I must admit, when I first read the announcement I thought bad things about Ximian, with a sort of gutteral "sell-out" feel.
But after having thought about it, I have no problem with this at all. I think what they are doing is something like what Troll tech or Cygnus has done -- having a complete Free product, but demanding payment when it's used in a non-free way. Troll and Cygnus have this with GPL libraries and providing alternate licenses for payment.
Ximian can't quite do this, unfortunately, since the GPL doesn't restrict any normal use, so it has to make the plugin proprietary. It's a shame that there isn't a way to do it otherwise, and it is dangerous to create precedence of proprietary plugins to GPLed products. But, I feel the basic intention is the same.
At some point, there'll be a Free program that basically matches the functionality of Exchange, and someone (maybe Ximian) will make a Connector for that. If that is proprietary, then we'll have something to complain about.
But until then, if Ximian can find a way to fund themselves, more power to them.
The judge is very likely to believe: I am not required to give this information by way of my fifth amendment rights. That implies you've got something to hide, and I assume the jury can take that into account, but they can't force you to give the key. All of this assuming, of course, that you are in the US and not the UK.
{72} Cryptography may provide a technical fix for Supreme Court decisions allowing the invasion of one's private papers. However, the effectiveness of that fix will depend on whether the Court holds that use immunity from the compulsory production of a cryptographic key extends to the incriminating documents decrypted with the key. Logic suggests that the Court should so hold.
But that means nothing compared to actual precendence, of which I am not aware, 'cause I don't really keep up with this stuff. I assume it's protected, as the recent case against that mobster was borderline and it wasn't a question of whether the guy would be forced to give his password, but what to do once it had been aquired.
The problem is, Stallman's viewpoint only serves to support the stereotype of the free software movement: "A bunch of opinionated geeks, who have all these high and mighty principles, but won't actually help Joe User learn how to use this stuff, because they don't consider him worthy."
Those opinionated geeks with high and mighty principles are what made GNU happen, why there's so much GPLed software, why Linux is as complete as it is. They deserve credit for working -- often without compensation -- towards that goal. They aren't businessmen, they aren't marketers, but it's those programmers that are why we are where we are.
Like it or not, those programmers are more important to GNU/Linux than all those users you speak of. Not that users aren't good -- but the real momentum of the system will come from the creators, not the users.
Since when does RMS have a duty to fight for you to do whatever you want as effectively as possible? Just because he talks about ideology and the common good doesn't mean he owes you anything. The good for dustpuppy != the good for everyone.
Do you want to run whatever proprietary program on Linux? Fine, go ahead, RMS could probably care less. But where do you get off making it sound like your goodwill towards yourself is some sort of moral virtue?
But, OTOH, he has shown himself in the past to be a purveyor of utterly ignorant dogma, almost on the level of religious zeal, that, as such, I dont consider him a leader of anything.
Can you give an example? I feel like a lot of people here are putting words into RMS's mouth -- but you don't need to do this, because this is the internet and you can link directly to most of what RMS has written. So exactly what utterly ignorant dogma are you speaking of?
Personally, as a businessman, I would never, ever have RMS on a board. He is quite good at pontificating his views, but he is absolutely horrible at seeing other sides of arguments. It's his way or the highway. Can you imagine the board meetings? He would drive everyone else crazy.
As a businessman, what perspective do you have on Gnome anyway? Gnome isn't a business, its board is not the board of a business, large parts of the work are done on a non-commercial basis. There are many, many people who have contributed to Gnome without expectation of monetary reward. To businesses these people are just dopes -- and yet they are the only reason Free Software has succeded, they are the ones who have actually given for the cause. Businesses just make deals. One day they help you, the next day they change their mind, or maybe just go bankrupt.
As a businessman, I think you can't appreciate what GNU is about, what Gnome can be, and what is important for the project. There will always be arguments and negotiations -- RMS will argue about important things, like freedom, while the businesses will be more apt to argue about territory.
I like Galeon too, but it's keyboard shortcuts still need a lot of work -- mostly in terms of focus. Half the time I try to use a shortcut it doesn't work, because the focus is on the URL or tabs or some other pointless area.
Apparently this is a problem with integrating with Gecko, though... IMHO the basic problem of keyboard focus exists across all GTK apps... I wish there was never any concept of focus besides being in a text-entry area. You always have to guess what the behavior will be when you hit a key otherwise. It bugs me a lot in evolution, too (which I otherwise quite like)
Re:good job mozilla, way to break everyone's stats
on
Mozilla 0.9.6 Released
·
· Score: 2
BTW -- has anyone found a good way to make favicon.ico on Linux? The Linux browsers don't mind if it's a PNG or something, but IE doesn't seem to like that at all. In order to get something that IE would accept I ultimately had to do it on Windows, which was annoying.
Special options to ImageMagick? Does a program exist just for this?
Re:Prettier outside, same junk inside
on
Concept PC 2001
·
· Score: 2
It's just not worth it to have an upgradable computer for most uses. I know how to do it fairly well, and my main computer is easily upgradable, but for instance the simple box I got for my girlfriend isn't worth bothering with -- it's easier and less of a pain to just get a new one. In the not-so-distant future I'm sure the same will be true with my own personal computer -- I might as well just replace the thing. USB, firewire, etc., only make this more true, as the more interesting, eclectic, or novel components won't be built into the computer.
The real problem is that current computers (or perhaps operating systems) don't allow for replacement of hardware without software. At this point with Windows, the operating system decays faster than the computer anyway, so people don't notice it so much. But it isn't true for Linux, and is less and less true for Windows, and people are going to want to keep their software environment even as they replace the hardware.
How exactly this will be possible, I'm not sure. I suppose I can imagine a piece of software that would mirror the old hard drive over to the new one, over USB or something -- it wouldn't even be a big deal if it took an hour or two to transfer, considering the benefits. Kind of like Ghost or something.
Licensing will make this very difficult on Windows -- anyone who writes that will be asking for trouble. Also, the nature of PC hardware makes this difficult -- you'd have the wrong drivers installed once you crossed over (it would work really nicely on Apple hardware, though).
But hey, maybe this is already available right now...? If it gets included with new computers, then there really wouldn't be much point to upgradable hardware.
Rather, non-relational databases already exist -- there's lots of object databases. And when you consider that many of these object databases support objects that can also be serialized to XML (e.g., for XMLRPC), then you see that they are pretty much orthogonal to an XML database.
Object databases are being used more and more, I think -- though they aren't taking off or even biting into RDBMS's much...
I think he confuses you with his definitions. If I interpret your argument correctly -- which I think I do, because I agree -- but use his terms, there are no programmers, only software engineers.
That is, he seems to think there's some dead-simple job that you can give to the grunts to do, and doesn't matter if they aren't that clever. You would counter -- there are no such jobs. A job worth doing at all is worth doing well, with a thought of architecture and design.
Of course, some jobs don't matter as much -- a well defined component can be programmed poorly but to spec and still work. If it has problems, they won't be the large architectural problems that haunt us for decades.
But then, you can counter -- we are already haunted by architectural problems, because the architects of yore were not prescient. And it takes a good programmer to turn that poor architecture into something good. And good programmers can do that! That is software engineering -- at least how he'd define it.
Bad programmers may not do as great harm when they are forced into very constrained environments by the definitions of the architect. But we should all know that doesn't lead to great programming -- passable, perhaps, but great programming is better than that -- great programming brings together all the pieces in a way that trancends anything that is possible in a top-down environment. But institutional people don't care about great programming. They make due with passable.
But the opinion of the original author -- for all the flaws I see in it -- was that software has to be better than passable. So there's two ways: we expect the Engineer-Man to be all-knowing and all-controlling, Ada/Eiffel-style (but with a particular passion for domination), or just maybe we make great programs with lots of great programmers, working together with everyone expected to be smart and have an eye on the overall design. Cathedral vs. the Bazaar, I suppose.
However, it's important to consider one thing -- how many bridges are built every year? How many have as many challenges as the Clark Bridge? Not many, certainly. How much software is written in a year?
The other thing about that -- the design of bridges isn't holding back the production. The economics are incredibly different.
If I had a "great" bridge idea, it wouldn't much matter, would it? Or a great building idea -- an Arcology or something. Skyscapers cost a billion dollars or so to produce, I believe. My great design simply isn't going to happen. And even the best building designs can't cut down the price that much -- no orders-of-magnitude advances to be made in one generation. There's just not nearly as much (relatively) at stake in the design of a building.
But a good design for a program -- if you have a good design, you're golden. Hell, you're finished -- running your design (the program) through a compiler does not cost a billion dollars, even for the largest program.
But maybe we should use the construction analogy, like you say: for a long time we've been building lots of mud huts. That was okay, because people need a place to live, even if it is humble. Too bad it melts in the rain.
Now we've learned how to make wood houses, and behold, they are better. Some people are using even more advanced techniques -- brick and concrete, let's say -- and it can be cumbersome, because we haven't built the infrastructure to make those techniques that efficient yet -- it's sometimes hard to get good brick, just like you sometimes must program in C because the client likely won't have Java installed, and you can't build a second story on a mud hut no matter how good the design or materials.
Some people use these more advanced techniques to build very good, very solid houses -- such things do exist, some software doesn't fail. Really. We have added more stories to our buildings, and built roads between them. Some people talk about skyscrapers, but everyone still needs a place to live, so most people make houses.
And we're getting better -- the problems of yesteryear are not the problems of today. Our construction techniques have improved quickly, not on the timescale of epochs like construction. We should be damned proud. And software pushes the envelope more than construction does, but people seldom die when our software collapses, so what the hell.
FWIW, I think that number must be for Debian stable, which is way out of date. My own calculations are that there's 13583 packages in sid. I can't for the life of me find an official number on the web.
I got this with: grep '^Package:'/var/lib/apt/lists/* | wc
13583 27166 1416287
If someone knows if that's a right or wrong calculation, please correct me. I know the number of packages has really ballooned recently, but I didn't think it was quite that much. OTOH, I think it must have had a lot to do with the opening up of the new-maintainer process, which happened between the last stable release and now. A lot more new maintainers means a lot more new packages.
Totally. They shouldn't even install dselect by default. It's horrible.
Quit out right away. Use apt-cache to search for packages (like apt-cache search icq). Pay particular attention to the task- packages (try dpkg -l 'task-*'), which will install a bunch of things at once (like everything you need for gcc, etc).
I'm sure there's tools I don't know about myself, but that's the problem with Debian -- great things exist and you just don't know about them (like apt-cache, which I only learned about a couple months ago, but it's a wonderful tool)
Debian comes close to this but in a much different way that is very top heavy in terms of people assembling packages, etc.
I don't really know FreeBSD that well, but I strongly suspect that Debian policy guidelines are much more agressive than FreeBSD policies. So if it takes more effort to make a Debian package than a FreeBSD port, it's because there's more issues to deal with.
I also am under the impression that there are considerably more debs than ports -- the Debian homepage says 3950 packages, but I don't know if that's in stable, testing, or sid. There starts to be combinatorial difficulties to maintaining a system when you get a really large number of packages. And Debian pulls it off (while the RPM based distributions don't).
If Debian works in this area, it's really a matter of social structures, not technical ones. FreeBSD has a very different structure, which provides good stability but not the inclusiveness of Debian.
Contrary to what other people say, I think Debian is closer to FreeBSD than Slackware. FreeBSD ports at least sound very similar to Debian packages -- Debian packages just happen to precompile for you, which saves a little time for you but isn't otherwise significantly different. apt handles building from source as well, if you really want to do that.
Slackware is more do-it-yourself than FreeBSD is, it seems to me. I don't know what the other BSDs are like -- maybe they are more like Slackware.
Building from source tarballs is okay, especially when you have a very limited set of functions and only a handful of user accounts. But if you want a richer environment and/or you want to have other users (who may not be entirely trusted), it seems like a lot of effort. A good packaging system -- be it binary or with source -- saves a ton of time.
I was just recently talking with someone who was firmly convinced that you couldn't do a decent job of administering a server without spending 10 hours a week on the server. I think he was stuck in the past -- at least for a simple server, dedicated to a few stable functions (like email or web hosting), it hardly takes any time at all to keep a good OS working. I think that guy just isn't familiar with modern OSes and distributions. Old school just isn't efficient.
I think you overestimate the intrinsic speed problems of CGI. Sure, you have to start a new process, but that doesn't take that many resources. If you have to start up a complicated interpreter, as you would for PHP, then yes it's slow. But a small C program starts fairly quickly.
When testing different adapters for an application server I was playing with, there were persistent versions written in Python, for use with mod_python/mod_snake -- the adapters were essentially small scripts that contacted the application server. Those persistant Python versions were actually slower than an equivalent C CGI program. Of course, the C version built as an apache module was somewhat faster, but they were both at the point when neither was a significant bottleneck. So CGI can be pretty fast.
You can actually do what is essentially CGI through PHP too -- if you have something that needs to be run suid, then run it through system() (which loads up a shell, which is annoying and slow) or some other way (I don't know of a way to call a program directly in PHP...?)
Or you can go the FastCGI (or FastCGI-like) direction, where you have a sub-server that handles certain requests. I don't know how easy that is to do in PHP -- it's very useful to have object serialization at that point, and I don't think PHP has that (?)
If he's not trying to get the attention that doing those things causes, why else would he want to do it? Surely, even if a project is part of the GNU, it is up to the project leaders and the developers to determine which way a project goes and how it does things. Not RMS as he seems so desperate to do...
Because he believes passionately in his ideals, and has formed his identity and dedicated his life to those ideals. He really, really, really wants software to be Free.
As to particular issues where technical decisions have led to conflict -- Emacs/XEmacs, The Tcl wars, and maybe glibc (though it's only hot words for now)... yes, when confronted with other strong personalities, things can get pretty hot, and he's not great at compromise. Hell, I heard him talking about splitting HURD from Debian because of some rather minor licensing issues, even though Debian is the just about the only best friend the FSF has among distributions.
But sometimes that sort of aproach actually works -- if everyone had been peaceful and compromising about Qt, it would have never been released under the GPL. He doesn't always do the best thing, but he does provide a sort of anchor.
You can make money like Troll Tech does (as well as Cygnus and a few others) -- release your code under the GPL, and force those who want to use your code in proprietary programs to pay you for the code under a different license.
I can't remember where I saw this brought up, but I'm pretty sure I've read RMS say that he sees nothing wrong with this. And there isn't anything wrong with this -- you are providing good code to anyone Free program and programmer, and still making money off the traditional system. In many ways you do more for Free Software than if you released your code under the LGPL. You aren't locking proprietary software out, but you are giving a stronger incentive for them to release their software under the GPL or BSD license.
It's a real shame that Troll Tech didn't switch to the GPL earlier, because I think we'd all be united behind KDE (and TT would probably be making more money). But they had every chance at the time -- too bad, really.
I believe nearly 50% of current Open Source-licensed software is GPLed. I think you underestimate it.
Also, while copyleft licenses, and the GPL in particular, is a subset of licenses classified as Open Source, the Free Software movement is distinct from the Open Source movement -- neither is a subset of the other.
Lastly, RMS's politics have not gotten kookier -- they've stayed pretty much the same all along. That's a record of consistency that no Open Source advocate can claim, since the OSS movement didn't exist back when the GNU Manifesto was written -- and OSS wouldn't come to exist until all the hard work had been finished, and there was a fast-moving wagon to hop onto.
For God's sake, someone mod this down as ignorant and uninformed. RMS is the undisputed father of GNU, and everyone should agree that he has been an extremely important figure in the Free Software movement. He's never claimed to be the father of Open Source, and in fact he goes to pains to distinguish himself from that philosophy whenever necessary. Had this guy ever read or heard anything from RMS, he'd know that.
Also, the idea that RMS is doing this for his personal glorification is absurd -- people may (perhaps rightly) criticize him for portraying himself as a martyr, but he's not starved for attention (*coughESRcough*).
If people want to start a debate on the place of proprietary software and the philosophies that RMS presents, can't it at least be hung off a slightly informed comment? Somebody who doesn't know any better will read this comment and think it's based in some sort of fact.
RMS's leadership in the GNU project has been largely constructive -- the GNU project has written a lot of software. It could have been different -- they could have spent their time just bitching about copyright law and philosophy. They didn't. I would hope (and expect) that it would be the same in GNOME.
What RMS absolutely wouldn't accept is the idea that a proprietary program is good enough to fill a niche. If people have a real need for an application and they can't do it with free software, to RMS that is as good as if they can't do it at all.
I think this is good. GNU has written a lot of boring, unfilling software because it was needed -- not because it was fun to program, not because it earned them praise, not because it was even scratching a personal itch -- and they wrote it in spite of commercial equivalents. They wrote it because of a long-term vision. That they did this is why GNU/Linux exists.
I think that sort of vision could be a damned good way to aproach GNOME development. I don't think RMS is the only one with that vision -- but his consistency provides a kind of anchor.
Some people seem to think this is a real (if sleazy) porn site that somehow thought that kdhxfm88.org would be a good domain name.
I think it's worse than that. He's probably putting up something that is specifically meant to be offensive and annoying, to further encourage KDHX (and other domains he registers) to pay him off. I doubt KDHX really cares about the domain -- they let it expire after all. But it's not that they don't get to use the domain, it's that the domain is offensive and slanders their organization. I think they would easily win any case against this guy -- not just to get back the domain name, but a libel/slander suit against him. If they were actually to try to do this, they would want to contact other people who have also suffered this extortion, to pool resources.
The server appears to just be on DSL (a traceroute stalls on netblock-66.51.198.150.dslextreme.com, which is probably some firewall just in front of his computer). Which makes it seem even less like a real porn site.
DSLExtreme is a DSL provider in California, apparently. From them you may be able to track down who registered the domain, who is presumably also hosting it. From there you can send a cease-and-desist letter directly to the actual owner, using his actual name, and not falsified information in the WHOIS database. Or if you really felt like it you could sue him (and more power to you).
Also, you can probably get DSLExtreme (whoever they are -- they seem like a normal sort of provider) to shut down his access. I somehow doubt the site fits in their terms of service. If nothing else that'll stop him for a while, and it'll annoy him.
Slashdot Mirror
But after having thought about it, I have no problem with this at all. I think what they are doing is something like what Troll tech or Cygnus has done -- having a complete Free product, but demanding payment when it's used in a non-free way. Troll and Cygnus have this with GPL libraries and providing alternate licenses for payment.
Ximian can't quite do this, unfortunately, since the GPL doesn't restrict any normal use, so it has to make the plugin proprietary. It's a shame that there isn't a way to do it otherwise, and it is dangerous to create precedence of proprietary plugins to GPLed products. But, I feel the basic intention is the same.
At some point, there'll be a Free program that basically matches the functionality of Exchange, and someone (maybe Ximian) will make a Connector for that. If that is proprietary, then we'll have something to complain about.
But until then, if Ximian can find a way to fund themselves, more power to them.
This article comes to the conclusion:
But that means nothing compared to actual precendence, of which I am not aware, 'cause I don't really keep up with this stuff. I assume it's protected, as the recent case against that mobster was borderline and it wasn't a question of whether the guy would be forced to give his password, but what to do once it had been aquired.Like it or not, those programmers are more important to GNU/Linux than all those users you speak of. Not that users aren't good -- but the real momentum of the system will come from the creators, not the users.
Do you want to run whatever proprietary program on Linux? Fine, go ahead, RMS could probably care less. But where do you get off making it sound like your goodwill towards yourself is some sort of moral virtue?
As a businessman, I think you can't appreciate what GNU is about, what Gnome can be, and what is important for the project. There will always be arguments and negotiations -- RMS will argue about important things, like freedom, while the businesses will be more apt to argue about territory.
Apparently this is a problem with integrating with Gecko, though... IMHO the basic problem of keyboard focus exists across all GTK apps... I wish there was never any concept of focus besides being in a text-entry area. You always have to guess what the behavior will be when you hit a key otherwise. It bugs me a lot in evolution, too (which I otherwise quite like)
Special options to ImageMagick? Does a program exist just for this?
The real problem is that current computers (or perhaps operating systems) don't allow for replacement of hardware without software. At this point with Windows, the operating system decays faster than the computer anyway, so people don't notice it so much. But it isn't true for Linux, and is less and less true for Windows, and people are going to want to keep their software environment even as they replace the hardware.
How exactly this will be possible, I'm not sure. I suppose I can imagine a piece of software that would mirror the old hard drive over to the new one, over USB or something -- it wouldn't even be a big deal if it took an hour or two to transfer, considering the benefits. Kind of like Ghost or something.
Licensing will make this very difficult on Windows -- anyone who writes that will be asking for trouble. Also, the nature of PC hardware makes this difficult -- you'd have the wrong drivers installed once you crossed over (it would work really nicely on Apple hardware, though).
But hey, maybe this is already available right now...? If it gets included with new computers, then there really wouldn't be much point to upgradable hardware.
Object databases are being used more and more, I think -- though they aren't taking off or even biting into RDBMS's much...
That is, he seems to think there's some dead-simple job that you can give to the grunts to do, and doesn't matter if they aren't that clever. You would counter -- there are no such jobs. A job worth doing at all is worth doing well, with a thought of architecture and design.
Of course, some jobs don't matter as much -- a well defined component can be programmed poorly but to spec and still work. If it has problems, they won't be the large architectural problems that haunt us for decades.
But then, you can counter -- we are already haunted by architectural problems, because the architects of yore were not prescient. And it takes a good programmer to turn that poor architecture into something good. And good programmers can do that! That is software engineering -- at least how he'd define it.
Bad programmers may not do as great harm when they are forced into very constrained environments by the definitions of the architect. But we should all know that doesn't lead to great programming -- passable, perhaps, but great programming is better than that -- great programming brings together all the pieces in a way that trancends anything that is possible in a top-down environment. But institutional people don't care about great programming. They make due with passable.
But the opinion of the original author -- for all the flaws I see in it -- was that software has to be better than passable. So there's two ways: we expect the Engineer-Man to be all-knowing and all-controlling, Ada/Eiffel-style (but with a particular passion for domination), or just maybe we make great programs with lots of great programmers, working together with everyone expected to be smart and have an eye on the overall design. Cathedral vs. the Bazaar, I suppose.
If I had a "great" bridge idea, it wouldn't much matter, would it? Or a great building idea -- an Arcology or something. Skyscapers cost a billion dollars or so to produce, I believe. My great design simply isn't going to happen. And even the best building designs can't cut down the price that much -- no orders-of-magnitude advances to be made in one generation. There's just not nearly as much (relatively) at stake in the design of a building.
But a good design for a program -- if you have a good design, you're golden. Hell, you're finished -- running your design (the program) through a compiler does not cost a billion dollars, even for the largest program.
But maybe we should use the construction analogy, like you say: for a long time we've been building lots of mud huts. That was okay, because people need a place to live, even if it is humble. Too bad it melts in the rain.
Now we've learned how to make wood houses, and behold, they are better. Some people are using even more advanced techniques -- brick and concrete, let's say -- and it can be cumbersome, because we haven't built the infrastructure to make those techniques that efficient yet -- it's sometimes hard to get good brick, just like you sometimes must program in C because the client likely won't have Java installed, and you can't build a second story on a mud hut no matter how good the design or materials.
Some people use these more advanced techniques to build very good, very solid houses -- such things do exist, some software doesn't fail. Really. We have added more stories to our buildings, and built roads between them. Some people talk about skyscrapers, but everyone still needs a place to live, so most people make houses.
And we're getting better -- the problems of yesteryear are not the problems of today. Our construction techniques have improved quickly, not on the timescale of epochs like construction. We should be damned proud. And software pushes the envelope more than construction does, but people seldom die when our software collapses, so what the hell.
I got this with: /var/lib/apt/lists/* | wc
grep '^Package:'
13583 27166 1416287
If someone knows if that's a right or wrong calculation, please correct me. I know the number of packages has really ballooned recently, but I didn't think it was quite that much. OTOH, I think it must have had a lot to do with the opening up of the new-maintainer process, which happened between the last stable release and now. A lot more new maintainers means a lot more new packages.
Quit out right away. Use apt-cache to search for packages (like apt-cache search icq). Pay particular attention to the task- packages (try dpkg -l 'task-*'), which will install a bunch of things at once (like everything you need for gcc, etc).
I'm sure there's tools I don't know about myself, but that's the problem with Debian -- great things exist and you just don't know about them (like apt-cache, which I only learned about a couple months ago, but it's a wonderful tool)
I also am under the impression that there are considerably more debs than ports -- the Debian homepage says 3950 packages, but I don't know if that's in stable, testing, or sid. There starts to be combinatorial difficulties to maintaining a system when you get a really large number of packages. And Debian pulls it off (while the RPM based distributions don't).
If Debian works in this area, it's really a matter of social structures, not technical ones. FreeBSD has a very different structure, which provides good stability but not the inclusiveness of Debian.
Slackware is more do-it-yourself than FreeBSD is, it seems to me. I don't know what the other BSDs are like -- maybe they are more like Slackware.
Building from source tarballs is okay, especially when you have a very limited set of functions and only a handful of user accounts. But if you want a richer environment and/or you want to have other users (who may not be entirely trusted), it seems like a lot of effort. A good packaging system -- be it binary or with source -- saves a ton of time.
I was just recently talking with someone who was firmly convinced that you couldn't do a decent job of administering a server without spending 10 hours a week on the server. I think he was stuck in the past -- at least for a simple server, dedicated to a few stable functions (like email or web hosting), it hardly takes any time at all to keep a good OS working. I think that guy just isn't familiar with modern OSes and distributions. Old school just isn't efficient.
When testing different adapters for an application server I was playing with, there were persistent versions written in Python, for use with mod_python/mod_snake -- the adapters were essentially small scripts that contacted the application server. Those persistant Python versions were actually slower than an equivalent C CGI program. Of course, the C version built as an apache module was somewhat faster, but they were both at the point when neither was a significant bottleneck. So CGI can be pretty fast.
You can actually do what is essentially CGI through PHP too -- if you have something that needs to be run suid, then run it through system() (which loads up a shell, which is annoying and slow) or some other way (I don't know of a way to call a program directly in PHP...?)
Or you can go the FastCGI (or FastCGI-like) direction, where you have a sub-server that handles certain requests. I don't know how easy that is to do in PHP -- it's very useful to have object serialization at that point, and I don't think PHP has that (?)
As to particular issues where technical decisions have led to conflict -- Emacs/XEmacs, The Tcl wars, and maybe glibc (though it's only hot words for now)... yes, when confronted with other strong personalities, things can get pretty hot, and he's not great at compromise. Hell, I heard him talking about splitting HURD from Debian because of some rather minor licensing issues, even though Debian is the just about the only best friend the FSF has among distributions.
But sometimes that sort of aproach actually works -- if everyone had been peaceful and compromising about Qt, it would have never been released under the GPL. He doesn't always do the best thing, but he does provide a sort of anchor.
I can't remember where I saw this brought up, but I'm pretty sure I've read RMS say that he sees nothing wrong with this. And there isn't anything wrong with this -- you are providing good code to anyone Free program and programmer, and still making money off the traditional system. In many ways you do more for Free Software than if you released your code under the LGPL. You aren't locking proprietary software out, but you are giving a stronger incentive for them to release their software under the GPL or BSD license.
It's a real shame that Troll Tech didn't switch to the GPL earlier, because I think we'd all be united behind KDE (and TT would probably be making more money). But they had every chance at the time -- too bad, really.
Also, while copyleft licenses, and the GPL in particular, is a subset of licenses classified as Open Source, the Free Software movement is distinct from the Open Source movement -- neither is a subset of the other.
Lastly, RMS's politics have not gotten kookier -- they've stayed pretty much the same all along. That's a record of consistency that no Open Source advocate can claim, since the OSS movement didn't exist back when the GNU Manifesto was written -- and OSS wouldn't come to exist until all the hard work had been finished, and there was a fast-moving wagon to hop onto.
Also, the idea that RMS is doing this for his personal glorification is absurd -- people may (perhaps rightly) criticize him for portraying himself as a martyr, but he's not starved for attention (*coughESRcough*).
If people want to start a debate on the place of proprietary software and the philosophies that RMS presents, can't it at least be hung off a slightly informed comment? Somebody who doesn't know any better will read this comment and think it's based in some sort of fact.
Who at Paypal do you have to ask about those APIs?
Maybe you meant right and left were replaced with "for sale" and "for rent".
What RMS absolutely wouldn't accept is the idea that a proprietary program is good enough to fill a niche. If people have a real need for an application and they can't do it with free software, to RMS that is as good as if they can't do it at all.
I think this is good. GNU has written a lot of boring, unfilling software because it was needed -- not because it was fun to program, not because it earned them praise, not because it was even scratching a personal itch -- and they wrote it in spite of commercial equivalents. They wrote it because of a long-term vision. That they did this is why GNU/Linux exists.
I think that sort of vision could be a damned good way to aproach GNOME development. I don't think RMS is the only one with that vision -- but his consistency provides a kind of anchor.
I think it's worse than that. He's probably putting up something that is specifically meant to be offensive and annoying, to further encourage KDHX (and other domains he registers) to pay him off. I doubt KDHX really cares about the domain -- they let it expire after all. But it's not that they don't get to use the domain, it's that the domain is offensive and slanders their organization. I think they would easily win any case against this guy -- not just to get back the domain name, but a libel/slander suit against him. If they were actually to try to do this, they would want to contact other people who have also suffered this extortion, to pool resources.
The server appears to just be on DSL (a traceroute stalls on netblock-66.51.198.150.dslextreme.com, which is probably some firewall just in front of his computer). Which makes it seem even less like a real porn site.
DSLExtreme is a DSL provider in California, apparently. From them you may be able to track down who registered the domain, who is presumably also hosting it. From there you can send a cease-and-desist letter directly to the actual owner, using his actual name, and not falsified information in the WHOIS database. Or if you really felt like it you could sue him (and more power to you).
Also, you can probably get DSLExtreme (whoever they are -- they seem like a normal sort of provider) to shut down his access. I somehow doubt the site fits in their terms of service. If nothing else that'll stop him for a while, and it'll annoy him.