You did work under contract, an appeared to do something not in their interest... i.e. maintain copyright and release under the GPL. Well, give them an unlimited license. I mean, by all rights they should own copyright, so give them what they need.
Additionally, the legal issues here are silly. This is the military, they have lots of guns, therefore, they make the rules.
Regardless, nobody here thinks that this was ethical of you. If you used GPL code (and can justify the GPL for that reason), then inform them of this and what that means. As they are unlikely to distribute, they won't care. If the code is being turned over to a company to make a program to distribute to the military, inform them of the situation. They can keep the system under the GPL by including a source code CD to the military and contract them into including a copy if they distribute it.
Contrary to/. wisdom, the GPL does not require a company to put there software for download, in CVS, etc. They only need to make source available to a third party if they don't include it originally.
I mean, the GPL doesn't limit companies doing a custom job... it only limits companies trying to do a "mass market" job... which makes sense, the Free world Stallman refers to was a world of custom code, and with custom code, there is no reason not to make source available (that way you don't need to be the only one that can make changes, you can ethically turn down the project). Generic mass-market applications (Office suites, GUI OSes, etc) have an incentive to keep proprietary, because you can charge per copy.
While there are other "clients" that code for the DoD can be sold to, charges like treason come to mind.:)
I really want to meet someone that can play a game of Heroes of Might and Magic in 8 hours... must be a genius to think and act that fast...
My games (2-3 players) will go 18-24 hours, easily... but 3 people on 5 computers with the max difficulty takes a while...:) Especially with enough cigarette and pizza breaks.
I got Heroes III Complete sitting on my desk, and it the game I play most often. They had a successful game, 2 expansions, and a series of standalone campaigns that seems to be selling. They are supposedly working on Heroes IV. Turn based games still exist, because some people prefer them. They are easier to write and have a following.
The most important part of multi-platform development, as this showed, was it catches programming errors. The errors that show up on ports are the same errors that show up on weird drivers, etc.
Yes, Linux gaming is good. Mac gaming is good. Anything that fights MS monopoly power is good.
Whether you like Windows or not, MS Monopoly power SUCKS.
The claims are OR'd together, the subparts of the claim are AND'd together.
Read the claims.
Each one involves the four characteristics I outlined... the different in the claims is predominately what is logged and HOW the log files are transmitted.
It's a more specific patent than you think.
I don't believe that any GET URL with a sender and recipient is a violation, I don't believe that it falls under any of the claims.
What are you talking about that Linux isn't the best Unix, it's more stable than Windows 98!:)
Yeah, it's a shame how much respect Linux gets compared to the actual Unixes. Linux makes Unix available to the masses, in a way the BSDs don't. If I hadn't been running a few Linux boxes for a few years, OpenBSD would have been impossible to pick up.
Besides, NT already has a POSIX layer, it's just incomplete. As we saw with Interix, getting a complete POSIX layer on NT isn't that tough, if they want to go down that route.
I mean, comparing Windows to Linux is silly. Windows is a desktop operating system where stability is traded for features. The comparison of NT to Linux is closer, but still silly. NT aims for a reasonable degree of stability but trades some for features. The software for NT massively beats out the "Linux" software.
The fair comparison of Linux is other Unixes, where Linux doesn't compare all that much. I mean, Redhat is a competitor of MS in the CE market for Kiosks, etc., but not on the desktop. MS really lacks a desktop competitor outside of Apple, which isn't significant unless OS X gets Apple really back in the game.
Server side, the only NT vs. Linux comparison is IIS on NT vs. Apache on Linux, the rest of the software is SO vastly different.
If you read the claims, it seems like a generally limited scope stupid patent. The problem is that this fails the non-obvious requirement and the Constitutional advancedment arguement.
Trade Secret laws are the problems, not patents. Anything that SHOULD be patented (monopoly for releasing information the society wouldn't get) is now protected under Trade Secret laws. As a result, the company gets rediculous protection... there is no societal reason for Trade Secret laws, it simply allows corporations to make more money.
This sort of system needs to be revealed to be used. Therefore, the patent is silly. Furthermore, the decision to put up a system like this is NOT related to the ability to patent it, so society is not advanced by the patent, the system would be released anyway and society would gain the knowledge.
HOWEVER, this system is pretty unique, so the patent is less of an issue. Read the claims, the logs portion of it is REALLY significant. They didn't patent GETs:)...
To fall under the patent's claims, all 7 of them, you pretty much need to have a system that does the following:
1. Accept submissions to your server
AND
2. Create a custom URL for EACH receipient of the data (meaning, if I send 1 URL to 5 people, I don't fall under the patent)
AND
3. The server logs all accesses (it would seem to reason that these aren't web logs, but database logs with specific access information)
4. Transmits the logs to the sender (there are multiple ways of this, on request, automatically, without confirmed requiest, etc., etc., etc.)
I mean, that is a relatively specific approach to things.
Additionally, it becomes questionable if the GET string qualifies as a unique URL.
For example, if I send the same URL (script access) with a different set of variables, I should potentially be able to escape the claims.
If you are doing something like this, it isn't THAT hard to work around the patent.
The problem with these patents, however, is that by merely reading the claims, one could design the system trivially with a mastery of the skill. I have no need to read the patent and gain knowledge (that I can put to use in 20 years).
OTOH: when they came up with this in 97, this was pretty unique. Now it seems very commonplace, and there probably is prior art for some of this. I also doubt that they would try to enforce these claims anyways, given how sketchy the patent seems today.
For example, when I post to USENET, I have sent something to many receipients. Deja/Google indexes this information, and sends it to other receipients. The log files are then sent me?
Nope, I don't get a log of every response to my message.
Don't get me wrong, this is a silly patent, but not QUITE as silly as you seem to think.:)
All the GPL being outlawed (which wouldn't happen, it would be more subtle), we wouldn't lose much. We'd rerelease it under the GPL. Very little is actually GPL'd because of the GPL. It is much more common for people to avoid the GPL because of it.
The Open Source community cares little for Free Software, but makes different demands.
If I announced that I was part of the "community" and would sell GPL licenses (which I can do, I just can't stop you from redistributing), you'd al flame me, while I'd be legally in the right.
The right to a downloadable ISO, web site, and ftp site, is NOT a right that the GPL ennumerates.
Why is this a war. RMS has the right idea. He didn't go and talk Netscape into open sourcing Navigator. He didn't talk AT&T into openning Unix, he went and wrote his own.
We need more coding, less Slashdot ranting. The goal isn't world domination. If you want the freedoms that the GPL provides, only use GPL'd (or less restrictive, a la BSD) licensed sosftware. If you want global domination, go grab your dice and a war game and play, otherwise stop with the mental masturbation.
RMS gets credit for many things. He is an idealist and DID something to make his ideals work. ESR is an opportunist who wrote an completely preposterous essay and tried to steal the thunder of someone actually doing something.
The whole "open source" process fails for large projects. All the significant projects are run by a small group. While Linux has many "contributers" the core team is small, and Linux's development is slowing because of the management issues.
The best free software (Emacs, Apache, Perl, Samba, GCC, etc.) are ALL small teams or individuals writing great code. Even if they get contributions, the direction is set by a small group. The "Open Source" projects that ESR talks about are among the weaker links (Mozilla is the prime example, but I'd argue that Linux falls in here... all the stability arguements are in contrast to Windows, which by engineering tradeoff has more features and less stability, not against other Unixes. Does anyone here REALLY feel that Linux stacks up against HP-UX, Solaris, AIX, or BSD 4.4 Lites in stability? I mean the kernel, not the mess of software).
The Mythical Man Month holds...
Open Source is the trojan horse to destroy Free Software. By focusing on the convenient instead of the ideals, once the Millions of Eyeballs theory falls to the Mythical Man Month, then Open Source becomes passe.
Show RMS some respect... y'all run his code, use his license, and piss on his work because it isn't convenient to placate your corporations?
As a result of OS X, I'm switching to a Mac. I haven't owned an Apple since an Apple//c, but I can't wait to be able to afford my new G4 Powerbook.
OS X seems incredible, and since I've switched from Linux -> BSD (Open), I've been a happy camper. OS X seems like it will let me play nice with all my applications, including the important option of coding for deployment on my BSD boxes.
I don't have time to hack Darwin, but I'm glad they openned it. Assuming they got a single modification submitted that was worthwhile, openning it was good, because it means when OS X runs on my shiny new laptop, it'll be better.
Slashdot ignorance is aggravating, but you take the good with the bad.
This is slight more newsworthy than the usual "Linux is the best OS ever, I mean, it's more stable than Windows 98!" mental masturbation that sits on the comment sections these days.
Alex
ESR never was relevant...
on
ESR On XML-RPC
·
· Score: 2
I don't get how he ever became relevant. He is definitely involved in some coding, but relatively trivial. He was the first person to decide that the secret to Free Software was for businesses to get involved. Additionally, he created the concept (in words) of Open Source development which has YET to succeed.
He talked Netscape into the Mozilla project. While I actually somewhat enjoy Mozilla, I find no reason to switch from IE at this time, and Mozilla shipped as Netscape 6 two years after starting, when his theory was that it would reduce development time.
There are two theories, the mythical man month and the Bazarre theory. The Mythical Man Month was written after the management of a HUGE software project. The Bazarre theory was written after writing an e-mail downloader and a whim... As someone who has managed small scale projects, the mythical man month is totally true.
I never understood how he became a spokesman. I don't agree with ANY of RMS's political theories, but I like his theories on software. I think that his contributions (GNU, the GPL, Emacs) are HUGE, while ESR wrote some essays.
While both the tech writer and engineer are important to a project, the engineer is more useful. You can run a project without a tech writer (you'll suffer later, but it can be done), but try an engineering project without an engineer.
His code contributions while nice, are trivial compared to the air time/free money he has received.
His code are extensions on GPL'd code. Yeah. The GPL is not magic. It does not declare code free. Code is copyrighted by the author, the GPL is a possible license.
Note: IANAL, but I have read the GPL and read some case law on copyrights and licenses.
He owns his extensions. The company is suggesting that they become their property (if he works on them while there). If he agrees, then they own his extensions. They can release them under the GPL or not. He has NO say.
However, they cannot build a derivative work (the GPL code + his extensions) and distribute it without meeting the requirements of the GPL. However, they can do WHATEVER they want with his extensions, including keeping them proprietary. The combination of GPL code + his extensions is a derivative work to the GPL code. His extensions are NOT a derivative work, they are an original work.
There is no magic to the GPL that says you must do something. If I take MS code, slap a GPL on it, and release it, it isn't GPL because the license has no meaning, I don't own the work.
In this scenario, the code becomes Company X code (again, assuming he improves it while there), who can release it or not under any license.
But he already released it under the GPL? Irrelevant. Those that have already obtained the code under the GPL from him are able to use it (and redistribute), but he could not. The GPL isn't magic, it is a way of licensing software.
You CAN reclose GPL code, but the code released under the GPL already is floating around under the GPL. However, if he signs it over to the company, who prohibits him from releasing it under the GPL, he can't release it, he doesn't own it.
Now it would be interesting for him to further relicense that code from a third party (under the GPL), but I'm not convinced that that would work given that he signed away all rights to it. However, if people haven't maintained copies with the licenses, he is SOL, because the company's version is no longer GPL'd.
Guys, the law is NOT code. You don't write an algorithm. There is judgement and grey.
For example, if you are providing a channel, or hosting a site entitled Kittie Porn, you should be responsible. If you just pull everything down, you shouldn't. It's a judgement thing.
For example, Napster enables MP3 sharing. In and of itself, this is innocent. However, Napster will scan your hard drive for you and offer to share out all your MP3s. The problem with this is that while it is legal for me to rip a CD that I own (or I'll argue, a friend owns) for my own purposes, I can't legally share it out. Napster encourages you to do so. Intent matters.
For example, if google happens to index a kiddie porn site, they didn't intentionally do it and shouldn't be responsible.
If I provide a site that is a directory of kiddie porn sites, I SHOULD be responsible, even if I'm just linking. If I provide a link to a site that is legal porn, and they add kiddie porn, I shouldn't be.
Basically, if you are intentionally doing something shady, you are responsible. If something happens beyond your control, safe harbor should protect you.
I went to a prep school for my last three years of high school, where I continued to slack in my classes. I did well in extra-curriculars, but I was a mediocre student. However, I rocked my SATs, SAT IIs, and AP tests. I used this to show schools that I was a capable individual and got an opportunity to attend an elite school.
Does this jive with America's sense of the Protestant Work Ethic and working hard to achieve? Absolutely not. Did I do reasonably well in College? Yes, although a little more effort and I'd have done well. However, my grades alone wouldn't have gotten me in.
I'm a middle-class white student, I'm NOT who he wants to help, I'm who he wants to hurt.
Here is the problem though. I went to a prep school with a decent reputation among many schools throughout the nation. As a result, the entire top 20% of my school attended elite schools. My school's reputation got relatively dumb kids with high GPAs at a respectable school and mediocre test scores into good schools.
Those same kids in a local Public school no doubt would have had high GPAs and low test scores, but not been considered, because the low test scores would show that they weren't that swift.
All in all, this sounds good, right?
No. Without test scores, the only means of evaluating GPA is to look at the schools reputation. While this is fine for Private Schools or elite Public schools (magnets focusing on education in the South, special enterance exams in northeastern metropolitan cities, for example), but what about the poor kids from unknown schools?
Yeah, they'll get a 4.0 unweigthed GPA and rank in the top five or whatever at their school, but do you accept them? At a school you've never heard of in a poor area, you can assume that a 4.0 isn't of reasonable caliber for your school. What do you use, AP classes? If you're from a middle class suburb or attend a private school, AP classes may be the answer. If you're from a poor school area, you don't have that option.
Yes, money helps. You can't change that. What you CAN do is give people without money an opportunity. The kids with the expensive education and test prep have an advantage on the tests, but it isn't a guarantee.
In a scenario without the tests, the kids HAVE a guarantee. They are the only ones from high schools that schools have heard of. Without the tests, your poor kid with the 4.0 but no APs because they aren't offered and no SAT scores to back him up doesn't get in, but the kids from the private schools with 3.6 GPAs with a reputation for excellence do.
That isn't useful.
Admissions already adjusts for background. The UC system is absurd, allowing admission by test scores alone, which reduces their paperwork, but doesn't help. Their admissions process is too numeric, but they are processing too many applications for hands on evaluations. Schools normally have some kind of cut where they reduce their applicant pool by screening out a certain GPA range. Then admissions officers can make decisions. The poor kid with a 4.0 from a crappy school with a 1300 GPA is probably the equivalent of the wealthy kid with a 1400 GPA based upon the extra training and education. They CAN take that into account. They can't take race into account, but they can figure out the student's background. A poor white kid OR poor black kid should have that taken into account. A rich white kid OR rich black kid should have their background into account.
No, test scores don't substitute for discretion. But they allow you to justify things. If the kid with a 4.0 GPA had a 1000 SAT, he probably doesn't get in. But with a 1300, he does. Why? In the first case, you can't confirm his aptitude, in the second you can. Without the tests, you CAN'T confirm his aptitude AT ALL. You'd assume he is a 1000 SAT kid because he's from a crappy school and admit the known candidate from the prep school.
This man HAS NOT thought through his situation. They need to add more discretion, but they don't need to scrap the best means for helping people out.
This isn't exactly an Open Source business, but a business catering to the "Open Source" community that is in trouble. Much like Tucows BSD section, the rest of the computing world really can't get in.
When I was an NT Administrator, I got Windows NT Magazine, and so did everybody I know. They had GREAT articles and tips that would help any NT Admin out in running a site.
However, the Linux group doesn't play by the same rules. I occaisionally buy a book (O'Reilly's SAMBA book comes to mind), but that's about it. The rest of the time I use online docs. For my BSD boxes, I find that the man pages are the most useful things. I never think to get a Linux magazine.
I think that it's a cultural thing. In the Linux world, I don't feel "mainstream" and don't feel like buying mainstream publications. I don't think to spend money.
For programmers, Open Source can work. I know that I've hired Open Source program writers to install applications, etc. (I mean, for corporate use, when my guys are worth $X/hr, and it will take Y hrs to learn and do it, if X*Y > Consultant fee... I hird consultants), but for people catering to hobbyiests, it may be the end.
The demographic that used to support all the PC hobbyist books/magazines are pretty much all using Linux. It was a group of tweakers, and you can't tweak Windows anymore (well, some, but it's not useful). That market is dead, replaced by websites and newsgroups.
I just wonder, why do Windows magazines/books sell and Linux ones don't? Is it just being abnormal?
I don't see anything in there giving authorization of the Trademark...
If you're not compatible, you can't use the name.
That doesn't mean you can use the registered mark. When that was done, there was no registered mark. The Mark is owned by a corporation (founded by this gentlemen, not that it's terribly relevant), and can't be appropriated without authorization.
For example, if I write some code, call my program Microsoft, and say in my license that you may use the name Microsoft in referring to my code, you aren't authorized to use it.
That is hardly a license to a Trademark issued two years after that was written.
The author of the original software asked you not to use the name ssh or Secure Shell if they are incompatible. That is a contractual agreement as part of a license.
The company is using the Trademark SSH to refer to their company and software. When you used that license, you contracted to not name is SSH if it is not compatible. That has NO bearing on the trademark.
The statement that the mark wasn't enforced was the WRONG tack to take.
Good job. We have now taken the position to the outside world of being total assholes.
The guy made an effort NOT to bully an Open Source group. He didn't send threatening lawyer letters. He asked people to be reasonable. Quite frankly, you're going to lose BADLY in a court of law, because not only is there the possibility of confusion, but there is DOCUMENTED confusion.
Now, this is totally absurd.
He didn't bully, DON'T use that against him. The arguement that he didn't bring it up until confusion happened may be legally correct, but now sets the horrible precedent that Open Source groups will be as shady as possible, and if you give them an inch, they'll take a mile.
Notice something, the original license (if anything tells people not to be nice and release their source code, this is the example, that was hardly a license, more of a nice gesture) allowed you to call your application ssh.
That's irrelevant.
OpenSSH could be renamed, OpenSecSH, which would be an open implementation of the SecSH protocol, which is the name of the working group. The FILE could still be called ssh/sshd.
SAMBA couldn't call itself SMB, confusion reasons, but the applications use SMB (like, smbd).
However, you can be polite, compile as osecsh and osecshd and include a symlink (automatically, but prompted) for ssh and sshd, so if you have both implementations, you can decide which one you want.
However, if the Open Source community insists on fighting on the Trademark grounds, we're in the wrong.
You can dispute the merits of Software Patents.
You can dispute the merits of long copyright terms.
You can dispute the merits of copyrights in general.
You can't dispute the merits of Trademarks.
Trademarks are the only thing that prevents confusion in the marketplace. If people are confused and think that OpenSSH is from SSH, then there is a legitamate issue.
Also, I don't think you're going to win on the enforcement. One year or so is reasonable, given that OpenSSH has minimal press coverage, etc. I think that it would be EASY for SSH to show that they found out about it, saw a problem, and then asked them to fix it.
Theo, you're going to lose, and you're being a bastard. He hasn't demanded that you stop making a version of SSH, just that you not use his product name.
Theo, I'm an OpenBSD user. I love OpenBSD. I love OpenSSH. You're in the wrong here... VERY WRONG.
Alex
Remembering is Copying...
on
License to Sit
·
· Score: 2
Kinda clever, however, isn't there case law that indicates that "copying to memory" is fair use? I believe software companies tried to use the copying to memory to get copyright law to cover the use of the software and the courts told them to blow off.
I also think that "copying to hard drive" falls under fair use, but less relevant to the satire.
I do think that the cracks about the Government ID card are extremely funny and amusing. This is overall a good satire, thanks/.
Most of country is a little less anti-business then California. Besides, California's laws would probably fail on Federal Appeal, as a restriction of Free Speech. The company has as much free speech as the individual.
Federal law doesn't cover this stuff, only prohibits certain kinds of discrimination. Most states allow corporations reasonable leeway.
Sorry dude, Linux is the ugliest designed Kernel ever. It ignores the last thirty years of software engineering, and puts forth a technologically primitive system. I would argue that the NT Kernel is more advanced than it.
Microsoft broke the NT Kernel's elegance in 4.0 and 5.0, but it was a reasonably well designed kernel.
I mean, the reason I run Linux is that it is "fun" and I can get precompiled things like KDE right out of the box (and I don't get funky things like Cronolog not compiling on my machine like I'm picking up with OpenBSD).
If I don't run Redhat, I lose these advantages.
But, I pick up a poorly designed kernel.
Thanks, but no thanks. I think that Linux is the best hacked together kernel, but I find the design really inelegant.
Besides, I have no time on my hands, and this isn't on the top of my todo list. However, I'll check out the site and keep it in mind in the future, thanks.
After getting hacked, we switched our servers from Redhat to OpenBSD. Part of this was the security audit/out of box security, and part was the belief that a more obscure system is safer. We got his with a rootkit, and avoiding Redhat/NT means we are out of the mainstream and less likely to get slammed.
I would never go back. We still have Linux boxes for Development because the Redhat KDE Workstation install is a dream, but for a server, wow, I love BSD.
The BSD methodology is more sane. The startup process is cleaner and easier to manage. My system is more organized. The Ports are terrific.
For a Workstation where I want the newest toys, I might use Linux because it always gets the whiz-bang features, although I may toss FreeBSD on a box to play with.
However, with OpenBSD I know what's up, with Linux, I find that millions of things install everywhere. Besides, my OpenBSD FTP install now takes about 10 minutes after I learned the install process, the Redhat one still takes forever with it crashing about half the time.
I realize that Redhat != Linux, but it is the most common distribution and aims for the corporate market. I feel like the Redhat system is held together with chewing gum with rediculous numbers of scripts that are hacks to hold different systems together. I should try Debian, but OpenBSD has done what I want and I look foward to trying FreeBSD. I've done NetBSD and OpenBSD systems, and they have been less aggravation than Linux installs.
Um, bzzzt, wrong. SCO isn't pretty, or sexy, but they sell a LOT of licenses. They are a UNIX on Intel, and have a large business presence.
There servers have never been the most powerful (running on Intel Architecture, which until 2-3 years ago was seriously sucking compared to real Iron), but they have sold a lot of licenses.
The big Unix shops used to make a lot of money from licenses, but they were fewer licenses more $/license.
Not getting hype doesn't make you insignificant.
Hell, Solaris had limited presence in the Server room until the Internet boom, because their Hardware was considered substandard to IBM and HP stuff. However, they were a workstation player that gave their machines to engineering schools on the cheap, so people equated Sun with Unix.
Marketing gives the perception of marketshare, not the reality.
I'm sure this will open up a feeding frenzy to post this on Slashdot, but some of NT Admins can't do that. While some consolidate on 1 or 2 boxes, a well designed NT Network will host at least 3 servers, and they may only have a few dozen users. For example, you should have a PDC, BDC, and Exchange Server. If you combine the functions (which I've done at small sites), you open yourself up to headaches later on.
Besides, my PDC/BDCs get hammered as people log in, but there is little login during the day. Exchange doesn't get hit much at night, as only the coders are logged in and less e-mail is sent. This is a silly arguement.
This list is to keep bug information to people that program with Bind. For Example, if you're a Redhat user, you will know that Redhat will buy into this, and therefore the Redhat patch will be out before the exploit becomes common knowledge and there is a downloadable kit that trashes a machine on it.
Combine this with Redhat's automatic update services, and you have used this to really help the business model.
If you are programming on Bind, you will be on this list (or should be on this list). If you just install Bind and don't touch the code, getting exploit information is of marginal benefit. However, giving info on the exploit makes it easy for someone who doesn't know much to code up an auto-exploiter.
This allows vendors to bust their ass on a fix before the vulnerability gets out.
There is an exception here. Right now, when there is a problem, people are in a frenzy to fix it. If this allows them to not feel as pressured to fix the bug, then this is very counter productive.
Slashdot Mirror
You did work under contract, an appeared to do something not in their interest... i.e. maintain copyright and release under the GPL. Well, give them an unlimited license. I mean, by all rights they should own copyright, so give them what they need.
/. wisdom, the GPL does not require a company to put there software for download, in CVS, etc. They only need to make source available to a third party if they don't include it originally.
:)
Additionally, the legal issues here are silly. This is the military, they have lots of guns, therefore, they make the rules.
Regardless, nobody here thinks that this was ethical of you. If you used GPL code (and can justify the GPL for that reason), then inform them of this and what that means. As they are unlikely to distribute, they won't care. If the code is being turned over to a company to make a program to distribute to the military, inform them of the situation. They can keep the system under the GPL by including a source code CD to the military and contract them into including a copy if they distribute it.
Contrary to
I mean, the GPL doesn't limit companies doing a custom job... it only limits companies trying to do a "mass market" job... which makes sense, the Free world Stallman refers to was a world of custom code, and with custom code, there is no reason not to make source available (that way you don't need to be the only one that can make changes, you can ethically turn down the project). Generic mass-market applications (Office suites, GUI OSes, etc) have an incentive to keep proprietary, because you can charge per copy.
While there are other "clients" that code for the DoD can be sold to, charges like treason come to mind.
I really want to meet someone that can play a game of Heroes of Might and Magic in 8 hours... must be a genius to think and act that fast...
:) Especially with enough cigarette and pizza breaks.
My games (2-3 players) will go 18-24 hours, easily... but 3 people on 5 computers with the max difficulty takes a while...
I got Heroes III Complete sitting on my desk, and it the game I play most often. They had a successful game, 2 expansions, and a series of standalone campaigns that seems to be selling. They are supposedly working on Heroes IV. Turn based games still exist, because some people prefer them. They are easier to write and have a following.
The most important part of multi-platform development, as this showed, was it catches programming errors. The errors that show up on ports are the same errors that show up on weird drivers, etc.
Yes, Linux gaming is good. Mac gaming is good. Anything that fights MS monopoly power is good.
Whether you like Windows or not, MS Monopoly power SUCKS.
Umm, unless the post is a 1-2 line Linux Rulez post, those always get a 3 or 4... sometimes a 5. :)
But I guess that is a "fact" on Slashdot, so...
Alex
The claims are OR'd together, the subparts of the claim are AND'd together.
Read the claims.
Each one involves the four characteristics I outlined... the different in the claims is predominately what is logged and HOW the log files are transmitted.
It's a more specific patent than you think.
I don't believe that any GET URL with a sender and recipient is a violation, I don't believe that it falls under any of the claims.
You must violate an ENTIRE claim.
What are you talking about that Linux isn't the best Unix, it's more stable than Windows 98! :)
Yeah, it's a shame how much respect Linux gets compared to the actual Unixes. Linux makes Unix available to the masses, in a way the BSDs don't. If I hadn't been running a few Linux boxes for a few years, OpenBSD would have been impossible to pick up.
Besides, NT already has a POSIX layer, it's just incomplete. As we saw with Interix, getting a complete POSIX layer on NT isn't that tough, if they want to go down that route.
I mean, comparing Windows to Linux is silly. Windows is a desktop operating system where stability is traded for features. The comparison of NT to Linux is closer, but still silly. NT aims for a reasonable degree of stability but trades some for features. The software for NT massively beats out the "Linux" software.
The fair comparison of Linux is other Unixes, where Linux doesn't compare all that much. I mean, Redhat is a competitor of MS in the CE market for Kiosks, etc., but not on the desktop. MS really lacks a desktop competitor outside of Apple, which isn't significant unless OS X gets Apple really back in the game.
Server side, the only NT vs. Linux comparison is IIS on NT vs. Apache on Linux, the rest of the software is SO vastly different.
If you read the claims, it seems like a generally limited scope stupid patent. The problem is that this fails the non-obvious requirement and the Constitutional advancedment arguement.
:)...
Trade Secret laws are the problems, not patents. Anything that SHOULD be patented (monopoly for releasing information the society wouldn't get) is now protected under Trade Secret laws. As a result, the company gets rediculous protection... there is no societal reason for Trade Secret laws, it simply allows corporations to make more money.
This sort of system needs to be revealed to be used. Therefore, the patent is silly. Furthermore, the decision to put up a system like this is NOT related to the ability to patent it, so society is not advanced by the patent, the system would be released anyway and society would gain the knowledge.
HOWEVER, this system is pretty unique, so the patent is less of an issue. Read the claims, the logs portion of it is REALLY significant. They didn't patent GETs
To fall under the patent's claims, all 7 of them, you pretty much need to have a system that does the following:
1. Accept submissions to your server
AND
2. Create a custom URL for EACH receipient of the data (meaning, if I send 1 URL to 5 people, I don't fall under the patent)
AND
3. The server logs all accesses (it would seem to reason that these aren't web logs, but database logs with specific access information)
4. Transmits the logs to the sender (there are multiple ways of this, on request, automatically, without confirmed requiest, etc., etc., etc.)
I mean, that is a relatively specific approach to things.
Additionally, it becomes questionable if the GET string qualifies as a unique URL.
For example, if I send the same URL (script access) with a different set of variables, I should potentially be able to escape the claims.
If you are doing something like this, it isn't THAT hard to work around the patent.
The problem with these patents, however, is that by merely reading the claims, one could design the system trivially with a mastery of the skill. I have no need to read the patent and gain knowledge (that I can put to use in 20 years).
OTOH: when they came up with this in 97, this was pretty unique. Now it seems very commonplace, and there probably is prior art for some of this. I also doubt that they would try to enforce these claims anyways, given how sketchy the patent seems today.
Alex
Not really, the log data is relevant.
:)
For example, when I post to USENET, I have sent something to many receipients. Deja/Google indexes this information, and sends it to other receipients. The log files are then sent me?
Nope, I don't get a log of every response to my message.
Don't get me wrong, this is a silly patent, but not QUITE as silly as you seem to think.
All the GPL being outlawed (which wouldn't happen, it would be more subtle), we wouldn't lose much. We'd rerelease it under the GPL. Very little is actually GPL'd because of the GPL. It is much more common for people to avoid the GPL because of it.
The Open Source community cares little for Free Software, but makes different demands.
If I announced that I was part of the "community" and would sell GPL licenses (which I can do, I just can't stop you from redistributing), you'd al flame me, while I'd be legally in the right.
The right to a downloadable ISO, web site, and ftp site, is NOT a right that the GPL ennumerates.
Why is this a war. RMS has the right idea. He didn't go and talk Netscape into open sourcing Navigator. He didn't talk AT&T into openning Unix, he went and wrote his own.
We need more coding, less Slashdot ranting. The goal isn't world domination. If you want the freedoms that the GPL provides, only use GPL'd (or less restrictive, a la BSD) licensed sosftware. If you want global domination, go grab your dice and a war game and play, otherwise stop with the mental masturbation.
RMS gets credit for many things. He is an idealist and DID something to make his ideals work. ESR is an opportunist who wrote an completely preposterous essay and tried to steal the thunder of someone actually doing something.
The whole "open source" process fails for large projects. All the significant projects are run by a small group. While Linux has many "contributers" the core team is small, and Linux's development is slowing because of the management issues.
The best free software (Emacs, Apache, Perl, Samba, GCC, etc.) are ALL small teams or individuals writing great code. Even if they get contributions, the direction is set by a small group. The "Open Source" projects that ESR talks about are among the weaker links (Mozilla is the prime example, but I'd argue that Linux falls in here... all the stability arguements are in contrast to Windows, which by engineering tradeoff has more features and less stability, not against other Unixes. Does anyone here REALLY feel that Linux stacks up against HP-UX, Solaris, AIX, or BSD 4.4 Lites in stability? I mean the kernel, not the mess of software).
The Mythical Man Month holds...
Open Source is the trojan horse to destroy Free Software. By focusing on the convenient instead of the ideals, once the Millions of Eyeballs theory falls to the Mythical Man Month, then Open Source becomes passe.
Show RMS some respect... y'all run his code, use his license, and piss on his work because it isn't convenient to placate your corporations?
What the hell?
As a result of OS X, I'm switching to a Mac. I haven't owned an Apple since an Apple //c, but I can't wait to be able to afford my new G4 Powerbook.
OS X seems incredible, and since I've switched from Linux -> BSD (Open), I've been a happy camper. OS X seems like it will let me play nice with all my applications, including the important option of coding for deployment on my BSD boxes.
I don't have time to hack Darwin, but I'm glad they openned it. Assuming they got a single modification submitted that was worthwhile, openning it was good, because it means when OS X runs on my shiny new laptop, it'll be better.
Slashdot ignorance is aggravating, but you take the good with the bad.
This is slight more newsworthy than the usual "Linux is the best OS ever, I mean, it's more stable than Windows 98!" mental masturbation that sits on the comment sections these days.
Alex
I don't get how he ever became relevant. He is definitely involved in some coding, but relatively trivial. He was the first person to decide that the secret to Free Software was for businesses to get involved. Additionally, he created the concept (in words) of Open Source development which has YET to succeed.
He talked Netscape into the Mozilla project. While I actually somewhat enjoy Mozilla, I find no reason to switch from IE at this time, and Mozilla shipped as Netscape 6 two years after starting, when his theory was that it would reduce development time.
There are two theories, the mythical man month and the Bazarre theory. The Mythical Man Month was written after the management of a HUGE software project. The Bazarre theory was written after writing an e-mail downloader and a whim... As someone who has managed small scale projects, the mythical man month is totally true.
I never understood how he became a spokesman. I don't agree with ANY of RMS's political theories, but I like his theories on software. I think that his contributions (GNU, the GPL, Emacs) are HUGE, while ESR wrote some essays.
While both the tech writer and engineer are important to a project, the engineer is more useful. You can run a project without a tech writer (you'll suffer later, but it can be done), but try an engineering project without an engineer.
His code contributions while nice, are trivial compared to the air time/free money he has received.
Alex
His code are extensions on GPL'd code. Yeah. The GPL is not magic. It does not declare code free. Code is copyrighted by the author, the GPL is a possible license.
Note: IANAL, but I have read the GPL and read some case law on copyrights and licenses.
He owns his extensions. The company is suggesting that they become their property (if he works on them while there). If he agrees, then they own his extensions. They can release them under the GPL or not. He has NO say.
However, they cannot build a derivative work (the GPL code + his extensions) and distribute it without meeting the requirements of the GPL. However, they can do WHATEVER they want with his extensions, including keeping them proprietary. The combination of GPL code + his extensions is a derivative work to the GPL code. His extensions are NOT a derivative work, they are an original work.
There is no magic to the GPL that says you must do something. If I take MS code, slap a GPL on it, and release it, it isn't GPL because the license has no meaning, I don't own the work.
In this scenario, the code becomes Company X code (again, assuming he improves it while there), who can release it or not under any license.
But he already released it under the GPL? Irrelevant. Those that have already obtained the code under the GPL from him are able to use it (and redistribute), but he could not. The GPL isn't magic, it is a way of licensing software.
You CAN reclose GPL code, but the code released under the GPL already is floating around under the GPL. However, if he signs it over to the company, who prohibits him from releasing it under the GPL, he can't release it, he doesn't own it.
Now it would be interesting for him to further relicense that code from a third party (under the GPL), but I'm not convinced that that would work given that he signed away all rights to it. However, if people haven't maintained copies with the licenses, he is SOL, because the company's version is no longer GPL'd.
Alex
Guys, the law is NOT code. You don't write an algorithm. There is judgement and grey.
For example, if you are providing a channel, or hosting a site entitled Kittie Porn, you should be responsible. If you just pull everything down, you shouldn't. It's a judgement thing.
For example, Napster enables MP3 sharing. In and of itself, this is innocent. However, Napster will scan your hard drive for you and offer to share out all your MP3s. The problem with this is that while it is legal for me to rip a CD that I own (or I'll argue, a friend owns) for my own purposes, I can't legally share it out. Napster encourages you to do so. Intent matters.
For example, if google happens to index a kiddie porn site, they didn't intentionally do it and shouldn't be responsible.
If I provide a site that is a directory of kiddie porn sites, I SHOULD be responsible, even if I'm just linking. If I provide a link to a site that is legal porn, and they add kiddie porn, I shouldn't be.
Basically, if you are intentionally doing something shady, you are responsible. If something happens beyond your control, safe harbor should protect you.
I went to a prep school for my last three years of high school, where I continued to slack in my classes. I did well in extra-curriculars, but I was a mediocre student. However, I rocked my SATs, SAT IIs, and AP tests. I used this to show schools that I was a capable individual and got an opportunity to attend an elite school.
Does this jive with America's sense of the Protestant Work Ethic and working hard to achieve? Absolutely not. Did I do reasonably well in College? Yes, although a little more effort and I'd have done well. However, my grades alone wouldn't have gotten me in.
I'm a middle-class white student, I'm NOT who he wants to help, I'm who he wants to hurt.
Here is the problem though. I went to a prep school with a decent reputation among many schools throughout the nation. As a result, the entire top 20% of my school attended elite schools. My school's reputation got relatively dumb kids with high GPAs at a respectable school and mediocre test scores into good schools.
Those same kids in a local Public school no doubt would have had high GPAs and low test scores, but not been considered, because the low test scores would show that they weren't that swift.
All in all, this sounds good, right?
No. Without test scores, the only means of evaluating GPA is to look at the schools reputation. While this is fine for Private Schools or elite Public schools (magnets focusing on education in the South, special enterance exams in northeastern metropolitan cities, for example), but what about the poor kids from unknown schools?
Yeah, they'll get a 4.0 unweigthed GPA and rank in the top five or whatever at their school, but do you accept them? At a school you've never heard of in a poor area, you can assume that a 4.0 isn't of reasonable caliber for your school. What do you use, AP classes? If you're from a middle class suburb or attend a private school, AP classes may be the answer. If you're from a poor school area, you don't have that option.
Yes, money helps. You can't change that. What you CAN do is give people without money an opportunity. The kids with the expensive education and test prep have an advantage on the tests, but it isn't a guarantee.
In a scenario without the tests, the kids HAVE a guarantee. They are the only ones from high schools that schools have heard of. Without the tests, your poor kid with the 4.0 but no APs because they aren't offered and no SAT scores to back him up doesn't get in, but the kids from the private schools with 3.6 GPAs with a reputation for excellence do.
That isn't useful.
Admissions already adjusts for background. The UC system is absurd, allowing admission by test scores alone, which reduces their paperwork, but doesn't help. Their admissions process is too numeric, but they are processing too many applications for hands on evaluations. Schools normally have some kind of cut where they reduce their applicant pool by screening out a certain GPA range. Then admissions officers can make decisions. The poor kid with a 4.0 from a crappy school with a 1300 GPA is probably the equivalent of the wealthy kid with a 1400 GPA based upon the extra training and education. They CAN take that into account. They can't take race into account, but they can figure out the student's background. A poor white kid OR poor black kid should have that taken into account. A rich white kid OR rich black kid should have their background into account.
No, test scores don't substitute for discretion. But they allow you to justify things. If the kid with a 4.0 GPA had a 1000 SAT, he probably doesn't get in. But with a 1300, he does. Why? In the first case, you can't confirm his aptitude, in the second you can. Without the tests, you CAN'T confirm his aptitude AT ALL. You'd assume he is a 1000 SAT kid because he's from a crappy school and admit the known candidate from the prep school.
This man HAS NOT thought through his situation. They need to add more discretion, but they don't need to scrap the best means for helping people out.
This isn't exactly an Open Source business, but a business catering to the "Open Source" community that is in trouble. Much like Tucows BSD section, the rest of the computing world really can't get in.
When I was an NT Administrator, I got Windows NT Magazine, and so did everybody I know. They had GREAT articles and tips that would help any NT Admin out in running a site.
However, the Linux group doesn't play by the same rules. I occaisionally buy a book (O'Reilly's SAMBA book comes to mind), but that's about it. The rest of the time I use online docs. For my BSD boxes, I find that the man pages are the most useful things. I never think to get a Linux magazine.
I think that it's a cultural thing. In the Linux world, I don't feel "mainstream" and don't feel like buying mainstream publications. I don't think to spend money.
For programmers, Open Source can work. I know that I've hired Open Source program writers to install applications, etc. (I mean, for corporate use, when my guys are worth $X/hr, and it will take Y hrs to learn and do it, if X*Y > Consultant fee... I hird consultants), but for people catering to hobbyiests, it may be the end.
The demographic that used to support all the PC hobbyist books/magazines are pretty much all using Linux. It was a group of tweakers, and you can't tweak Windows anymore (well, some, but it's not useful). That market is dead, replaced by websites and newsgroups.
I just wonder, why do Windows magazines/books sell and Linux ones don't? Is it just being abnormal?
Alex
I don't see anything in there giving authorization of the Trademark...
If you're not compatible, you can't use the name.
That doesn't mean you can use the registered mark. When that was done, there was no registered mark. The Mark is owned by a corporation (founded by this gentlemen, not that it's terribly relevant), and can't be appropriated without authorization.
For example, if I write some code, call my program Microsoft, and say in my license that you may use the name Microsoft in referring to my code, you aren't authorized to use it.
That is hardly a license to a Trademark issued two years after that was written.
The author of the original software asked you not to use the name ssh or Secure Shell if they are incompatible. That is a contractual agreement as part of a license.
The company is using the Trademark SSH to refer to their company and software. When you used that license, you contracted to not name is SSH if it is not compatible. That has NO bearing on the trademark.
The statement that the mark wasn't enforced was the WRONG tack to take.
Good job. We have now taken the position to the outside world of being total assholes.
The guy made an effort NOT to bully an Open Source group. He didn't send threatening lawyer letters. He asked people to be reasonable. Quite frankly, you're going to lose BADLY in a court of law, because not only is there the possibility of confusion, but there is DOCUMENTED confusion.
Now, this is totally absurd.
He didn't bully, DON'T use that against him. The arguement that he didn't bring it up until confusion happened may be legally correct, but now sets the horrible precedent that Open Source groups will be as shady as possible, and if you give them an inch, they'll take a mile.
Notice something, the original license (if anything tells people not to be nice and release their source code, this is the example, that was hardly a license, more of a nice gesture) allowed you to call your application ssh.
That's irrelevant.
OpenSSH could be renamed, OpenSecSH, which would be an open implementation of the SecSH protocol, which is the name of the working group. The FILE could still be called ssh/sshd.
SAMBA couldn't call itself SMB, confusion reasons, but the applications use SMB (like, smbd).
However, you can be polite, compile as osecsh and osecshd and include a symlink (automatically, but prompted) for ssh and sshd, so if you have both implementations, you can decide which one you want.
However, if the Open Source community insists on fighting on the Trademark grounds, we're in the wrong.
You can dispute the merits of Software Patents.
You can dispute the merits of long copyright terms.
You can dispute the merits of copyrights in general.
You can't dispute the merits of Trademarks.
Trademarks are the only thing that prevents confusion in the marketplace. If people are confused and think that OpenSSH is from SSH, then there is a legitamate issue.
Also, I don't think you're going to win on the enforcement. One year or so is reasonable, given that OpenSSH has minimal press coverage, etc. I think that it would be EASY for SSH to show that they found out about it, saw a problem, and then asked them to fix it.
Theo, you're going to lose, and you're being a bastard. He hasn't demanded that you stop making a version of SSH, just that you not use his product name.
Theo, I'm an OpenBSD user. I love OpenBSD. I love OpenSSH. You're in the wrong here... VERY WRONG.
Alex
Kinda clever, however, isn't there case law that indicates that "copying to memory" is fair use? I believe software companies tried to use the copying to memory to get copyright law to cover the use of the software and the courts told them to blow off.
/.
I also think that "copying to hard drive" falls under fair use, but less relevant to the satire.
I do think that the cracks about the Government ID card are extremely funny and amusing. This is overall a good satire, thanks
Most of country is a little less anti-business then California. Besides, California's laws would probably fail on Federal Appeal, as a restriction of Free Speech. The company has as much free speech as the individual.
Federal law doesn't cover this stuff, only prohibits certain kinds of discrimination. Most states allow corporations reasonable leeway.
Sorry dude, Linux is the ugliest designed Kernel ever. It ignores the last thirty years of software engineering, and puts forth a technologically primitive system. I would argue that the NT Kernel is more advanced than it.
Microsoft broke the NT Kernel's elegance in 4.0 and 5.0, but it was a reasonably well designed kernel.
I mean, the reason I run Linux is that it is "fun" and I can get precompiled things like KDE right out of the box (and I don't get funky things like Cronolog not compiling on my machine like I'm picking up with OpenBSD).
If I don't run Redhat, I lose these advantages.
But, I pick up a poorly designed kernel.
Thanks, but no thanks. I think that Linux is the best hacked together kernel, but I find the design really inelegant.
Besides, I have no time on my hands, and this isn't on the top of my todo list. However, I'll check out the site and keep it in mind in the future, thanks.
Alex
After getting hacked, we switched our servers from Redhat to OpenBSD. Part of this was the security audit/out of box security, and part was the belief that a more obscure system is safer. We got his with a rootkit, and avoiding Redhat/NT means we are out of the mainstream and less likely to get slammed.
I would never go back. We still have Linux boxes for Development because the Redhat KDE Workstation install is a dream, but for a server, wow, I love BSD.
The BSD methodology is more sane. The startup process is cleaner and easier to manage. My system is more organized. The Ports are terrific.
For a Workstation where I want the newest toys, I might use Linux because it always gets the whiz-bang features, although I may toss FreeBSD on a box to play with.
However, with OpenBSD I know what's up, with Linux, I find that millions of things install everywhere. Besides, my OpenBSD FTP install now takes about 10 minutes after I learned the install process, the Redhat one still takes forever with it crashing about half the time.
I realize that Redhat != Linux, but it is the most common distribution and aims for the corporate market. I feel like the Redhat system is held together with chewing gum with rediculous numbers of scripts that are hacks to hold different systems together. I should try Debian, but OpenBSD has done what I want and I look foward to trying FreeBSD. I've done NetBSD and OpenBSD systems, and they have been less aggravation than Linux installs.
I love BSD now, and I won't go back.
Alex
Um, bzzzt, wrong. SCO isn't pretty, or sexy, but they sell a LOT of licenses. They are a UNIX on Intel, and have a large business presence.
There servers have never been the most powerful (running on Intel Architecture, which until 2-3 years ago was seriously sucking compared to real Iron), but they have sold a lot of licenses.
The big Unix shops used to make a lot of money from licenses, but they were fewer licenses more $/license.
Not getting hype doesn't make you insignificant.
Hell, Solaris had limited presence in the Server room until the Internet boom, because their Hardware was considered substandard to IBM and HP stuff. However, they were a workstation player that gave their machines to engineering schools on the cheap, so people equated Sun with Unix.
Marketing gives the perception of marketshare, not the reality.
I'm sure this will open up a feeding frenzy to post this on Slashdot, but some of NT Admins can't do that. While some consolidate on 1 or 2 boxes, a well designed NT Network will host at least 3 servers, and they may only have a few dozen users. For example, you should have a PDC, BDC, and Exchange Server. If you combine the functions (which I've done at small sites), you open yourself up to headaches later on.
Besides, my PDC/BDCs get hammered as people log in, but there is little login during the day. Exchange doesn't get hit much at night, as only the coders are logged in and less e-mail is sent. This is a silly arguement.
This list is to keep bug information to people that program with Bind. For Example, if you're a Redhat user, you will know that Redhat will buy into this, and therefore the Redhat patch will be out before the exploit becomes common knowledge and there is a downloadable kit that trashes a machine on it.
Combine this with Redhat's automatic update services, and you have used this to really help the business model.
If you are programming on Bind, you will be on this list (or should be on this list). If you just install Bind and don't touch the code, getting exploit information is of marginal benefit. However, giving info on the exploit makes it easy for someone who doesn't know much to code up an auto-exploiter.
This allows vendors to bust their ass on a fix before the vulnerability gets out.
There is an exception here. Right now, when there is a problem, people are in a frenzy to fix it. If this allows them to not feel as pressured to fix the bug, then this is very counter productive.
Alex