I think there are other reasons as well. CommonLisp is a real mess in some areas and some of its designers had such disdain for UNIX that it seems to try hard not to live well in a UNIX environment. Both CommonLisp and Smalltalk were pushed along for many years by greedy companies with big dollar signs in their eyes, killing any possibility of grassroots adoption. Dylan and Self missed the boat compared to Java and Python because they did try to be so much more.
I have to say, though: of the currently popular scripting languages, Python seems to closest to becoming a real programming language, and it's still retaining its original utility and simplicity. What Python is missing at this point is a new implementation that provides incremental compilation to efficient native code, plus a few language cleanups to go with that.
let's call it by its true name
on
CPRM Voted Down
·
· Score: 1
People keep referring to these kinds of measures as "anti-piracy". Let's call it by its true name: "anti-fair-use". That's because there is no known technological measure that can distinguish between fair use and piracy. Companies have merely tried to preemptively redefine "piracy" as any kind of copying, even copying that traditionally falls under fair use.
By "these languages", I mean languages that are used like Perl and Python in the UNIX and Linux world.
Dylan is in the same boat with Smalltalk, CommonLisp, Scheme, and Self: great languages with great technology, but they are little used among UNIX and Linux users to do the kinds of things Perl or Python are used for.
Yes, I realize lots of people get really attached to Perl and particular other scripting languages. But the world is full of scripting languages like that. Overall, they offer roughly the same performance and they offer similar functionality. Some of them have slightly better libraries, others have somewhat better error detection, and yet others are somewhat easier to embed in C/C++ programs. Writing one of those byte-code interpreters isn't particularly hard, which is probably while there are so many of them around (Perl, Python, Tcl, Ruby, Lua, awk, etc.).
To me, things will get interesting once one of these languages actually starts delivering a non-trivial, high-performance native code compiler. I don't mean the kind of simple translation to C code that exists for Perl and Python, but something that actually genuinely increases performance and figures out stuff about data types. Will anybody be up to the challenge? That remains to be seen.
Until then, I won't hold my breath for Perl6. In fact, I don't much care. Perl5 is good enough for what I use it for, and there are lots of other choices around when Perl5 isn't good for the task.
One issue often overlooked in these things is the problem that plagues some publishers and causes them to reject unsolicited submissions: what the hell do you do when somebody hands you the outline for something very similar to a project you have under development?
What makes you think they are under any obligation to pay you for your suggestions without those disclaimers? The implementation of such suggestions is not protected by copyright law, and they are only protected by patent law if you file for a patent before they did (in which case Microsoft should pay).
In fact, much of the value and user friendliness (such as it is) of Microsoft products derives from the information they derive from the aggregate feedback of millions of users. That's why it is particularly cynical when people point to Microsoft and say "See how innovative they are? Look at all those things they came up with.". Any product that has a large user community and a company that listens even half-heartedly to its customers will have a big leg up in terms of features and user-friendliness on any of its competitors.
This is just the kind of thing that caused Stallman to start the GNU project and create the GPL in the first place: he believes that if you use a piece of software, you should be able to customize it to your tastes; it shouldn't be driven by the convenience, marketing whims, or fashion of whatever commercial company supplied it to you. The hope and intent is that not only you get software that you like, but that in aggregate, the preferences and customizations that people make to software will result in software that is effective and useful to its user community. Of course, 20 years ago, GNU wasn't about what happened to your digital VCR when you pushed a button on your remote control, it was what happened to your editor when you pushed the meta key, but it's roughly the same idea.
Now, you may argue about whether GNU and the GPL is the right way of achieving that and whether Stallman has been a good spokesperson for these kinds of issues. And GNU/GPL won't automatically make software easy to use for the masses of computer users who aren't sufficiently skilled to roll up their sleeves and modify it to their tastes. But, despite all its problems, I have seen no better approach so far than the approach GNU has taken.
It is, of course, ironic that the complaint about ReplayTV comes from someone who has worked for Apple for so long, a company that has taken many positions in the past that were really in conflict with the GNU project (Apple seems to have mellowed more recently, though).
to have people redistribute it for you: you want them to make it available on FTP sites and on CD-ROMs. It's the end-users that are responsible for licensing the software. Ariston's inclusion of shareware with some piece of hardware might technically violate the license or it might not, but as long as they didn't deactivate the "shareware screen", it's not a blatant and outrageous violation of shareware terms. If Warner didn't want this sort of thing to happen, he shouldn't call his software "shareware". This kind of fuzzy, borderline case has no implications for the GPL or its enforcement.
Another provision of most "shareware" is that the only charge for it can be reasonable costs of distribution and media. By bundling this software with their hardware, and by touting this software on the box, one can reasonably assume that part of the price of the hardware was going to fund the software.
One might. But one would discover that one is wrong when one actually opened the program. Then one could complain to the hardware vendor about it.
And, of course, shareware isn't a licensing term, it's more of a generic category. Different shareware can have wildly different licenses.
So what? If Warner calls it "shareware" and uses a license that actually is much more restrictive than that term implies, he has added misleading advertising to his list of problems.
I'm sorry, I don't see any reason for Warner to complain. Legally, what Ariston has done may have been marginal or even crossed the line slightly, but if Warner doesn't want people to distribute his software as shareware, he shouldn't call it "shareware". This has no implications for the GPL or other copyright enforcement for free software.
Apparently, Warner doesn't understand the concept of shareware. The idea behind shareware is to get people to distribute software freely and as widely as possible. Shareware puts the responsibility of paying for the software on the end user that actually uses the software. If Warner didn't want people to distribute his software, he shouldn't call it "shareware".
The implications of this for the GPL are exactly nil. As far as I can tell, Warner was at best unclear about what his license terms were, and probably actually just didn't understand what he was doing when he put out his software as "shareware".
Even if this were actually a case of copyright violation, the GPL isn't there to recover damages. It isn't even there as a hidden trap to force unwary companies to reveal their source code. It's there to keep the software in question open, and it seems to work reasonably well at that.
If you run a 10M emacs process, you either have to lock it into memory or you have to accept that it will occasionally get paged out and you have to wait for it to repaint its windows. QNX can't do any better in that respect than Linux.
In different words, you could set up your Linux system so that it would "feel" just like a QNX system, but most people aren't willing to live with the limitations that that imposes. The fact is that we like to use the big, featureful, complex applications that occasionally get paged out, and we just buy more memory if that gets too frequent.
So would you like to retrain the 1000's of employees my company has on how to use KDE or Gnome?
No, but I think they should never have to see the Windows desktop or deal with anything system related either.
What they should get is high-quality custom applications that let them get their work done easily. Those kinds of applications are more expensive to purchase, but they save money in the long term. You can deliver those kinds of applications most easily to a web or X11 client machine, from a mainframe or UNIX server.
Picture putting Linux on one of your sales force's desk. They wouldn't know what to do with it.
Traditionally, these kinds of applications on UNIX and mainframes are written to hide the operating system completely; the sales force would get a product that's highly customized and specific to their applications, something that's easy to administer, and requires virtually no training. They'd get something on their desk (often an X terminal) that requires no maintenance or upgrades, doesn't crash, and is up all the time.
Windows has lowered expectations. On Windows, people get a Windows desktop with Excel and some poorly written custom VisualBasic program. With Windows, your sales force has to deal with all sorts of computer machinery, moving files around, defragmenting their disk, upgrading, etc., that has nothing to do with the task at hand. To be sure, the Windows solution is lower cost in terms of purchase price, but in terms of usability, support, maintenance, or training, it is worse in pretty much every respect.
Microsoft isn't interested in just seeing a bunch of license documents, they are interested in seeing all your machines and correlating them with a bunch of license documents.
AOL has no power, they simply have a lot of clueless users. Clueless users can be helped with education and better, free software. That, rather than AIM clones, should be where open source efforts go.
How much effort and money do you think it took to develop the Internet infrastructure (software, protocols, etc.) that AOL got to use for free to build their systems around? How much advertising revenue do they derive from having tied millions of people to their substandard proprietary service, stifling the development of high-quality free services? And even when it comes to bandwidth itself, are you sure that non-AOL users don't actually end up subsidizing AOL's external traffic when all is counted up?
I think on balance, AOL has not even begun to pay their debt to the Internet; in fact, they probably still are imposing more costs on non-AOL users than they are paying. So, don't talk about "stealing" when people are trying to interoperate with their proprietary infrastructure.
But in the end, we agree: don't use AOL's services.
AOL should never have placed itself at the center of an instant messaging infrastructure in the first place. They did so out of greed, trying to hook as many users on their systems as possible. In the process they stopped the evolution of open, distributed, scalable chat systems and delivered something with a questionable business model and questionable security.
The injustice is not that UN*X clients talk to their servers and use up their bandwidth, the injustice is that AOL succeeded at inserting itself into what should have been an open part of the Internet infrastructure in the first place.
I quite agree: people shouldn't use AIM, not out of respect for AOL's bandwidth, but because AIM is a bad idea to begin with. Develop something new and better that uses entirely open protocols and fits in with the existing infrastructure; you can build on IRC or http or Gnutella or any number of other protocols and existing systems. Stop copying commercial Windows hacks and start using your imagination.
There are ways of working around the problem. But why bother? The AIM infrastructure is a proprietary, centralized service, built on proprietary software. It should never have caught on and displaced the distributed chat systems that preceded it in the first place. I'm glad to see open source support for it end; this should have happened much sooner.
How should things work? You don't go to AOL for proprietary mail service--your ISP provides you with mail service. What we need is a distributed chat protocol based on servers run by your ISPs. For example, each ISP could run an IRC server to which their users can connect, and those servers could form a global network. Now, IRC has some problems, but all that effort that has been wasted on building AOL clients could have gone into fixing those problems.
The patent describes an obvious engineering solution. Walid Tout (the "inventor") should go into the engineering hall of shame: either he wasn't smart enough to realize that this was obvious, or he was irresponsible enough to patent it anyway.
I can never quite figure out why people keep working for Microsoft. Do you believe that Microsoft actually makes worthwhile technical contributions or innovations? Do you think it is liked very much by customers? Don't you have enough money by now to retire? Or are you just waiting for your stock options to become worth something again? Even if you believe in Microsoft's mission, wouldn't you rather be in a smaller company again?
The Titanium G4 looks nice; it gets high points for style, but I don't think it's a good value. The screen is wide but pretty low resolution and there is no choice of pointing device other than the touch pad. You can get a wide choice of equivalent PC laptops for significantly less money.
The content provider lobby doesn't want you to get symmetric high speed access. After all, you might assert your fair use rights and share some CD you like with your mother. They're going to hold your ISP responsible for that, if they don't buy your ISP outright.
But worse than you doing free advertising for them by distributing their stuff to friends, you might actually shun their content completely and do something on your own. Instead of mindlessly consuming addictive and useless content while being fed advertising, you might actually engage in dialog with other human beings--no revenue opportunities there, no way to get you to consume more. You might even broadcast a school production of an out-of-copyright Shakespeare or Bach work and not have to pay anyone for it. Such civic activities would be downright un-American.
And if the excuse of bandwidth limitations doesn't work anymore (128Mbps downstream, 128kbps might start looking ridiculous to even the most clueless), restrictive terms of service will keep you from using fast uplinks for anything that might endanger anybody's revenue stream, whether you have a legal right to or not.
Much as I think CVS is a great program, for simple copying or backup, rsync or unison are likely to be more efficient and more convenient. (Rsync propagates changes in one direction, unison propagates changes in both directions.)
pvcs sucks because of the rule that the suckiest and most selfish programmer on the
team also happens to be a bright and cheery early riser,
He used to be a sad nightowl: he could never get any work done because someone else always had the files locked he needed. But then he started getting up early, got to lock his files first, and became bright and cheery.:-)
Second, I believe an algorithm is known that can do lookups in unsorted, unindexed lists in O(log(n)) time. That is certainly an interesting proposition.
I'm not sure there would be a lot of practical consequences. "Polynomial time" does not mean "tractable". And you are right that approximations are pretty good. In fact, an exact solution would only be a one-time gain, dwarfed by many more down-to-earth improvements in efficiency and technology.
In any case, whether quantum computing can solve NP-complete problems in polynomial time is an open question.
Slashdot Mirror
I have to say, though: of the currently popular scripting languages, Python seems to closest to becoming a real programming language, and it's still retaining its original utility and simplicity. What Python is missing at this point is a new implementation that provides incremental compilation to efficient native code, plus a few language cleanups to go with that.
People keep referring to these kinds of measures as "anti-piracy". Let's call it by its true name: "anti-fair-use". That's because there is no known technological measure that can distinguish between fair use and piracy. Companies have merely tried to preemptively redefine "piracy" as any kind of copying, even copying that traditionally falls under fair use.
Dylan is in the same boat with Smalltalk, CommonLisp, Scheme, and Self: great languages with great technology, but they are little used among UNIX and Linux users to do the kinds of things Perl or Python are used for.
To me, things will get interesting once one of these languages actually starts delivering a non-trivial, high-performance native code compiler. I don't mean the kind of simple translation to C code that exists for Perl and Python, but something that actually genuinely increases performance and figures out stuff about data types. Will anybody be up to the challenge? That remains to be seen.
Until then, I won't hold my breath for Perl6. In fact, I don't much care. Perl5 is good enough for what I use it for, and there are lots of other choices around when Perl5 isn't good for the task.
What makes you think they are under any obligation to pay you for your suggestions without those disclaimers? The implementation of such suggestions is not protected by copyright law, and they are only protected by patent law if you file for a patent before they did (in which case Microsoft should pay).
In fact, much of the value and user friendliness (such as it is) of Microsoft products derives from the information they derive from the aggregate feedback of millions of users. That's why it is particularly cynical when people point to Microsoft and say "See how innovative they are? Look at all those things they came up with.". Any product that has a large user community and a company that listens even half-heartedly to its customers will have a big leg up in terms of features and user-friendliness on any of its competitors.
Now, you may argue about whether GNU and the GPL is the right way of achieving that and whether Stallman has been a good spokesperson for these kinds of issues. And GNU/GPL won't automatically make software easy to use for the masses of computer users who aren't sufficiently skilled to roll up their sleeves and modify it to their tastes. But, despite all its problems, I have seen no better approach so far than the approach GNU has taken.
It is, of course, ironic that the complaint about ReplayTV comes from someone who has worked for Apple for so long, a company that has taken many positions in the past that were really in conflict with the GNU project (Apple seems to have mellowed more recently, though).
to have people redistribute it for you: you want them to make it available on FTP sites and on CD-ROMs. It's the end-users that are responsible for licensing the software. Ariston's inclusion of shareware with some piece of hardware might technically violate the license or it might not, but as long as they didn't deactivate the "shareware screen", it's not a blatant and outrageous violation of shareware terms. If Warner didn't want this sort of thing to happen, he shouldn't call his software "shareware". This kind of fuzzy, borderline case has no implications for the GPL or its enforcement.
One might. But one would discover that one is wrong when one actually opened the program. Then one could complain to the hardware vendor about it.
And, of course, shareware isn't a licensing term, it's more of a generic category. Different shareware can have wildly different licenses.
So what? If Warner calls it "shareware" and uses a license that actually is much more restrictive than that term implies, he has added misleading advertising to his list of problems.
I'm sorry, I don't see any reason for Warner to complain. Legally, what Ariston has done may have been marginal or even crossed the line slightly, but if Warner doesn't want people to distribute his software as shareware, he shouldn't call it "shareware". This has no implications for the GPL or other copyright enforcement for free software.
The implications of this for the GPL are exactly nil. As far as I can tell, Warner was at best unclear about what his license terms were, and probably actually just didn't understand what he was doing when he put out his software as "shareware".
Even if this were actually a case of copyright violation, the GPL isn't there to recover damages. It isn't even there as a hidden trap to force unwary companies to reveal their source code. It's there to keep the software in question open, and it seems to work reasonably well at that.
In different words, you could set up your Linux system so that it would "feel" just like a QNX system, but most people aren't willing to live with the limitations that that imposes. The fact is that we like to use the big, featureful, complex applications that occasionally get paged out, and we just buy more memory if that gets too frequent.
No, but I think they should never have to see the Windows desktop or deal with anything system related either.
What they should get is high-quality custom applications that let them get their work done easily. Those kinds of applications are more expensive to purchase, but they save money in the long term. You can deliver those kinds of applications most easily to a web or X11 client machine, from a mainframe or UNIX server.
Traditionally, these kinds of applications on UNIX and mainframes are written to hide the operating system completely; the sales force would get a product that's highly customized and specific to their applications, something that's easy to administer, and requires virtually no training. They'd get something on their desk (often an X terminal) that requires no maintenance or upgrades, doesn't crash, and is up all the time.
Windows has lowered expectations. On Windows, people get a Windows desktop with Excel and some poorly written custom VisualBasic program. With Windows, your sales force has to deal with all sorts of computer machinery, moving files around, defragmenting their disk, upgrading, etc., that has nothing to do with the task at hand. To be sure, the Windows solution is lower cost in terms of purchase price, but in terms of usability, support, maintenance, or training, it is worse in pretty much every respect.
Microsoft isn't interested in just seeing a bunch of license documents, they are interested in seeing all your machines and correlating them with a bunch of license documents.
AOL has no power, they simply have a lot of clueless users. Clueless users can be helped with education and better, free software. That, rather than AIM clones, should be where open source efforts go.
I think on balance, AOL has not even begun to pay their debt to the Internet; in fact, they probably still are imposing more costs on non-AOL users than they are paying. So, don't talk about "stealing" when people are trying to interoperate with their proprietary infrastructure.
But in the end, we agree: don't use AOL's services.
The injustice is not that UN*X clients talk to their servers and use up their bandwidth, the injustice is that AOL succeeded at inserting itself into what should have been an open part of the Internet infrastructure in the first place.
I quite agree: people shouldn't use AIM, not out of respect for AOL's bandwidth, but because AIM is a bad idea to begin with. Develop something new and better that uses entirely open protocols and fits in with the existing infrastructure; you can build on IRC or http or Gnutella or any number of other protocols and existing systems. Stop copying commercial Windows hacks and start using your imagination.
How should things work? You don't go to AOL for proprietary mail service--your ISP provides you with mail service. What we need is a distributed chat protocol based on servers run by your ISPs. For example, each ISP could run an IRC server to which their users can connect, and those servers could form a global network. Now, IRC has some problems, but all that effort that has been wasted on building AOL clients could have gone into fixing those problems.
The patent describes an obvious engineering solution. Walid Tout (the "inventor") should go into the engineering hall of shame: either he wasn't smart enough to realize that this was obvious, or he was irresponsible enough to patent it anyway.
I can never quite figure out why people keep working for Microsoft. Do you believe that Microsoft actually makes worthwhile technical contributions or innovations? Do you think it is liked very much by customers? Don't you have enough money by now to retire? Or are you just waiting for your stock options to become worth something again? Even if you believe in Microsoft's mission, wouldn't you rather be in a smaller company again?
The Titanium G4 looks nice; it gets high points for style, but I don't think it's a good value. The screen is wide but pretty low resolution and there is no choice of pointing device other than the touch pad. You can get a wide choice of equivalent PC laptops for significantly less money.
But worse than you doing free advertising for them by distributing their stuff to friends, you might actually shun their content completely and do something on your own. Instead of mindlessly consuming addictive and useless content while being fed advertising, you might actually engage in dialog with other human beings--no revenue opportunities there, no way to get you to consume more. You might even broadcast a school production of an out-of-copyright Shakespeare or Bach work and not have to pay anyone for it. Such civic activities would be downright un-American.
And if the excuse of bandwidth limitations doesn't work anymore (128Mbps downstream, 128kbps might start looking ridiculous to even the most clueless), restrictive terms of service will keep you from using fast uplinks for anything that might endanger anybody's revenue stream, whether you have a legal right to or not.
Much as I think CVS is a great program, for simple copying or backup, rsync or unison are likely to be more efficient and more convenient. (Rsync propagates changes in one direction, unison propagates changes in both directions.)
He used to be a sad nightowl: he could never get any work done because someone else always had the files locked he needed. But then he started getting up early, got to lock his files first, and became bright and cheery. :-)
Grover's algorithm is O(n^0.5).
In any case, whether quantum computing can solve NP-complete problems in polynomial time is an open question.