No worries. The BBC could always "move into the 21st century" by buying programmes outright and freely distributing them on the web. Expect your TV license fee to rise by a couple of orders of magnitude - maybe £5k a year would cover it. Great idea, Einstein!
I've not seen him IRL, but I've read a couple of his articles in NewScientist. Abridged versions of that book, I guess. Read any of his stuff, and you can't help thinking "Von Daniken rides again!" It's the same scattershot, unsubstantiated arguments.
As far as his "singularity" goes, his definition of it is so astoundingly vague that it could apply to anything. IIRC he defines it as an "invention after which human civilisation cannot continue in the same form". OK, so choose one. For travelling, we have the longitude problem, hydrodynamics, steam-ships, aeroplanes, the jet engine, automobiles. For communication we have the public postal system, telegraph, radio, telephone, mobile phones, internet. For data storage we have paper, the printing press, punched cards, microfiche, hard disks, floppy disks, writeable CDs. And don't forget the cheap personal computer. Every *individual* one of these inventions has changed Western civilisation in such a way that continuing in the previous form is impossible if you want to continue existing within the same society. The combined effect of all those inventions is that every few years, civilisation reaches a new "singularity". The only way you can keep the "old" ways is by opting out of the society altogether, viz the Amish and other groups.
Kurzweil's premise is that new inventions are happening more and more frequently, so we're riding an exponential increase. But the problem here is to fix what you call an "invention". Is a DVD-R an invention? Yes, it allows you to manipulate video on your PC in a way you couldn't before, but in another way it's just building on CD-R. And CD-R just builds on tapes and floppy disks. Even the iPod Nano is just a more modern version of the wind-up gramophone. We might think that things are developing at an outrageous pace today, but think of the silly sod who resigned from the Patent Office in the 1890s because "everything had been invented". Kurzweil is just the optimistic side of the same coin - but equally hard-of-thinking.
Go hang-gliding instead! That'll stop you moaning. It might also give you an idea how bloody useless most people are at thinking in 3D. If you can't control a hang-glider (which is the aerial equivalent of a pedal-bike) then you ain't going nowhere *near* a flying car!
But for FOSS projects, we're theoretically dealing with something in which anyone can participate. "Many eyeballs" and all that. This should mean not only that anyone *can* participate, but that anyone should be *able* to. Now if you're going to put some arbitrary restrictions on it - "we won't tell you what you need to know until you've hung out on IRC long enough to join the club" - then you're majorly restricting your eyeball count.:-/
This isn't a commercial benefit, and it isn't a benefit to the project. The only people it benefits are those people in the club who get a nice warm feeling from their exclusivity.
Yes, good spec writing *is* hard, and it's the mark of a good software engineer. It worries me that Linus doesn't appreciate specs, because this suggest to me that he is not a good software engineer. A good coder, yes, but coder != engineer. It suggests that not only does he not realise the benefits of a good spec, but also that he possibly would not be capable of writing one. And that's a pretty sad situation.
Alternatively, maybe you could wish that the fuckwit coders who gave the thing to you had read the documentation and done a little testing to make sure that it works according to the docs, instead of changing things arbitrarily without telling anyone.
All that "my code is my design" bollocks is just that - bollocks. I can spend a week reading code to find how stuff actually works, or I can spend a few hours reading the spec that says how it *should* work. My job is *not* debugging someone else's shit code, my job is writing something that uses or interfaces to their code. If they've not done their job properly, why should I be expected to be the one to find the holes, just because they can't be arsed testing it properly?
but with Linux where there is quite chaotic development... it may not be desirable to have an overall spec
Surely this is exactly where you *do* want some kind of spec. Suppose several people are building separate parts of a car independently. If one person has built the wheel-arches with an 18" diameter, if the person responsible for building the wheels hands you a 20" wheel then you're screwed. It might be the most beautiful wheel in the world, but you'll have to either throw it away or take an angle-grinder to the already-assembled wheel-arches.
In the case of Linux, a spec which tries to specify in detail exactly what each part should do is doomed. But if it at least says what each part's responsibilities are, and what the interface is, then that's all you need. "Does my driver free this memory, or does yours?" "Whose code issues the callback?" Those kind of questions.
I'm actually amused by the hypocrisy of Linus here. Without POSIX, Linux, BSD and the rest would all just be yet another failed machine-specific OS. It's only POSIX compliance that's made them workable, because it's only POSIX compliance that's allowed people to move from regular UNIX to Linux/BSD. Similarly, without X then they'd be without a windowing system, and X is 100% fixed in standards documents. The whole thing is written in C or C++, which again is fixed in standards so that the code will still compile on Intel or other platforms. Sadly this seems symptomatic of many F/OSS leaders. They've taken their college hobby, managed to get paid for it afterwards, and never looked back - and this means they've missed out on any real-life experience. Of those who have had "work", too often this means university teaching, Bell Labs, Xerox Park or somewhere similar where there were no targets, no deadlines and no quality control. In other words somewhere away from *real* software engineering.
Basically, Linus seems a very, very skilled coder/hacker. However as a software engineer, I *really* wouldn't want him around me, working on safety-related software.
Thank you Mr Xerox - if I'd wanted a copy of a FAQ, I would have got one.:-/
Instead we're on the good point that Minna raises of "how do we know this is required from the text of the GPL itself?". As Minna correctly says, the GPL FAQ doesn't count for shit in court, so what does the GPL itself actually say about this? So we're down to interpreting the GPL in real detail, and it turns out to be pretty interesting. Minna's dead right, it doesn't seem to be as clear-cut as the FAQs say. Minna knows how he/she reads it (and has put up good arguments that have tied up other posters). I know how I read it, and although it seems to match the FAQ (and contradict Minna's position), it does indeed seem a pretty gnarly bit of logic to get to the position that the FAQ states. Certainly not as clear-cut as the FAQ, and open to various legal weasellings if you so wished.
If you've got reasoning based on lines of the GPL and US copyright code to support your interpretation, let's hear it. Summaries of FAQs don't cut it.
In the example you gave, a webservice or other based upon another GPLed server is being _published_ on the web
Client-server lesson 101. Code exists on your client computer. Code exists on the server. *Data* passes between them. At no point is either piece of code "published" on the web. Some of the data may itself be separate code (Java or Javascript) which runs on the client, but there is always a chunk of code which only exists on the server (Apache, etc in the case of the web - the things which make a server *be* a server).
This is not hard to understand. It's an essential piece of knowledge about how the net and other client-server systems work, and without it you're not equipped to deal with the GPL, particularly as it affects client-server systems.
What makes you think Google's algorithms are based on GPLed algorithms?
I'm certain they aren't. However, I'm equally certain that they'll be tied into Apache and some database system, probably hard-coded for speed. (I'm sure Google has a page telling you, but I can't be arsed to go look.) And I'm sure that Google have spent some time hacking the system as a whole for maximum performance, so removing various non-essential bits of the database or web-server systems. Whether they've released these hacks to the world, I don't know, but in many ways it makes sense for them to keep them hidden (at least for a bit) to maintain their "edge".
Fair enough - I'll take on the "distribute" point then.:-)
I wouldn't argue with "distribute" and "publish" being different. It seems we agree on "publish" meaning specifically to give copies to anyone in the public who requests it, and "distribute" meaning to generally give copies to people. (I checked that US copyright site, and that was pretty clear about "publish" meaning to "distribute" to members of the public, although it didn't clarify "distribute" further.) Where we differ is in what is happening when modified software is installed on employees' computers.
You say it's being "distributed" because it's being presented to the employee to use. I'd disagree, because I don't think it is being presented to the employee. Rather, a copy of the software modified by the company is being put on a computer owned by the company which just *happens* to be used by that employee. As I understand it (and I may be wrong), you can't "distribute" to yourself, because "distribute" requires some transfer to take place from one person/entity to another person/entity. A transfer therefore clearly doesn't happen if the company owns the PC, owns the CD/server from which the modified software is installed, and owns the modified software (OK, I know they don't exactly "own" the software, but we both know the real score and it saves typing;-). The employee is allowed to use the modified software on the PC, but at no point is it ever transferred to them (even if the employee happens to be the one who does the installation). So this doesn't quality as "distribution" - instead we're under the "copy" rules by which data is transferred from one storage medium (a CD or server) to another storage medium (the PC's hard drive), both of which are owned by the company.
Note that under this interpretation, if the company handed out a CD so that the employee could install the modified software on their home (personal) PC, the company would be screwed under the GPL. This *would* become "distribution" and therefore they'd be liable for GPL infringement. Unless of course they had some hotshot lawyers draw up some kind of contract under which the employee's home PC became the property of the company for some period of time, or some other legal weaselling. Of course, that relies on the employee to rat out their employer for violating software licensing, but the possibility still exists.
I like your point about any random employee being able to release it from the server - I'd not thought of that!:-) Of course, I think it's invalidated by the first issue (over which we disagree) but it's still a nice hack.
Doh - they can do this already by repeatedly downloading one bitmap/page/whatever.
There's no rule says that the bandwidth has to be any good. Me, if I was doing this then I'd cap the bandwidth available for people to download source - after the first 10MB in a day, reduce source bandwidth to 1Kbaud. That screws the gangbangers.
Anyway, since when was the GPL supposed to explicitly enable DOS attacks?
Sorry, but there are two ways in which your interpretation fails.
Firstly, releasing code in-house is *not* "distribution". This has been covered pretty well by other posters, but anyway. Every in-house PC is owned by the same entity, the company. Consider if you install Linux on three PCs at home - you're not "distributing" it to yourself each time, you're copying it from a single install CD/DVD/ISO that you already have. And copying explicitly is *not* covered, because the GPL differentiates between copying and distributing...
You may copy and distribute...
...and it places some specific restrictions on distributing only:-
You may modify your copy or copies of the Program... provided that you also meet all of these conditions...
b) You must cause any work that you distribute or publish... to be licensed as a whole at no charge to all third parties under the terms of this License.
But there's a second point as well, though: even if this *was* classed as "distribution", the modified code would *still* not have to be globally released. And here's for why...
You may copy and distribute the Program... provided that you also do one of the following:
a) Accompany it with the complete corresponding machine-readable source code, which must be distributed under the terms of Sections 1 and 2 above on a medium customarily used for software interchange...
If distribution of executable or object code is made by offering access to copy from a designated place, then offering equivalent access to copy the source code from the same place counts as distribution of the source code, even though third parties are not compelled to copy the source along with the object code.
This shows that even if internal in-house use counted as "distribution" (which it doesn't), then all you would have to do to satisfy the GPL is place the modified code on the same internal server from which your employees install it. There is no requirement that the rest of the world be able to access that server. Or if your employees install it from CD, put the source code on that CD - again there's no requirement to send the CD to anyone else. So in neither case will the modified code leave the company. And since the servers are company-internal, I wish you luck in finding out whether that happens or not!;-/
How does website owners being idiots affect the GPL? Will there be an explicit clause in the GPL saying "thou shalt not exclude Firefox/Konqueror/Opera/whatever from thy website?"
They really, *really* haven't thought about this. The existing GPL said that if the derivative code stayed in-house, then you didn't have to release your changes. Now they're saying "well, your software is staying in-house, but you have limitations on what you can do with it, depending on what data you handle or that data's source". Well screw that. That's precisely what everyone hates about DRM - it's restricting how software is allowed to use data that you already own.
Personally, I can see GPL3 getting zero use if that gets in. Or if anyone adopts it, there'll be an instant fork of that application, simply due to the licensing (Google for one are majorly unlikely to be releasing their search algorithms to the world), and all the active users will adopt the GPL2 fork, leaving a few people on a wasting-away GPL3 fork. That really doesn't help anyone.
Basically, this proposal is exactly what we all hate about closed-source software licensing - their ability to bait-and-switch. Get people using the software, with data that's tied to the software, and then change the licensing terms on your next upgrade. "Oh, you don't want this restriction? and you don't want to pay $x to keep your software? Then goodbye, and good luck getting your data back." In this case, the GPL team are doing the same with a software install base. "Don't like this new license? then forget about using Apache, GTK, etc. Oh, that screws your business which was previously using them legally? Too bad."
On the same topic of not thinking things through, consider the proposal to ban selected companies from using GPL software. This is even crazier. Again we're back to the ability of software licensors to arbitrarily revoke your license to use software and leave you high and dry.
Luckily we have the ability to keep going with the GTK2 license, which I predict will be the result - the GPL3 license will die, unused and unloved. The only result will be a permanent loss of credibility for RMS and the GPL in general, which would be a shame.
The cry of the man who only uses Linux on a desktop and has never heard the word "server".
Talk to your IT person. Suggest that a server crashing once a week (at random times) and losing people's work when it happens is merely a case of "1 min wasted". Then watch them laugh in your face. Alternatively, wise up...
The story I heard when I was in Scouts was based on African traditions (since Baden Powell got his military education in Africa), IIRC the Zulus. Anyway, the story goes that they hold their spear in their left hand and their shield in their right hand (or the majority will, assuming righthandedness is the majority). If you're shaking hands with someone you don't trust, you use right hands because then you still have your shield to protect yourself from him and his mates. But to show trust, you discard your shield and shake with your left hand. You're then unprotected so you're trusting them not to kill you.
Don't know whether it's true or not, but it's a nice story.:-)
FWIW, spiral staircases also spiral up clockwise so that a defender can retreat up the stairs and still use their sword in their right hand. Hence there's a historical urban legend that various European kings had squads of left-handed swordsmen specifically for digging people out of towers!
There was a billboard advert in the UK, I think for Tizer or Irn Bru or some similar drink. Anyway, the line was "I love it, and so do my bitches." Picture was of some old English upper-class bloke, with a couple of labradors...
But say someone goes into your shop, buys a soda, and then shouts, "WTF? there's a cockroach in this soda!" Bizarre sense of entitlement? Only in that you're entitled to expect what you paid for, and in the case of Sims it was people expecting "Sims Online" and not "sad-CGI-porn online" or "mugged-for-your-dinner-money online".
If the GPL code is in an EXE file, with data only being passed to other places by standard OS services, then only that EXE file needs to be made available under the GPL. However that puts some serious limits on what you can do with it, and on its efficiency.
If the GPL code is in a DLL though, so that other places call the functions directly, then things are much more efficient. However all the other places that use the DLL must also be GPL. This is the so-called "viral" nature of the GPL (which the LGPL doesn't have). Cygwin is a perfect example of this in action.
Whether you use GPL or LGPL for releasing your libraries depends 100% on your philosophy. If you want to establish a de-facto standard so that everyone can use it, LGPL is what you want (or maybe BSD if you don't care about what happens to it after you've written it). If you only want the ideologically pure to use it, you go GPL. If you want to make money, you only release it with a dollar-cost license.
Meanwhile, Microsoft itself had been making use of GPLed code for years without any of their dire warnings coming true.
Which code was GPL that MS used?
BSD now, there's plenty of cases where MS have used BSD-licensed code. But I've not heard of any GPL code. And if there had been, I really would have expected to have heard, given the publicity coup that'd be.
"The sticks" to me doesn't necessarily mean that they're 100% offroad. Nor does it mean that their roads are untouched by snowplough. And here's some news - most SUVs are not significantly better than a regular car on tackling the rough stuff, simply because most are designed for looks and not performance. To the original poster's credit, Jeeps are one of the real performers, but anyone with a Lexus, BMW, Ford Escape, Toyota RAV4 or any similar Barbie crap can forget it. Especially if you've been sucked in by one of the FWD ones instead of a proper 4x4.
More important is having the right tires (a regular car with snow tires or snow chains is a damn sight better off in the snow than any SUV), and more important than that is having the skills to use it properly. The same applies for rough tracks, only there the skills thing is your only get-out. FYI, I hang-glide, and that's taken me up a decent number of rough tracks. So far, I know of two that I wouldn't go up in my car, and one of those isn't legal for car access anyway!;-) And I'm not a very skilled driver either.
Cheaper is *the* major factor. Hybrids are way expensive right now.
Cleaner isn't really in there though. Diesels are bad for particulates, and currently there isn't anything forcing people to sort it (although future standards will do).
The ultimate clean engine though would be a diesel-electric hybrid, because the diesel emissions happen at low-efficiency load sites. Run the engine at max efficiency, and you're sorted - this is why trains have been diesel-electric for years, because even with the conversion inefficiencies, the efficiency of diesel at its best point is mindblowing. But the American market won't buy diesel, sadly, due to the disastrous diesels rushed out in the 70s and 80s. Ho hum.
More realistically, he could get it by not driving a big fuck-off tank with a big fuck-off engine. But that would be un-American...:-/ Sigh.
Grab.
PS. Yes, I know most SUV owners say "oh, but I need them to get up gravel tracks, in snow, carry stuff, etc". Similarly, most obese people say "I have problems with my glands". Both are self-deluded.
I hope no-one thinks you're serious!:-) Yeah, most of the time it's not needed - but if you do need it and it ain't there then you're a thin smear of raspberry jam...
I doubt your average squaddie has enough baggage allowance to carry 5 laptops.
In a dusty, sandy environment, the most important thing *has* to be that it's sealed. If not, you can guaranteed that all the shit that gets inside it will kill the fans, CD/DVD and possibly hard drive in pretty short order.
Plus a ruggedised laptop will be squaddie-proof - throw it across the room, drop your bags on it, throw it out of the truck or whatever, it'll survive. The same could not be said of any five regular laptops!:-/
Because you need energy to separate the "H2" from the "O". With a 100% efficiency system, you could convert H20 to H2 + O, then use a fuel cell or whatever to generate electricity/heat/motion, and you'll get back precisely the energy in electricity/heat/motion that you used to separate "H2" from "O" initially. In fact the world sucks, so you're actually at some crappy low percentage.
The only benefit from using hydrogen is that this conversion process can be done somewhere outside your town so the emissions in town are clean. Same thing with electric vehicles. So what we need is some way of storing lots of potential energy in a car, which means you need high-density batteries for pure electric, or high-density hydrogen storage for fuel cell. Hydrogen is currently looking more likely. Initial versions just used high-pressure tanks, but that needs all sorts of high pressure pumps and heavy tanks. Trouble is that molecules in a gas in a confined space will naturally want to jump about (it's what creates gas pressure), so that's a pain.
The latest approach is to stash molecules of H2 in the gaps between molecules in various fancy compounds, kind of like dropping marbles into a tube (you may have heard of carbon nanotubes being used for it, which almost exactly mimics that analogy). Being "slotted in place" allows more H2 storage without the high pressure tank. And that seems to be what this one is about.
Slashdot Mirror
No worries. The BBC could always "move into the 21st century" by buying programmes outright and freely distributing them on the web. Expect your TV license fee to rise by a couple of orders of magnitude - maybe £5k a year would cover it. Great idea, Einstein!
Grab.
I've not seen him IRL, but I've read a couple of his articles in NewScientist. Abridged versions of that book, I guess. Read any of his stuff, and you can't help thinking "Von Daniken rides again!" It's the same scattershot, unsubstantiated arguments.
As far as his "singularity" goes, his definition of it is so astoundingly vague that it could apply to anything. IIRC he defines it as an "invention after which human civilisation cannot continue in the same form". OK, so choose one. For travelling, we have the longitude problem, hydrodynamics, steam-ships, aeroplanes, the jet engine, automobiles. For communication we have the public postal system, telegraph, radio, telephone, mobile phones, internet. For data storage we have paper, the printing press, punched cards, microfiche, hard disks, floppy disks, writeable CDs. And don't forget the cheap personal computer. Every *individual* one of these inventions has changed Western civilisation in such a way that continuing in the previous form is impossible if you want to continue existing within the same society. The combined effect of all those inventions is that every few years, civilisation reaches a new "singularity". The only way you can keep the "old" ways is by opting out of the society altogether, viz the Amish and other groups.
Kurzweil's premise is that new inventions are happening more and more frequently, so we're riding an exponential increase. But the problem here is to fix what you call an "invention". Is a DVD-R an invention? Yes, it allows you to manipulate video on your PC in a way you couldn't before, but in another way it's just building on CD-R. And CD-R just builds on tapes and floppy disks. Even the iPod Nano is just a more modern version of the wind-up gramophone. We might think that things are developing at an outrageous pace today, but think of the silly sod who resigned from the Patent Office in the 1890s because "everything had been invented". Kurzweil is just the optimistic side of the same coin - but equally hard-of-thinking.
Grab.
Go hang-gliding instead! That'll stop you moaning. It might also give you an idea how bloody useless most people are at thinking in 3D. If you can't control a hang-glider (which is the aerial equivalent of a pedal-bike) then you ain't going nowhere *near* a flying car!
Grab.
But for FOSS projects, we're theoretically dealing with something in which anyone can participate. "Many eyeballs" and all that. This should mean not only that anyone *can* participate, but that anyone should be *able* to. Now if you're going to put some arbitrary restrictions on it - "we won't tell you what you need to know until you've hung out on IRC long enough to join the club" - then you're majorly restricting your eyeball count. :-/
This isn't a commercial benefit, and it isn't a benefit to the project. The only people it benefits are those people in the club who get a nice warm feeling from their exclusivity.
Yes, good spec writing *is* hard, and it's the mark of a good software engineer. It worries me that Linus doesn't appreciate specs, because this suggest to me that he is not a good software engineer. A good coder, yes, but coder != engineer. It suggests that not only does he not realise the benefits of a good spec, but also that he possibly would not be capable of writing one. And that's a pretty sad situation.
Grab.
Alternatively, maybe you could wish that the fuckwit coders who gave the thing to you had read the documentation and done a little testing to make sure that it works according to the docs, instead of changing things arbitrarily without telling anyone.
All that "my code is my design" bollocks is just that - bollocks. I can spend a week reading code to find how stuff actually works, or I can spend a few hours reading the spec that says how it *should* work. My job is *not* debugging someone else's shit code, my job is writing something that uses or interfaces to their code. If they've not done their job properly, why should I be expected to be the one to find the holes, just because they can't be arsed testing it properly?
Grab.
but with Linux where there is quite chaotic development ... it may not be desirable to have an overall spec
Surely this is exactly where you *do* want some kind of spec. Suppose several people are building separate parts of a car independently. If one person has built the wheel-arches with an 18" diameter, if the person responsible for building the wheels hands you a 20" wheel then you're screwed. It might be the most beautiful wheel in the world, but you'll have to either throw it away or take an angle-grinder to the already-assembled wheel-arches.
In the case of Linux, a spec which tries to specify in detail exactly what each part should do is doomed. But if it at least says what each part's responsibilities are, and what the interface is, then that's all you need. "Does my driver free this memory, or does yours?" "Whose code issues the callback?" Those kind of questions.
I'm actually amused by the hypocrisy of Linus here. Without POSIX, Linux, BSD and the rest would all just be yet another failed machine-specific OS. It's only POSIX compliance that's made them workable, because it's only POSIX compliance that's allowed people to move from regular UNIX to Linux/BSD. Similarly, without X then they'd be without a windowing system, and X is 100% fixed in standards documents. The whole thing is written in C or C++, which again is fixed in standards so that the code will still compile on Intel or other platforms. Sadly this seems symptomatic of many F/OSS leaders. They've taken their college hobby, managed to get paid for it afterwards, and never looked back - and this means they've missed out on any real-life experience. Of those who have had "work", too often this means university teaching, Bell Labs, Xerox Park or somewhere similar where there were no targets, no deadlines and no quality control. In other words somewhere away from *real* software engineering.
Basically, Linus seems a very, very skilled coder/hacker. However as a software engineer, I *really* wouldn't want him around me, working on safety-related software.
Grab.
Thank you Mr Xerox - if I'd wanted a copy of a FAQ, I would have got one. :-/
Instead we're on the good point that Minna raises of "how do we know this is required from the text of the GPL itself?". As Minna correctly says, the GPL FAQ doesn't count for shit in court, so what does the GPL itself actually say about this? So we're down to interpreting the GPL in real detail, and it turns out to be pretty interesting. Minna's dead right, it doesn't seem to be as clear-cut as the FAQs say. Minna knows how he/she reads it (and has put up good arguments that have tied up other posters). I know how I read it, and although it seems to match the FAQ (and contradict Minna's position), it does indeed seem a pretty gnarly bit of logic to get to the position that the FAQ states. Certainly not as clear-cut as the FAQ, and open to various legal weasellings if you so wished.
If you've got reasoning based on lines of the GPL and US copyright code to support your interpretation, let's hear it. Summaries of FAQs don't cut it.
In the example you gave, a webservice or other based upon another GPLed server is being _published_ on the web
Client-server lesson 101. Code exists on your client computer. Code exists on the server. *Data* passes between them. At no point is either piece of code "published" on the web. Some of the data may itself be separate code (Java or Javascript) which runs on the client, but there is always a chunk of code which only exists on the server (Apache, etc in the case of the web - the things which make a server *be* a server).
This is not hard to understand. It's an essential piece of knowledge about how the net and other client-server systems work, and without it you're not equipped to deal with the GPL, particularly as it affects client-server systems.
What makes you think Google's algorithms are based on GPLed algorithms?
I'm certain they aren't. However, I'm equally certain that they'll be tied into Apache and some database system, probably hard-coded for speed. (I'm sure Google has a page telling you, but I can't be arsed to go look.) And I'm sure that Google have spent some time hacking the system as a whole for maximum performance, so removing various non-essential bits of the database or web-server systems. Whether they've released these hacks to the world, I don't know, but in many ways it makes sense for them to keep them hidden (at least for a bit) to maintain their "edge".
Sheesh indeed.
Grab.
Fair enough - I'll take on the "distribute" point then. :-)
;-). The employee is allowed to use the modified software on the PC, but at no point is it ever transferred to them (even if the employee happens to be the one who does the installation). So this doesn't quality as "distribution" - instead we're under the "copy" rules by which data is transferred from one storage medium (a CD or server) to another storage medium (the PC's hard drive), both of which are owned by the company.
:-) Of course, I think it's invalidated by the first issue (over which we disagree) but it's still a nice hack.
I wouldn't argue with "distribute" and "publish" being different. It seems we agree on "publish" meaning specifically to give copies to anyone in the public who requests it, and "distribute" meaning to generally give copies to people. (I checked that US copyright site, and that was pretty clear about "publish" meaning to "distribute" to members of the public, although it didn't clarify "distribute" further.) Where we differ is in what is happening when modified software is installed on employees' computers.
You say it's being "distributed" because it's being presented to the employee to use. I'd disagree, because I don't think it is being presented to the employee. Rather, a copy of the software modified by the company is being put on a computer owned by the company which just *happens* to be used by that employee. As I understand it (and I may be wrong), you can't "distribute" to yourself, because "distribute" requires some transfer to take place from one person/entity to another person/entity. A transfer therefore clearly doesn't happen if the company owns the PC, owns the CD/server from which the modified software is installed, and owns the modified software (OK, I know they don't exactly "own" the software, but we both know the real score and it saves typing
Note that under this interpretation, if the company handed out a CD so that the employee could install the modified software on their home (personal) PC, the company would be screwed under the GPL. This *would* become "distribution" and therefore they'd be liable for GPL infringement. Unless of course they had some hotshot lawyers draw up some kind of contract under which the employee's home PC became the property of the company for some period of time, or some other legal weaselling. Of course, that relies on the employee to rat out their employer for violating software licensing, but the possibility still exists.
I like your point about any random employee being able to release it from the server - I'd not thought of that!
Grab.
Doh - they can do this already by repeatedly downloading one bitmap/page/whatever.
There's no rule says that the bandwidth has to be any good. Me, if I was doing this then I'd cap the bandwidth available for people to download source - after the first 10MB in a day, reduce source bandwidth to 1Kbaud. That screws the gangbangers.
Anyway, since when was the GPL supposed to explicitly enable DOS attacks?
Sorry, but there are two ways in which your interpretation fails.
...and it places some specific restrictions on distributing only:-
... provided that you also meet all of these conditions ...
... to be licensed as a whole at no charge to all third parties under the terms of this License.
... provided that you also do one of the following:
;-/
Firstly, releasing code in-house is *not* "distribution". This has been covered pretty well by other posters, but anyway. Every in-house PC is owned by the same entity, the company. Consider if you install Linux on three PCs at home - you're not "distributing" it to yourself each time, you're copying it from a single install CD/DVD/ISO that you already have. And copying explicitly is *not* covered, because the GPL differentiates between copying and distributing...
You may copy and distribute...
You may modify your copy or copies of the Program
b) You must cause any work that you distribute or publish
But there's a second point as well, though: even if this *was* classed as "distribution", the modified code would *still* not have to be globally released. And here's for why...
You may copy and distribute the Program
a) Accompany it with the complete corresponding machine-readable source code, which must be distributed under the terms of Sections 1 and 2 above on a medium customarily used for software interchange...
If distribution of executable or object code is made by offering access to copy from a designated place, then offering equivalent access to copy the source code from the same place counts as distribution of the source code, even though third parties are not compelled to copy the source along with the object code.
This shows that even if internal in-house use counted as "distribution" (which it doesn't), then all you would have to do to satisfy the GPL is place the modified code on the same internal server from which your employees install it. There is no requirement that the rest of the world be able to access that server. Or if your employees install it from CD, put the source code on that CD - again there's no requirement to send the CD to anyone else. So in neither case will the modified code leave the company. And since the servers are company-internal, I wish you luck in finding out whether that happens or not!
Grab.
How does website owners being idiots affect the GPL? Will there be an explicit clause in the GPL saying "thou shalt not exclude Firefox/Konqueror/Opera/whatever from thy website?"
They really, *really* haven't thought about this. The existing GPL said that if the derivative code stayed in-house, then you didn't have to release your changes. Now they're saying "well, your software is staying in-house, but you have limitations on what you can do with it, depending on what data you handle or that data's source". Well screw that. That's precisely what everyone hates about DRM - it's restricting how software is allowed to use data that you already own.
Personally, I can see GPL3 getting zero use if that gets in. Or if anyone adopts it, there'll be an instant fork of that application, simply due to the licensing (Google for one are majorly unlikely to be releasing their search algorithms to the world), and all the active users will adopt the GPL2 fork, leaving a few people on a wasting-away GPL3 fork. That really doesn't help anyone.
Basically, this proposal is exactly what we all hate about closed-source software licensing - their ability to bait-and-switch. Get people using the software, with data that's tied to the software, and then change the licensing terms on your next upgrade. "Oh, you don't want this restriction? and you don't want to pay $x to keep your software? Then goodbye, and good luck getting your data back." In this case, the GPL team are doing the same with a software install base. "Don't like this new license? then forget about using Apache, GTK, etc. Oh, that screws your business which was previously using them legally? Too bad."
On the same topic of not thinking things through, consider the proposal to ban selected companies from using GPL software. This is even crazier. Again we're back to the ability of software licensors to arbitrarily revoke your license to use software and leave you high and dry.
Luckily we have the ability to keep going with the GTK2 license, which I predict will be the result - the GPL3 license will die, unused and unloved. The only result will be a permanent loss of credibility for RMS and the GPL in general, which would be a shame.
Grab.
No problems at all.
h ang_glider.jpg
;-)
http://www.mobydisk.com/mobydisk/hanggliding/big/
More fun than should be legal!
Oh no, an entire 1 min wasted
The cry of the man who only uses Linux on a desktop and has never heard the word "server".
Talk to your IT person. Suggest that a server crashing once a week (at random times) and losing people's work when it happens is merely a case of "1 min wasted". Then watch them laugh in your face. Alternatively, wise up...
Grab.
The story I heard when I was in Scouts was based on African traditions (since Baden Powell got his military education in Africa), IIRC the Zulus. Anyway, the story goes that they hold their spear in their left hand and their shield in their right hand (or the majority will, assuming righthandedness is the majority). If you're shaking hands with someone you don't trust, you use right hands because then you still have your shield to protect yourself from him and his mates. But to show trust, you discard your shield and shake with your left hand. You're then unprotected so you're trusting them not to kill you.
:-)
Don't know whether it's true or not, but it's a nice story.
FWIW, spiral staircases also spiral up clockwise so that a defender can retreat up the stairs and still use their sword in their right hand. Hence there's a historical urban legend that various European kings had squads of left-handed swordsmen specifically for digging people out of towers!
Grab.
There was a billboard advert in the UK, I think for Tizer or Irn Bru or some similar drink. Anyway, the line was "I love it, and so do my bitches." Picture was of some old English upper-class bloke, with a couple of labradors...
But say someone goes into your shop, buys a soda, and then shouts, "WTF? there's a cockroach in this soda!" Bizarre sense of entitlement? Only in that you're entitled to expect what you paid for, and in the case of Sims it was people expecting "Sims Online" and not "sad-CGI-porn online" or "mugged-for-your-dinner-money online".
Grab.
If the GPL code is in an EXE file, with data only being passed to other places by standard OS services, then only that EXE file needs to be made available under the GPL. However that puts some serious limits on what you can do with it, and on its efficiency.
If the GPL code is in a DLL though, so that other places call the functions directly, then things are much more efficient. However all the other places that use the DLL must also be GPL. This is the so-called "viral" nature of the GPL (which the LGPL doesn't have). Cygwin is a perfect example of this in action.
Whether you use GPL or LGPL for releasing your libraries depends 100% on your philosophy. If you want to establish a de-facto standard so that everyone can use it, LGPL is what you want (or maybe BSD if you don't care about what happens to it after you've written it). If you only want the ideologically pure to use it, you go GPL. If you want to make money, you only release it with a dollar-cost license.
Grab.
Meanwhile, Microsoft itself had been making use of GPLed code for years without any of their dire warnings coming true.
Which code was GPL that MS used?
BSD now, there's plenty of cases where MS have used BSD-licensed code. But I've not heard of any GPL code. And if there had been, I really would have expected to have heard, given the publicity coup that'd be.
Grab.
What definition of "helium" are you using, fuckwit?
"The sticks" to me doesn't necessarily mean that they're 100% offroad. Nor does it mean that their roads are untouched by snowplough. And here's some news - most SUVs are not significantly better than a regular car on tackling the rough stuff, simply because most are designed for looks and not performance. To the original poster's credit, Jeeps are one of the real performers, but anyone with a Lexus, BMW, Ford Escape, Toyota RAV4 or any similar Barbie crap can forget it. Especially if you've been sucked in by one of the FWD ones instead of a proper 4x4.
;-) And I'm not a very skilled driver either.
More important is having the right tires (a regular car with snow tires or snow chains is a damn sight better off in the snow than any SUV), and more important than that is having the skills to use it properly. The same applies for rough tracks, only there the skills thing is your only get-out. FYI, I hang-glide, and that's taken me up a decent number of rough tracks. So far, I know of two that I wouldn't go up in my car, and one of those isn't legal for car access anyway!
Grab.
Cheaper is *the* major factor. Hybrids are way expensive right now.
Cleaner isn't really in there though. Diesels are bad for particulates, and currently there isn't anything forcing people to sort it (although future standards will do).
The ultimate clean engine though would be a diesel-electric hybrid, because the diesel emissions happen at low-efficiency load sites. Run the engine at max efficiency, and you're sorted - this is why trains have been diesel-electric for years, because even with the conversion inefficiencies, the efficiency of diesel at its best point is mindblowing. But the American market won't buy diesel, sadly, due to the disastrous diesels rushed out in the 70s and 80s. Ho hum.
Grab.
More realistically, he could get it by not driving a big fuck-off tank with a big fuck-off engine. But that would be un-American... :-/ Sigh.
Grab.
PS. Yes, I know most SUV owners say "oh, but I need them to get up gravel tracks, in snow, carry stuff, etc". Similarly, most obese people say "I have problems with my glands". Both are self-deluded.
I hope no-one thinks you're serious! :-) Yeah, most of the time it's not needed - but if you do need it and it ain't there then you're a thin smear of raspberry jam...
Grab.
I doubt your average squaddie has enough baggage allowance to carry 5 laptops.
:-/
In a dusty, sandy environment, the most important thing *has* to be that it's sealed. If not, you can guaranteed that all the shit that gets inside it will kill the fans, CD/DVD and possibly hard drive in pretty short order.
Plus a ruggedised laptop will be squaddie-proof - throw it across the room, drop your bags on it, throw it out of the truck or whatever, it'll survive. The same could not be said of any five regular laptops!
Grab.
Grab.
Because you need energy to separate the "H2" from the "O". With a 100% efficiency system, you could convert H20 to H2 + O, then use a fuel cell or whatever to generate electricity/heat/motion, and you'll get back precisely the energy in electricity/heat/motion that you used to separate "H2" from "O" initially. In fact the world sucks, so you're actually at some crappy low percentage.
The only benefit from using hydrogen is that this conversion process can be done somewhere outside your town so the emissions in town are clean. Same thing with electric vehicles. So what we need is some way of storing lots of potential energy in a car, which means you need high-density batteries for pure electric, or high-density hydrogen storage for fuel cell. Hydrogen is currently looking more likely. Initial versions just used high-pressure tanks, but that needs all sorts of high pressure pumps and heavy tanks. Trouble is that molecules in a gas in a confined space will naturally want to jump about (it's what creates gas pressure), so that's a pain.
The latest approach is to stash molecules of H2 in the gaps between molecules in various fancy compounds, kind of like dropping marbles into a tube (you may have heard of carbon nanotubes being used for it, which almost exactly mimics that analogy). Being "slotted in place" allows more H2 storage without the high pressure tank. And that seems to be what this one is about.
Grab.