Sorry, we can't blame HP-UX either. The Sabre system was converted to Tandem Non-Stop hardware running Unix. This is only and HP issue because Compaq bought Tandem before HP bought Compaq. Doncha love the computer business?
This is undoubtedly a problem with Sabre, which EDS runs on behalf of Sabre Holdings. Both American Airlines and US Airways use Sabre for much of their operations.
Sabre started it's life as an American Airlines internal system (SABER, slight spelling difference), running on a rare operating system (PARS, later called ACP and currently TPF) on IBM mainframes. In the last few years Sabre completed a lengthy migration to HP Unix on Non-Stop (i.e. ex-Tandem) hardware. The mainframe systems were rock solid, but software talent was hard to come by, so they decided the time had come to switch.
I think this is in fact good for the project. This goes to show that the patents issue (worldwide, not only in Europe) is becoming a growing concern for more and more sectors. Seeing that they are being careful about this actually makes me think they remain very serious about seeing this project get finished well.
That's an excellant point. This isn't necessarily a bad thing for Open Source. In fact, as things stand, it sounds like it's in Munich's best interest to press for an anti-patent answer from the EU. And as the parent notes, the two named individuals are pro-source.
Closed source is better for patent-threatened users because there are contracts in place and because of "fitness for use" laws. I can't speak about Germany, but in the USA if you sell me something and it violates somebody else's patent, you need to make things right for me. And if "you" are some large company, making it right can involve patent cross-licensing and no cash changes hands. Outfits like Microsoft, Sun, HP, and IBM do that all the time.
Open Source is great, but as the licenses make clear, *you* wind up holding all the liabilities. There aren't any warranties, and there's no implied fitness for use. If Open Source software violates somebody's patent, it may be possible for them to sue you for infringement. They can certainly sue you to require you to "destroy" your copies of that software.
All fine and good, but what exactly constitutes unauthorized exhibition of a motion picture or video tape?
According to 17 USC 101 and 106:
To perform or display a work ''publicly'' means -
(1)
to perform or display it at a place open to the public or at any place where a substantial number of persons outside of a normal circle of a family and its social acquaintances is gathered; or
...
Subject to sections 107 through 121, the owner of copyright under this title has the exclusive rights to do and to authorize any of the following:
...
(4)
in the case of literary, musical, dramatic, and choreographic works, pantomimes, and motion pictures and other audiovisual works, to perform the copyrighted work publicly;
So, yes, churches, summer camps etc. movies are illegal. When I was in high school, we showed films (16mm, multi-reel) every few weeks after school. The rental included a performance fee that legalized our doing so. Blockbuster doesn't pay that fee for you, and neither does NetFlix.
"Just because we can do a thing does not mean that we must do it." The technology does not imply the right.
We should never forget that the first software company that the FSF and Stallman openly attacked was Apple. Long ago, and under Jobs' watch as once again, Apple sued all and sundry over the "look and feel" of their systems. Of course they're going to threaten anyone who breaks their stranglehold on the iPod. It's what Apple does - they develop monopolies and try to maintain hard them, often at the expense of their own "partners". Anybody seen a "clone Mac" lately?
Somewhere I still have an LPF button with the seven-color apple logo and a snake coming out of it, with the caption "Keep your lawyers off my computer!".
It *is* your responsibility to provide the source, but you do not have to include it under section 3A. You can provide it under section 3B or 3C if you like. Providing the source "when you provide the binary" is an option, not a requirement, of the GPL.
Correct, mostly. You're right about having the choice of providing either the source or an offer of the source. But unless you were given a section 3B offer, you don't have one to pass along under section 3C, so your 3B option would be to become a source distribution point yourself. So I guess we're both right.
Redhat is also bound by the GPL, so they have to provide source code.
Sure. To the people they provide binaries to. And last time I checked, the official Red Hat FTP sites had both the i386 and source CD-ROM images and RPMs, so they're in compliance.
If Redhat goes out of business you're right, you'd be obligated to provide source code (or find someone else that does).
Nope, as long as Red Hat complied with 3A as above, it's your job to comply with it yourself by handing out source CD-ROMs. The 3C exemption doesn't apply to you because you didn't get the 3B offer.
I don't mean to be pedantic about this, but people often misunderstand the GPL and their responsibilities under it.
And the question there is whether this "unreleased" source is GPL'd. One would expect that anything based on a GPL'd codebase would be GPL'd if made available officially or unofficially.
That's correct - anytime you distribute a GPL'ed program, you have to make the source available, even for "unreleased" versions. The FSF's GPL FAQ says so.
In this case, pre-releases are available, and the distribution of pre-release source should be protected by the GPL. If we consider released pre-releases to be releases, then, Sveasoft's actions against those distributing pre-release source are illegal - but it's a complex 'if'.
If Sveasoft gives out the binaries, they have to give out the source - it's just that simple. Now, I have to admit that I've always felt the FSF was off-base on this issue. Even in the world of Free Software and Open Source, there is reasonable justification for "controlled release" of code that is still under test. But the GPL doesn't have a "restricted beta" concept, so that's just the way the cookie crumbles.
No, you only have to offer source code if you've modified the code.
Nope, you have to provide the source if you received it, or the information about the offer you received. This has nothing to do with modifications. See GPL section 3 for the details.
If I download a redhat install disk, then give it to my friend I'm not violating the GPL by not providing him with source code.
Red Hat provided you with the source, so you have to provide it to anyone you provide binaries to (per GPL sec. 3A). Of course, Red Hat may also have provided you with a written offer for the source (per sec. 3B). If they did, then you're entitled by sec. 3C to pass that offer along, but if they didn't then you'd better have burned your friend the SRPM CDs as well.
Ummm no. If I receive a binary to a GPL'd program I can distribute it without source simply by telling the person I distribute it to where I got the source from. The person I got the source from is responsible for supplying the source to the third party, not me.
Only if you received the binary without the source and with the "offer valid... to... any third party". If you received (or even had simultaneous access to) the source, it's your responsibility to provide it when you provide the binary. Check out GPL section 3C. And if you didn't receive either the source or the offer, the binary was distributed in violation of the GPL.
Using already-defined search engines? Sure. But defining your own? I don't see any way to do it, and the Opera help system says there isn't (ref.http://www.opera.com/support/search/supsearch.dml?index=351). I run 7.50, which isn't the newest but it's darn close.
It's nice to see this appear in Free/Open Software, but just like tabbed browsing, Opera had the search thing licked first. And it's a lot easier to choose an alternative search engine (although it isn't infinitely configurable like Firefox). And I love the "right click and choose 'Search' or 'Search with'" model that Opera 7 has.
... these are my standard recommendations, and the books that have survived up to 25 years of periodic home- and office-library purges. They all have the benefit of not being the "soup du jour", and are still relevant many years after I first read them.
"The Soul of a New Machine", Tracy Kidder. A must-have for anyone interested in managing or leading geeks, and a good read to boot.
"The Art of Computer Programming", Donald Knuth. Especially volume 1, "Fundamental Algoritms".
"Principles of Compiler Design", Alfred Aho & Jeffrey Ullman (aka "the dragon book"). Kind of specialized, but appropriate for Comp Sci students.
"Computer Networks", Andrew Tanenbaum. Dated but still accurate and useful in understanding modern networks, especially radio-LANs (WiFi et al.).
"Operating Systems Design and Implementation", Andrew Tannenbaum. The book that got Linus Torvalds started on writing the Linux kernel. I disagree with Prof. Tannenbaum on the supremacy of microkernels, but it's still important.
"Software Reliability", Glenford Myers. We should all care about this stuff, and Myers attacks the topic in a reachable fashion.
"Metamagical Themas: Questing for the Essence of Mind and Pattern", and "Godel, Escher Bach: The Eternal Golden Braid" (aka "GEB"), both by Douglas Hofstadter. They're both tough to read but useful, although I only recommend them to people I know very well, and even then with a caveat. But if you like to *think about* computing, they're inspirational.
"The Visual Display of Quantitative Information", Edward Tutfe. Forget GUIs and other cutesy-pie "graphical" eye candy - Tufte teaches the real deal.
So, it sucks how they try to shove the PSP's up your ass like that, but they *are* pretty good service plans.
Yup, for some products these plans (Best Buy's and others) can be VERY good. For example, the Radio Shack service plans for cell phones include battery replacements. If you've every had to replace one of them, you know that they cost at least half as much as the phone's list price. But for 10-15% of list price you can buy a three-year service plan that offers you a brand new battery every year. Given the lifetime of some batteries, this is a major win.
"Do not mention the fucking PSP again, this is abusive, I have stated clearly 3 times that I am uninterested and have also kindly asked you to simply just stop talking to me about it at all. You don't seem to get it.
Here can you understand this "I don't want the fucking PSP"
Rule #1: Never drop the f-bomb. It gets you noplace and changes the basis of the conversation. Forty years ago, it was shocking and people would react by surrendering. Today it just makes them angry and you immediately lose any opportunity to win the argument.
Stock options don't have a clear value. Since you can't say "12,000,000 options are outstanding and excercisable, at a cost to the company of US$120M", you can't apply it as an expense. If you think Enron and Worldcom cooked their books, just wait until you see how the "expense" of stock options winds up being calculated. It's just as bad as requiring businesses to value their "goodwill" and take an earnings hit when it "goes down".
Options are issued from an "options pool". In any company large enough to be subject to FASB rules, that pool has already been set aside for that purpose. The dilution happened to the early investors (angels, pre-VC folks, etc.), which it was a small private company.
When it comes to copyrights and licensing, the law is on your side to begin with. If you deliver 1000 units with your software, and no license, then they can use the software in the usual way, but they cannot redistribute, copy, etc., etc.
If these guys have only one customer and no contract and have already turned over samples, they should watch out. The code might be deemed a "work for hire", especially if they took cash to produce the samples. If so, the copyright rests with the customer, not the programmers.
Licenses and contracts aren't BS, they're a very basic part of business. If you don't like having written rules and paying lawyers to draft them, you shouldn't be trying to conduct business yourself. Just be someone else's wage slave and enjoy the freedom of not worrying how you're going to make payroll next week and whether you'll have to fire your friends because you screwed up.
In either case, just ask them honestly what there budget for the project is. Do this/after/ explaining your expenses; that is, how much you want and why you want it, broken up (honestly) into sunk costs, future development, and the stuff for your pocket.
No no no, don't do that! Your costs have nothing to do with the price a customer will be willing to pay. One of the basic tenets of capitalism is that goods trade at or near their perceived value. If that value exceeds your cost, you make a profit (congratulations!), if not you take a loss. How much profit you make is direct compensation for the value you added to the parts you obtained before selling the result to your customer. This is the true meaning of "value added" (not to be confused with the "value added reseller", a synonym for "retailer").
Once they quote a budget, if it's reasonable, take 80% of it. If it's not reasonable, say so. That'll make them want to come back.
Now that's good advice, or at least almost. There are two schools of though on deals: "Everyone should feel some pain" and "Everyone should feel happy". The former is based on non-recurring sales, where you can squeeze all the profit you want and then scorch the ground. The latter assumes that your repuration with this customer is valuable to you, and theirs with you likewise. If you believe in recurring sales (usually a good thing!), you should always leave some money on the table, although I'd say 20% is too much.
I'm reminded of a SciFi Channel ad: "Even numbered Trek movies don't suck". Enough already - the "Halloween Surprise" was good info. This is just getting tired.
Is it only me, or has Raymond played the "I wrote fetchmail" card for a whole lot more than it was ever worth? The last time he tried any hackery, the Linux kernel developers sent him and his Config Markup Language 2 packing. And for good reason: developed in a vacuum, presented as a fait accompli, and defended bitterly against all criticism, it was everything he argued against in "The Cathedral and the Bazaar"
Kidder's "The Soul of a New Machine" should be required reading for anyone considering managing technical people. The lessons Kidder noted from the Data General team he observed are timeless.
And yes, Brooks' "The Mythical Man Month" is still valid, because it isn't about code, it's about software project management. Like it or not, nothing has really changed in the field in the last 30 years. Yes, the languages have changed (although APL programs and C programs typically have the same number of comments, excluding the lawyerese). Yes, we type in front of LCD screens now (although code windows still default to 80 "columns"). But programmers are still writing programs and those programs are still interacting with each other and still suffering from complex-system effects. And the "second system effect"? Just ask anybody who's had to pay for a rewrite of their pre-existing system in Java, C++ or whatever else the paradigm du jour is.
Slashdot Mirror
Sorry, we can't blame HP-UX either. The Sabre system was converted to Tandem Non-Stop hardware running Unix. This is only and HP issue because Compaq bought Tandem before HP bought Compaq. Doncha love the computer business?
This is undoubtedly a problem with Sabre, which EDS runs on behalf of Sabre Holdings. Both American Airlines and US Airways use Sabre for much of their operations.
Sabre started it's life as an American Airlines internal system (SABER, slight spelling difference), running on a rare operating system (PARS, later called ACP and currently TPF) on IBM mainframes. In the last few years Sabre completed a lengthy migration to HP Unix on Non-Stop (i.e. ex-Tandem) hardware. The mainframe systems were rock solid, but software talent was hard to come by, so they decided the time had come to switch.
Sorry, no Microsoft to blame here!
I think this is in fact good for the project. This goes to show that the patents issue (worldwide, not only in Europe) is becoming a growing concern for more and more sectors. Seeing that they are being careful about this actually makes me think they remain very serious about seeing this project get finished well.
That's an excellant point. This isn't necessarily a bad thing for Open Source. In fact, as things stand, it sounds like it's in Munich's best interest to press for an anti-patent answer from the EU. And as the parent notes, the two named individuals are pro-source.
Closed source is better for patent-threatened users because there are contracts in place and because of "fitness for use" laws. I can't speak about Germany, but in the USA if you sell me something and it violates somebody else's patent, you need to make things right for me. And if "you" are some large company, making it right can involve patent cross-licensing and no cash changes hands. Outfits like Microsoft, Sun, HP, and IBM do that all the time.
Open Source is great, but as the licenses make clear, *you* wind up holding all the liabilities. There aren't any warranties, and there's no implied fitness for use. If Open Source software violates somebody's patent, it may be possible for them to sue you for infringement. They can certainly sue you to require you to "destroy" your copies of that software.
Hence all the concern about software patents.
All fine and good, but what exactly constitutes unauthorized exhibition of a motion picture or video tape?
According to 17 USC 101 and 106:
So, yes, churches, summer camps etc. movies are illegal. When I was in high school, we showed films (16mm, multi-reel) every few weeks after school. The rental included a performance fee that legalized our doing so. Blockbuster doesn't pay that fee for you, and neither does NetFlix."Just because we can do a thing does not mean that we must do it." The technology does not imply the right.
We should never forget that the first software company that the FSF and Stallman openly attacked was Apple. Long ago, and under Jobs' watch as once again, Apple sued all and sundry over the "look and feel" of their systems. Of course they're going to threaten anyone who breaks their stranglehold on the iPod. It's what Apple does - they develop monopolies and try to maintain hard them, often at the expense of their own "partners". Anybody seen a "clone Mac" lately?
Somewhere I still have an LPF button with the seven-color apple logo and a snake coming out of it, with the caption "Keep your lawyers off my computer!".
The earliest versions of make replied "not war?".
:-)
I guess you had to be there
It *is* your responsibility to provide the source, but you do not have to include it under section 3A. You can provide it under section 3B or 3C if you like. Providing the source "when you provide the binary" is an option, not a requirement, of the GPL.
Correct, mostly. You're right about having the choice of providing either the source or an offer of the source. But unless you were given a section 3B offer, you don't have one to pass along under section 3C, so your 3B option would be to become a source distribution point yourself. So I guess we're both right.
Redhat is also bound by the GPL, so they have to provide source code.
Sure. To the people they provide binaries to. And last time I checked, the official Red Hat FTP sites had both the i386 and source CD-ROM images and RPMs, so they're in compliance.
If Redhat goes out of business you're right, you'd be obligated to provide source code (or find someone else that does).
Nope, as long as Red Hat complied with 3A as above, it's your job to comply with it yourself by handing out source CD-ROMs. The 3C exemption doesn't apply to you because you didn't get the 3B offer.
I don't mean to be pedantic about this, but people often misunderstand the GPL and their responsibilities under it.
And the question there is whether this "unreleased" source is GPL'd. One would expect that anything based on a GPL'd codebase would be GPL'd if made available officially or unofficially.
That's correct - anytime you distribute a GPL'ed program, you have to make the source available, even for "unreleased" versions. The FSF's GPL FAQ says so.
In this case, pre-releases are available, and the distribution of pre-release source should be protected by the GPL. If we consider released pre-releases to be releases, then, Sveasoft's actions against those distributing pre-release source are illegal - but it's a complex 'if'.
If Sveasoft gives out the binaries, they have to give out the source - it's just that simple. Now, I have to admit that I've always felt the FSF was off-base on this issue. Even in the world of Free Software and Open Source, there is reasonable justification for "controlled release" of code that is still under test. But the GPL doesn't have a "restricted beta" concept, so that's just the way the cookie crumbles.
No, you only have to offer source code if you've modified the code.
Nope, you have to provide the source if you received it, or the information about the offer you received. This has nothing to do with modifications. See GPL section 3 for the details.
If I download a redhat install disk, then give it to my friend I'm not violating the GPL by not providing him with source code.
Red Hat provided you with the source, so you have to provide it to anyone you provide binaries to (per GPL sec. 3A). Of course, Red Hat may also have provided you with a written offer for the source (per sec. 3B). If they did, then you're entitled by sec. 3C to pass that offer along, but if they didn't then you'd better have burned your friend the SRPM CDs as well.
Ummm no. If I receive a binary to a GPL'd program I can distribute it without source simply by telling the person I distribute it to where I got the source from. The person I got the source from is responsible for supplying the source to the third party, not me.
Only if you received the binary without the source and with the "offer valid ... to ... any third party". If you received (or even had simultaneous access to) the source, it's your responsibility to provide it when you provide the binary. Check out GPL section 3C. And if you didn't receive either the source or the offer, the binary was distributed in violation of the GPL.
Using already-defined search engines? Sure. But defining your own? I don't see any way to do it, and the Opera help system says there isn't (ref.http://www.opera.com/support/search/supsearch .dml?index=351). I run 7.50, which isn't the newest but it's darn close.
Prove me wrong - I'd love to know how!
It's nice to see this appear in Free/Open Software, but just like tabbed browsing, Opera had the search thing licked first. And it's a lot easier to choose an alternative search engine (although it isn't infinitely configurable like Firefox). And I love the "right click and choose 'Search' or 'Search with'" model that Opera 7 has.
... these are my standard recommendations, and the books that have survived up to 25 years of periodic home- and office-library purges. They all have the benefit of not being the "soup du jour", and are still relevant many years after I first read them.
"The Soul of a New Machine", Tracy Kidder. A must-have for anyone interested in managing or leading geeks, and a good read to boot.
"The Art of Computer Programming", Donald Knuth. Especially volume 1, "Fundamental Algoritms".
"Principles of Compiler Design", Alfred Aho & Jeffrey Ullman (aka "the dragon book"). Kind of specialized, but appropriate for Comp Sci students.
"Computer Networks", Andrew Tanenbaum. Dated but still accurate and useful in understanding modern networks, especially radio-LANs (WiFi et al.).
"Operating Systems Design and Implementation", Andrew Tannenbaum. The book that got Linus Torvalds started on writing the Linux kernel. I disagree with Prof. Tannenbaum on the supremacy of microkernels, but it's still important.
"Software Reliability", Glenford Myers. We should all care about this stuff, and Myers attacks the topic in a reachable fashion.
"Metamagical Themas: Questing for the Essence of Mind and Pattern", and "Godel, Escher Bach: The Eternal Golden Braid" (aka "GEB"), both by Douglas Hofstadter. They're both tough to read but useful, although I only recommend them to people I know very well, and even then with a caveat. But if you like to *think about* computing, they're inspirational.
"The Visual Display of Quantitative Information", Edward Tutfe. Forget GUIs and other cutesy-pie "graphical" eye candy - Tufte teaches the real deal.
So, it sucks how they try to shove the PSP's up your ass like that, but they *are* pretty good service plans.
Yup, for some products these plans (Best Buy's and others) can be VERY good. For example, the Radio Shack service plans for cell phones include battery replacements. If you've every had to replace one of them, you know that they cost at least half as much as the phone's list price. But for 10-15% of list price you can buy a three-year service plan that offers you a brand new battery every year. Given the lifetime of some batteries, this is a major win.
"Do not mention the fucking PSP again, this is abusive, I have stated clearly 3 times that I am uninterested and have also kindly asked you to simply just stop talking to me about it at all. You don't seem to get it. Here can you understand this "I don't want the fucking PSP"
Rule #1: Never drop the f-bomb. It gets you noplace and changes the basis of the conversation. Forty years ago, it was shocking and people would react by surrendering. Today it just makes them angry and you immediately lose any opportunity to win the argument.
On the CLI, this is solved with a simple "apropos integrated development."
Nly n Unx wld ll th vwls b sqzd t nd thn wld ppl b skd t spll "apropos" crrctly. Tht's Pgrro's pnt xctly.
Genhis Khan would be the Mandarin version of Grendel.
Stock options don't have a clear value. Since you can't say "12,000,000 options are outstanding and excercisable, at a cost to the company of US$120M", you can't apply it as an expense. If you think Enron and Worldcom cooked their books, just wait until you see how the "expense" of stock options winds up being calculated. It's just as bad as requiring businesses to value their "goodwill" and take an earnings hit when it "goes down".
Options are issued from an "options pool". In any company large enough to be subject to FASB rules, that pool has already been set aside for that purpose. The dilution happened to the early investors (angels, pre-VC folks, etc.), which it was a small private company.
When it comes to copyrights and licensing, the law is on your side to begin with. If you deliver 1000 units with your software, and no license, then they can use the software in the usual way, but they cannot redistribute, copy, etc., etc.
If these guys have only one customer and no contract and have already turned over samples, they should watch out. The code might be deemed a "work for hire", especially if they took cash to produce the samples. If so, the copyright rests with the customer, not the programmers.
Licenses and contracts aren't BS, they're a very basic part of business. If you don't like having written rules and paying lawyers to draft them, you shouldn't be trying to conduct business yourself. Just be someone else's wage slave and enjoy the freedom of not worrying how you're going to make payroll next week and whether you'll have to fire your friends because you screwed up.
In either case, just ask them honestly what there budget for the project is. Do this /after/ explaining your expenses; that is, how much you want and why you want it, broken up (honestly) into sunk costs, future development, and the stuff for your pocket.
No no no, don't do that! Your costs have nothing to do with the price a customer will be willing to pay. One of the basic tenets of capitalism is that goods trade at or near their perceived value. If that value exceeds your cost, you make a profit (congratulations!), if not you take a loss. How much profit you make is direct compensation for the value you added to the parts you obtained before selling the result to your customer. This is the true meaning of "value added" (not to be confused with the "value added reseller", a synonym for "retailer").
Once they quote a budget, if it's reasonable, take 80% of it. If it's not reasonable, say so. That'll make them want to come back.
Now that's good advice, or at least almost. There are two schools of though on deals: "Everyone should feel some pain" and "Everyone should feel happy". The former is based on non-recurring sales, where you can squeeze all the profit you want and then scorch the ground. The latter assumes that your repuration with this customer is valuable to you, and theirs with you likewise. If you believe in recurring sales (usually a good thing!), you should always leave some money on the table, although I'd say 20% is too much.
I'm reminded of a SciFi Channel ad: "Even numbered Trek movies don't suck". Enough already - the "Halloween Surprise" was good info. This is just getting tired.
Is it only me, or has Raymond played the "I wrote fetchmail" card for a whole lot more than it was ever worth? The last time he tried any hackery, the Linux kernel developers sent him and his Config Markup Language 2 packing. And for good reason: developed in a vacuum, presented as a fait accompli, and defended bitterly against all criticism, it was everything he argued against in "The Cathedral and the Bazaar"
Kidder's "The Soul of a New Machine" should be required reading for anyone considering managing technical people. The lessons Kidder noted from the Data General team he observed are timeless.
And yes, Brooks' "The Mythical Man Month" is still valid, because it isn't about code, it's about software project management. Like it or not, nothing has really changed in the field in the last 30 years. Yes, the languages have changed (although APL programs and C programs typically have the same number of comments, excluding the lawyerese). Yes, we type in front of LCD screens now (although code windows still default to 80 "columns"). But programmers are still writing programs and those programs are still interacting with each other and still suffering from complex-system effects. And the "second system effect"? Just ask anybody who's had to pay for a rewrite of their pre-existing system in Java, C++ or whatever else the paradigm du jour is.
Brooks rules, plain and simple.