult to compare. But if one suits your need, take it by any means:-)
> Think about what your goals are. Then find the best tool to get there.
I couldn't say it any better. If you want to make sure that functionality stays intact while you change things, write system tests. If you want to use code in a different context, document it and write unit tests first.
Whatever you do, you need some "absolute reference" to find out what is right and what is a bug. Tests are no good if they just preserve old bugs for eternity.
> I suppose there's a difference between serving just 500GB or a few terabytes.
If it is only that, you could just stick 8 disks in a nice server system, that gets you a few terabytes. Use your filesystem of least mistrust, such as reiserfs, XFS or JFS.
RAID 1+0 or 10 is very nice, but you still have a single point of failure. NetRAID and some kind of fail-over redundancy with another machine way be the way. You will not get a full SAN/NAS this way, but you also avoid a lot of complexity.
Re:How Debian (really) works...
on
Sarge is Now Frozen
·
· Score: 2, Insightful
> And now it is my turn to disagree. When you start talking about the enterprise, a whole different set of rules come in to play.
Yes, but you are confusing to different situations: running an old server with Debian Woody, and installing a new server with Debian Woody. Since servers have a (hardware) lifetime of about 5 years, it makes a lot of sense to run the same software release on it for this time. You might do a mayor update if you need to, but usually you would just do security fixes. That's why you see lots of RedHat 8, Solaris 5.6 or Windown NT 4 around.
However, for new machines the picture is completely different. Getting hardware that runs NT 4 is nearly impossible, which is why every enterprise that does not have a clear migration strategy by now has a serious problem. Debian Woody is similar: finding a recent machine where you can install Woody is probably a challenge (think graphics card on the desktop, and gigabit ethernet/RAID/SATA on the server side).
So Debian is fine for machines that you installed 2 or 3 years ago, but it is not at all a possible choice for a new machine. Which, incidentally, completely defeats the purpose of stability.
Re:Big woop now it's only 3 years behind. FP and F
on
Sarge is Now Frozen
·
· Score: 2, Informative
> I realize a lot of the posts here are in jest, but what's wrong with being a little slow on the release schedule?
The same as always: software aging (also termed "bit rot"). Look at woody (the last release), and you know what is wrong: standard kernel is 2.2 (therefore no support for most USB devices), XFree version 4.1 (good luck finding a graphics card that is supported), mozilla is version 1.0, openoffice.org is not even included, and KDE is still stuck at version 2 (which conflicts with version 3, so that you cannot run *any* KDE 3 application).
So whatever happens to sarge, it is going to be *way* to late. Plus sarge is not nearly up to date either. It is still based on XFree 4.3, and has no support for amd64. Which means it will be obsolet way before Etch can possibly be released.
It is a shame, because apt-get is so much superior to any rpm based solution I have seen. But unless you can run testing, Debian is not really an option on the desktop.
> The 1080i signal is 2 540 line scenes, shot 1/60th of a second apart, with half the picture data. If anything on the screen is moving, it will be in a different place 1/60th of a second later. if you mash those two scenes together into a 1080 line image, those items in motion will blur
That is the basic problem of deinterlacing. You want to increase the vertical resolution for stationary scenes, and the temporal resolution for fast moving scenes. It can be done, but it involves a lot of guesswork, and it can produce artifact.
So for most applications, scaling up half images is a very reasonable solution. If you care about vertical resolution, you should not use an interlaced signal in the first place.
> The claim "intelligent design is a valid alternative" is LOGICALLY FLAWED, and here is why:
Actually, the reason is a lot simpler. "Intelligent design" is not a theory at all, because it does not explain anything. All it says is that live is as it is because it is designed by someone more clever then us, and so we should not bother to find out why it is as it is.
Evolution can be tested, and it has been tested so many times, that it is very difficult to question on factual grounds. Intelligent design cannot be tested, and it cannot be used for anything useful either.
I have no problem with the teaching of intelligent design, as long as it is not taught as a theory.
> I use OpenOffice everyday, and although it's usable I wouldn't say for a second that it was a $300 competitor. It's not that it's bad, more that it's arcane.
I absolutely agree. What's more, development seems to come to a standstill (a bit like TeX, really). Remember the weird Staroffice 4.0 for Linux? It was emulating a complete Windows desktop. And I could bet all that code is still in there. E.g. Openoffice has its own font rendering, which again is arcane: it cannot even do subpixel rendering.
And so many features are still missing or not really working. Try to get multi-level section numbers (1.1, 1.2 etc), it is a pain, and simply does not work properly. Try embedding a dia diagram, and again it seems mostly broken. The PDF export creates wrong logical page numbers, and the TOC is not even hyperlinked.
And now they announce 2.0 as the great new release, but they have to resort to features like "count the words in the selection" as mayor improvements. Man, that is a feature I would take for granted. I wonder whether printing from the command line will finally work. Oh, and whether rtf export/import will keep my document intact, or still garble it up.
No, the codebase of Openoffice seems to be hopelessly bloated. It can do a lot, but you can't teach it any new tricks. It is doomed, I am sure. Maybe koffice will do better.
Yep, that is pretty sweet. Or, if it is not an option, you can use the KDE Printer Admin. Assuming that the access rights are set up, configuring a printer is pretty straight forward. It is certainly a lot better than the web GUI included with CUPS.
No, I think I would charge per page. Nessus is really good about producing length reports:-) My discount price is $50 per page.
One had I got one in hard copy that was 120 pages long. 110 pages where speculating about possibly dangerous applications on port 10000. All that was running there was webmin, bound to the local network. But nessus was run from the outside, and so it used its imagination to come up with hunderds of dangerous applications.
Although I wonder how dangerous an application is that you cannot reach?
The guy is a genius. Sometimes I feel like compiling a list like this myself, but I rarely install a system from scratch, so it is difficult to point them back to a single source.
But Linux needs more people like. Interface bugs are bugs, because the confuse the user, and (thus) the software does not work for them. Calling the user "stupid" wont help either, because you are still stuck with the same user:-).
The most obvious UI bug I remember is the GNOME pop up box when you exit a program without saving. They keep changing it, but it still makes me hesitate every time. It is just extremely nonintuitive. (Yes, and MacOS also took many revisions before they got it right. Microsoft this didn't get it...) Openoffice is a lot better, as is KDE.
Now if the developer would take these issues serious and fix them, Desktop Linux would be a lot closer already.
Dell "said it would begin shipping its Dimension brand of PCs with the new chips relatively soon with prices starting at around $3,000."
So why would you pay 3000 bucks for two throttled CPUs on one die, if you could get a dual PROCESSOR system for the same price? I mean, the second heat sink is not going to raise the price of the system to another level... and you can go with proven technology.
Actually, I would only consider a dual AMD64 system worthwile. With NUMA support improving in Linux, this should be a lot faster than 2 P4 cores competing for the same memory, already suffering from high latency.
> One of the biggest differences between KDE and Gnome is that KDE's use of the Qt library dramatically cuts down on dependancies.
Yes, that is a very valid point. I have compiled ethereal from scratch. It is only a gtk app (not even gnome), but the dependencies where killing me. config, libz, libexpat, glib, atk, freetype, fontconfig, pango, gtk... it is a nightmare, especially since you have it figure out and find every package yourself (and the relation seem to change from version to version).
> GNOME is hellbent on cloning Windows internals while 'innovating' the look & feel
Yes, it feals like that. Hate the "registry", every once in a while it will screw up, and you have to kill processes, delete lock files etc before any gnome application can start.
And than there are the common dialogs. You know, "Do you really want to quit without saving?", or the file selecter. They keep changing them around, but they never seem to get it right. I would put up with it, if they where at least consistent...
> while KDE is hellbent on cloning the look & feel of Windows while pushing new innovative internals
Which is probably a smart move. An open source project does not have the resources in "UI research" that MS has. So if MS software did actually work, I think it would be quite useable. Luckily, KDE also integrates ideas from MAC OS X, which are sometimes brilliant.
There are FAR fewer holes in W2K than people would like to admit. IE may have some problems but not the base OS.
Ok, if you tell me how to install W2K without IE, I would even accept this argument.
Even IIS has been tighted up a great deal.
You mean it is a lot less insecure now than it used to be?:-)
BTW, are "user gains access he shouldn't have" really considered on an W2K system? The majority of "linux" bugs seem to be of this type (symlink attack allowing to read some log file or something). Since W2K is still basically a single user system, I would imagine these are not taken to seriously.
> So I must raise a question of symantics: Is this technically "porting" or mearly something akin to "cross-compiling"?
Can you cross-compile without porting first?
I think it is porting. Porting to cygwin, which is just another POSIX API, like Linux, Solaris, BSD etc. So porting to cygwin is very much like porting to another UNIX variant. It is not like porting to native Windows, though. But why should you, if cygwin works?
Slashdot Mirror
> The VLIW architecture is beautiful in many ways,
That is true. It is similar to RISC, but it makes more efficient use of execution units and code size. However, in one aspect VLIW is even worth than RISC: the predicted efficiency is only realised if the ISA matches the CPU architecture. Otherwise, a translation layer is required, and the main advantage against CISC evaporates.
Sun had big troubles with than on SPARC. The ISA worked well for the SPARC processors in the early nineties, but it is difficult to execute efficiently on a modern CPU.
I am sure that Intel ran into that problem when designing new revisions of the Itanium, and eventually it is doomed. The marked place just doesn't permit to change the ISA with every CPU architecture, binary compatibility is to important for that.
So: ISA does not matter for efficiency, unless you want to recompile your apps for every new CPU. And most people don't want to do that for a meager performance advantage.
> My own home network contains multiple Windows, OS X, and Linux boxes.
:-)
Under the assumption, that most people don't want to set up their own private computer lab, this can only mean that at least two of these OSes suck. In the best of worlds, all three OSes would run happily side by side.
Of course, commercial interest is against that. While Linux (in theory) can coexist with other systems, MacOS runs into licenses issues, and Windows plays every dirty trick from to book to not coexist with another system.
> So when people come to me with problems or for advice, I don't preach from the Gospel According to Steve or the Revelation to St. Bill (or the Epistles of Linus). I listen to what their needs are, and I suggest whatever offers the best solution for them.
Get a Mac, a Wintel PC, a Linux AMD64 system, a fileserver and a firewall? I am not sure they would be happy with that answer
I stay with my judgement that computers as of today suck, and it is absolutely amazing that people to up with it. Ok, considering the alternative of not using a computer, it may be understandable...
Just give it a try, it is actually quite a nice browser. Sure, while it does block pop-ups very reliably, it cannot block ads on the page itself. Apart from that it works really well.
Certainly the title bar and the controls are nice and efficient. They got a lot of things right there: the controls are intuitive, relevant and mostly complete, which is more than you can say about Firefox. They also take up less precious desktop space (it is not just eye candy).
Apart from that, I think it is "good enough" for most users. I would certainly recommend it over IE, even if it falls short of delivering "the safest browsing experience".
> 1,000,000,000 bytes != 1 Gigabyte
Actually, it is. The standard was updated in 1998 to avoid confusion (Standard IEC 60027-2). Giga is 10^9, and it is constant, which means it does not change just because you use it for hard disks or memory.
If you mean 2^30, then you have to say gigabinary, abbreviated as gibi or Gi. Having different name for different things can avoid an awful lot of confusion, so it would very much recommend using them.
And now please put the following events into the correct order: America goes metric, hell freezes over, people use Gibi correctly.
Yes, you can append a . to the name. http://www.google.com./ will only ever get you google, or nothing at all.
This "trick" is a lot older than mozilla, it applies to all DNS lookups. It also prevents the name from matching a machine on the local network. Mozilla also seems to recognise the dot, and it avoids the "guessing" step.
I think it is a shame that this "episode" was so destructive. Many similar open source projects have benefited quite a lot from a company that is "productising" the development versions. Think wine, think Qemu, think Mozilla: they all have commercial partners that apply the last polish and charge money for that.
The same could have been true here, if MXS had been a bit more open and honest. After all, the GPL does not prevent you from charging money.
As it is right now, PearPC seems pretty dead. By now, Qemu clearly has the better PPC CPU emulation, it only takes a clever delevoper to make it boot Mac OS X. Which would be awesome.
> Look, ID says that science should leave room for evidence of a plan in the way things happen.
:-)
The is always room for evidence, but in this case there is no evidence. Look at designed things: cars, airplanes, computers. They show precisely repeated features (cylinders, fuselage windows, bolts), and simple geometric features (lines, circles etc). That would be evidence of design. Conspicuously absent in nature, if you ask me.
On the other hand, you find oddities in nature that are very difficult to explain without evolution. Take the wing of a bird, the leg of a dinosaur and the human hand, they all have the same basic topological features (not dimensions, of course). Since they have a completely different purpose, this could not have happened by design.
Or what about the wheel? The single most important design idea. It is completely absent by nature, because it cannot evolve (easily) by evolution. Just imagine wheel and bodies evolving along different lines
> As a scientist, I'd like to think that if a supernatural explanation fits the evidence better than the alternatives, and enables us to make accurate predictions about future events (and is thus able to be invalidated by those predictions being incorrect), then it would eventually pass into the scientific mainstream.
:-)
Of course. But you certainly have noticed that most supernatural explanations have no predictive power whatsoever. How would you expect a "created lifeform" to look like? Do humans fit this description? It is undecidable, because there is no such thing as "evidence for creation".
The big problem is that evolution is not a law of nature, it is more of a concept. And "intelligent design" is just a polite way of declaring intellectual defeat. So both are very difficult to compare. But if one suits your need, take it by any means
> Think about what your goals are. Then find the best tool to get there.
I couldn't say it any better. If you want to make sure that functionality stays intact while you change things, write system tests. If you want to use code in a different context, document it and write unit tests first.
Whatever you do, you need some "absolute reference" to find out what is right and what is a bug. Tests are no good if they just preserve old bugs for eternity.
> I suppose there's a difference between serving just 500GB or a few terabytes.
If it is only that, you could just stick 8 disks in a nice server system, that gets you a few terabytes. Use your filesystem of least mistrust, such as reiserfs, XFS or JFS.
RAID 1+0 or 10 is very nice, but you still have a single point of failure. NetRAID and some kind of fail-over redundancy with another machine way be the way. You will not get a full SAN/NAS this way, but you also avoid a lot of complexity.
> And now it is my turn to disagree. When you start talking about the enterprise, a whole different set of rules come in to play.
Yes, but you are confusing to different situations: running an old server with Debian Woody, and installing a new server with Debian Woody. Since servers have a (hardware) lifetime of about 5 years, it makes a lot of sense to run the same software release on it for this time. You might do a mayor update if you need to, but usually you would just do security fixes. That's why you see lots of RedHat 8, Solaris 5.6 or Windown NT 4 around.
However, for new machines the picture is completely different. Getting hardware that runs NT 4 is nearly impossible, which is why every enterprise that does not have a clear migration strategy by now has a serious problem. Debian Woody is similar: finding a recent machine where you can install Woody is probably a challenge (think graphics card on the desktop, and gigabit ethernet/RAID/SATA on the server side).
So Debian is fine for machines that you installed 2 or 3 years ago, but it is not at all a possible choice for a new machine. Which, incidentally, completely defeats the purpose of stability.
> I realize a lot of the posts here are in jest, but what's wrong with being a little slow on the release schedule?
The same as always: software aging (also termed "bit rot"). Look at woody (the last release), and you know what is wrong: standard kernel is 2.2 (therefore no support for most USB devices), XFree version 4.1 (good luck finding a graphics card that is supported), mozilla is version 1.0, openoffice.org is not even included, and KDE is still stuck at version 2 (which conflicts with version 3, so that you cannot run *any* KDE 3 application).
So whatever happens to sarge, it is going to be *way* to late. Plus sarge is not nearly up to date either. It is still based on XFree 4.3, and has no support for amd64. Which means it will be obsolet way before Etch can possibly be released.
It is a shame, because apt-get is so much superior to any rpm based solution I have seen. But unless you can run testing, Debian is not really an option on the desktop.
> it is probably a gallium alloy
r mo4.html . They even claim it is nontoxic (no cadmium or mercury).
:-) Cooling wise, I mean.
Or some other weird alloy. You can buy "liquid metal" for fun at http://www.scitoys.com/scitoys/scitoys/thermo/the
But I still wonder what this has that water does not?
> The 1080i signal is 2 540 line scenes, shot 1/60th of a second apart, with half the picture data. If anything on the screen is moving, it will be in a different place 1/60th of a second later. if you mash those two scenes together into a 1080 line image, those items in motion will blur
That is the basic problem of deinterlacing. You want to increase the vertical resolution for stationary scenes, and the temporal resolution for fast moving scenes. It can be done, but it involves a lot of guesswork, and it can produce artifact.
So for most applications, scaling up half images is a very reasonable solution. If you care about vertical resolution, you should not use an interlaced signal in the first place.
> The claim "intelligent design is a valid alternative" is LOGICALLY FLAWED, and here is why:
Actually, the reason is a lot simpler. "Intelligent design" is not a theory at all, because it does not explain anything. All it says is that live is as it is because it is designed by someone more clever then us, and so we should not bother to find out why it is as it is.
Evolution can be tested, and it has been tested so many times, that it is very difficult to question on factual grounds. Intelligent design cannot be tested, and it cannot be used for anything useful either.
I have no problem with the teaching of intelligent design, as long as it is not taught as a theory.
That is about as soon is Hurd is ready? Or was it nearly ready? I always mix those Hirds up.
> I use OpenOffice everyday, and although it's usable I wouldn't say for a second that it was a $300 competitor. It's not that it's bad, more that it's arcane.
I absolutely agree. What's more, development seems to come to a standstill (a bit like TeX, really). Remember the weird Staroffice 4.0 for Linux? It was emulating a complete Windows desktop. And I could bet all that code is still in there. E.g. Openoffice has its own font rendering, which again is arcane: it cannot even do subpixel rendering.
And so many features are still missing or not really working. Try to get multi-level section numbers (1.1, 1.2 etc), it is a pain, and simply does not work properly. Try embedding a dia diagram, and again it seems mostly broken. The PDF export creates wrong logical page numbers, and the TOC is not even hyperlinked.
And now they announce 2.0 as the great new release, but they have to resort to features like "count the words in the selection" as mayor improvements. Man, that is a feature I would take for granted. I wonder whether printing from the command line will finally work. Oh, and whether rtf export/import will keep my document intact, or still garble it up.
No, the codebase of Openoffice seems to be hopelessly bloated. It can do a lot, but you can't teach it any new tricks. It is doomed, I am sure. Maybe koffice will do better.
> By using Mac OS X's interface to CUPS.
Yep, that is pretty sweet. Or, if it is not an option, you can use the KDE Printer Admin. Assuming that the access rights are set up, configuring a printer is pretty straight forward. It is certainly a lot better than the web GUI included with CUPS.
No, I think I would charge per page. Nessus is really good about producing length reports :-) My discount price is $50 per page.
One had I got one in hard copy that was 120 pages long. 110 pages where speculating about possibly dangerous applications on port 10000. All that was running there was webmin, bound to the local network. But nessus was run from the outside, and so it used its imagination to come up with hunderds of dangerous applications.
Although I wonder how dangerous an application is that you cannot reach?
The guy is a genius. Sometimes I feel like compiling a list like this myself, but I rarely install a system from scratch, so it is difficult to point them back to a single source.
:-).
But Linux needs more people like. Interface bugs are bugs, because the confuse the user, and (thus) the software does not work for them. Calling the user "stupid" wont help either, because you are still stuck with the same user
The most obvious UI bug I remember is the GNOME pop up box when you exit a program without saving. They keep changing it, but it still makes me hesitate every time. It is just extremely nonintuitive. (Yes, and MacOS also took many revisions before they got it right. Microsoft this didn't get it...) Openoffice is a lot better, as is KDE.
Now if the developer would take these issues serious and fix them, Desktop Linux would be a lot closer already.
Dell "said it would begin shipping its Dimension brand of PCs with the new chips relatively soon with prices starting at around $3,000."
So why would you pay 3000 bucks for two throttled CPUs on one die, if you could get a dual PROCESSOR system for the same price? I mean, the second heat sink is not going to raise the price of the system to another level... and you can go with proven technology.
Actually, I would only consider a dual AMD64 system worthwile. With NUMA support improving in Linux, this should be a lot faster than 2 P4 cores competing for the same memory, already suffering from high latency.
> One of the biggest differences between KDE and Gnome is that KDE's use of the Qt library dramatically cuts down on dependancies.
Yes, that is a very valid point. I have compiled ethereal from scratch. It is only a gtk app (not even gnome), but the dependencies where killing me. config, libz, libexpat, glib, atk, freetype, fontconfig, pango, gtk... it is a nightmare, especially since you have it figure out and find every package yourself (and the relation seem to change from version to version).
> GNOME is hellbent on cloning Windows internals while 'innovating' the look & feel
Yes, it feals like that. Hate the "registry", every once in a while it will screw up, and you have to kill processes, delete lock files etc before any gnome application can start.
And than there are the common dialogs. You know, "Do you really want to quit without saving?", or the file selecter. They keep changing them around, but they never seem to get it right. I would put up with it, if they where at least consistent...
> while KDE is hellbent on cloning the look & feel of Windows while pushing new innovative internals
Which is probably a smart move. An open source project does not have the resources in "UI research" that MS has. So if MS software did actually work, I think it would be quite useable. Luckily, KDE also integrates ideas from MAC OS X, which are sometimes brilliant.
There are FAR fewer holes in W2K than people would like to admit. IE may have some problems but not the base OS.
:-)
Ok, if you tell me how to install W2K without IE, I would even accept this argument.
Even IIS has been tighted up a great deal.
You mean it is a lot less insecure now than it used to be?
BTW, are "user gains access he shouldn't have" really considered on an W2K system? The majority of "linux" bugs seem to be of this type (symlink attack allowing to read some log file or something). Since W2K is still basically a single user system, I would imagine these are not taken to seriously.
Can you cross-compile without porting first?
I think it is porting. Porting to cygwin, which is just another POSIX API, like Linux, Solaris, BSD etc. So porting to cygwin is very much like porting to another UNIX variant. It is not like porting to native Windows, though. But why should you, if cygwin works?