Yes, I'm sure that Adobe will have no problem with him using their library in his program that he gives out to all his friends and clients.
The Flash player? Go right ahead: "Adobe provides a free license to allow you to redistribute Adobe Flash Player or Adobe Shockwave Player on your company's intranet, or with your software product or service." (here)
Ditto the XAML solution: the render is included in.NET 3.
The pragmatic reality, to borrow your phrase, is that more people have a Flash player installed than an SVG renderer.
DF is a standard and more office suites read it [wikipedia.org] than just OO.o.
Sure, but that doesn't mean any of them will still read or write it in 15 years time, which is the problem we're discussing here. And if you're arguing safety in numbers you're better of with.doc.
This doesn't solve the problem of an activation server for the OS. You boot into your VM and it can't activate because the server is non-existent. Any problem you have with open source software is magnified 10X with closed source.
So activate it before you archive it?
I still don't think re-using a 15-year old VM is non-trivial though. Why is porting the VM software easier than other software?
It's not just the source itself. OpenOffice has a really nasty slew of dependencies. Do you see all of those, and all of their interfaces, still hanging around in 15 years? As it stands now you'll struggle to build recent OpenOffice on RHEL4, say, which is still widely in use - e.g. it needs a dbus API version you can't supply without upgrading half the OS packages beyond EL4.
And I don't think having the source is necessarily any help - you can't get a contributor-standard understanding of any large codebase overnight, and OOo is huge. Certainly not a job for a novice. Our hypothetical coder in 2025 would do better reading the format documents.
Interesting idea, I wonder if this will lead to people purposely adding lines of code just to generate more revenue.
FTFA there's 46,000+ eligable lines. So until they're turning over $2m a year there's no scarcity. I don't see them getting $2m but then I'd never heard of Miro.
So, clearly, they had some reason for wanting to make the change. I'm guessing that that reason, whatever it is, didn't just vanish.
The previous blog entry explains the reason: when you post your data it spills over to your friends accounts via inboxes etc. When you delete your account they don't want to have to hunt around all of your friends' and ex-friends' accounts to clean up all of that data, and they don't want to get in a legal mess by not cleaning it up.
I'm not sure I buy that completely: unless I use Facebook's messaging to send my email address say to a friend then it will only ever be stored against my record and deleting my record should clean it all up. And deleting all messages I created, and all notifications generated by my account should clean up the rest.
Lately there has been a huge push by Certified Microsoft Professionals and their companies to call (potential) clients and warn them of the dangers of open source.
I am a Microsoft Certified Professional and I work for a Microsoft Gold Certified Partner company. I'm not aware of any push nor have I seen material from Microsoft to encourage / support us making this push. Citation needed.
I mean, can you seriously name one product that's come out of MS R&D that counts as a success (discount anything that's a blatant knockoff of a pre-existing product, embrace and extend/extinguish is not R&D)?
It's not a runaway commercial success but Office Roundtable is a very neat gadget that came from Microsoft Research - it's a 360 degree videoconferencing camera which intelligently tracks the current speaker. We don't really need videoconf here but if I could justify one I'd get one:-)
Don't port, write portable using libraries that were made for that purpose.
In an ideal world, sure, but it doesn't usually work like that. Portable libraries, where they exist, don't always have all the function you need. They don't always expose everything in a completely platform neutral way e.g. because the hardware drivers underneath have subtle differences in the way they handle things. They sometimes have bugs themselves. etc. (I'll concede I haven't done this for games but I have written more mundane apps portably.)
Bottom line: you can't assume something will work unless you explicitly test it. Sure, testing one platform will catch common bugs but you still have to playtest the game thoroughly on all platforms after you've fixed the common bugs.
Some bugs are instantly visible (e.g. code won't even compile) on some platforms and cause only hidden damage on other platforms.
Now hang on a minute - that'd only be true if you're using different compilers on each platform and that isn't something you can take lightly. Even switching between different versions of the same compiler can be a significant porting effort - you can't just assume your code will compile and work on three separate compilers for free.
Targeting a larger audience results in more sales. Who'd have guessed?:p
But at the cost of porting then testing and supporting three separate releases. TFA doesn't address the cost of that, and tripling your testing along must be huge.
Sure, it worked for him, he got a huge volume of Mac sales, but that doesn't mean the numbers will work for everyone.
If this is typical of stuff used throughout... then open sourcing would be very difficult, you'd have to separate everything like that.
That's largely irrelevant, though - it's the game code we need open sourcing and in particular the server code. As long as the game *data* (i.e. the sounds etc.) is generally available we can still use it even if it doesn't get GPLed or whatever.
No, not electrical storage density but physical density. As in mass / volume.
We're talking about replacing the Tesla car's battery with some number of these to increase it's range. Whereas we know you could install three for the same weight we don't know that three will fit inside the Tesla battery compartment.
so... what was the site license I used for years and years when I worked at a university?
Fair enough, I'd forgotten about that - that's the MSDN Academic Alliance programme. It's a special case, there's no equivalent in the commercial world. I don't think it's supposed to be unaudited though.
If you're an OEM reseller and you had a site-licence then you'd just be installing Windows on PCs and sending them out of the door. If you buy individual OEM licences then you can sell on CDs, activation keys and the basic licence / warranty docs.
NCSoft was publically adamant that Tabula Rasa development would continue too, right up until they weren't. MMO companies are always like that, they need to maintain the illusion that everything is fine even while the ship sinks.
Now, since you started with GPL, then entire project is GPL. It doesn't matter what the university or company states they owned. GPL overrules...always as previous court cases have shown.
Uh, citation needed.
If you extend a GPL work but your changes belong to someone else then you are unable to licence your changes under the GPL (because you don't own them). GPL and copyright are separate. See the GPL FAQ, item "How do I get a copyright on my program in order to release it under the GPL?" or the copyright disclaimer section at the bottom of the GPL itself.
Rubbish. If that were true, then it would be really easy to sink any GPL project.
Yes - which is why you can't contribute to GNU projects without signing over copyright to the FSF. They're paranoid about that and for good reason. (Though I suspect it's you that would get nailed to the wall for it, not the project.)
That *is* what liquidpete's talking about; he's saying if "I start with a GPL project then it's all GPL and so I've screwed them". And he's wrong.
so I will go on and say that this isn't 1994, pretty much every single OS/Programming language/browsers supports lots of characters(ok, so maybe Windows doesn't, but that hardly qualifies as an OS.)
Oh ahahahahaha. Windows NT was fully Unicode as far back as 1994. Win95 had Unicode APIs (if not good support).
Windows has always been ahead of the game on multilanguage support (at the OS level at least).
OK, fair enough. Yes, it's the 6502 and Z80 that I know best. Variable length instructions hadn't occured to me, although the 6502 was generally opcode, one-byte-argument except for a few specific cases such as long jump - same as most RISCs I know e.g. MIPS.
Slashdot Mirror
Yes, I'm sure that Adobe will have no problem with him using their library in his program that he gives out to all his friends and clients.
The Flash player? Go right ahead: "Adobe provides a free license to allow you to redistribute Adobe Flash Player or Adobe Shockwave Player on your company's intranet, or with your software product or service." (here)
Ditto the XAML solution: the render is included in .NET 3.
The pragmatic reality, to borrow your phrase, is that more people have a Flash player installed than an SVG renderer.
DF is a standard and more office suites read it [wikipedia.org] than just OO.o.
Sure, but that doesn't mean any of them will still read or write it in 15 years time, which is the problem we're discussing here. And if you're arguing safety in numbers you're better of with .doc.
This doesn't solve the problem of an activation server for the OS. You boot into your VM and it can't activate because the server is non-existent. Any problem you have with open source software is magnified 10X with closed source.
So activate it before you archive it?
I still don't think re-using a 15-year old VM is non-trivial though. Why is porting the VM software easier than other software?
It's not just the source itself. OpenOffice has a really nasty slew of dependencies. Do you see all of those, and all of their interfaces, still hanging around in 15 years? As it stands now you'll struggle to build recent OpenOffice on RHEL4, say, which is still widely in use - e.g. it needs a dbus API version you can't supply without upgrading half the OS packages beyond EL4.
And I don't think having the source is necessarily any help - you can't get a contributor-standard understanding of any large codebase overnight, and OOo is huge. Certainly not a job for a novice. Our hypothetical coder in 2025 would do better reading the format documents.
Interesting idea, I wonder if this will lead to people purposely adding lines of code just to generate more revenue.
FTFA there's 46,000+ eligable lines. So until they're turning over $2m a year there's no scarcity. I don't see them getting $2m but then I'd never heard of Miro.
Try qzone.qq.com rather than just qq.com.
HTTP/1.1 200 OK
Date: Tue, 24 Feb 2009 17:41:06 GMT
Server: QZHTTP-2.3
Content-type: text/html
Content-length: 1728
Connection: close
Ah, what does it all mean? I dunno, are the Chinese proposing some sort of new web server protocol standard? Is there a new RFC out?
They've just called their software 'QZHTTP'. Try 'telnet qzone.qq.com 80' and 'HEAD / HTTP/1.0' and you'll see for yourself:
Server: QZHTTP-2.3
I don't think anyone's suggesting there's a new protocol here.
So, clearly, they had some reason for wanting to make the change. I'm guessing that that reason, whatever it is, didn't just vanish.
The previous blog entry explains the reason: when you post your data it spills over to your friends accounts via inboxes etc. When you delete your account they don't want to have to hunt around all of your friends' and ex-friends' accounts to clean up all of that data, and they don't want to get in a legal mess by not cleaning it up.
I'm not sure I buy that completely: unless I use Facebook's messaging to send my email address say to a friend then it will only ever be stored against my record and deleting my record should clean it all up. And deleting all messages I created, and all notifications generated by my account should clean up the rest.
Lately there has been a huge push by Certified Microsoft Professionals and their companies to call (potential) clients and warn them of the dangers of open source.
I am a Microsoft Certified Professional and I work for a Microsoft Gold Certified Partner company. I'm not aware of any push nor have I seen material from Microsoft to encourage / support us making this push. Citation needed.
I mean, can you seriously name one product that's come out of MS R&D that counts as a success (discount anything that's a blatant knockoff of a pre-existing product, embrace and extend/extinguish is not R&D)?
It's not a runaway commercial success but Office Roundtable is a very neat gadget that came from Microsoft Research - it's a 360 degree videoconferencing camera which intelligently tracks the current speaker. We don't really need videoconf here but if I could justify one I'd get one :-)
He was defining civilisation as how quickly the nation gets to see the new BSG.
I know it isn't free as in speech, but Half Life 2 Deathmatch is free to all nvidia card owners.
It's also free to ATI card owners:
http://www.steampowered.com/ati_offer1a/
so between the two that'll covers most bases - or Windows-using ones anyway.
Your /. post testing process failed. Are you underfunding your testing department?
No. But I don't waste their time proofreading my /. posts either :-p.
Don't port, write portable using libraries that were made for that purpose.
In an ideal world, sure, but it doesn't usually work like that. Portable libraries, where they exist, don't always have all the function you need. They don't always expose everything in a completely platform neutral way e.g. because the hardware drivers underneath have subtle differences in the way they handle things. They sometimes have bugs themselves. etc. (I'll concede I haven't done this for games but I have written more mundane apps portably.)
Bottom line: you can't assume something will work unless you explicitly test it. Sure, testing one platform will catch common bugs but you still have to playtest the game thoroughly on all platforms after you've fixed the common bugs.
Some bugs are instantly visible (e.g. code won't even compile) on some platforms and cause only hidden damage on other platforms.
Now hang on a minute - that'd only be true if you're using different compilers on each platform and that isn't something you can take lightly. Even switching between different versions of the same compiler can be a significant porting effort - you can't just assume your code will compile and work on three separate compilers for free.
Targeting a larger audience results in more sales. Who'd have guessed? :p
But at the cost of porting then testing and supporting three separate releases. TFA doesn't address the cost of that, and tripling your testing along must be huge.
Sure, it worked for him, he got a huge volume of Mac sales, but that doesn't mean the numbers will work for everyone.
If this is typical of stuff used throughout... then open sourcing would be very difficult, you'd have to separate everything like that.
That's largely irrelevant, though - it's the game code we need open sourcing and in particular the server code. As long as the game *data* (i.e. the sounds etc.) is generally available we can still use it even if it doesn't get GPLed or whatever.
No, not electrical storage density but physical density. As in mass / volume.
We're talking about replacing the Tesla car's battery with some number of these to increase it's range. Whereas we know you could install three for the same weight we don't know that three will fit inside the Tesla battery compartment.
so... what was the site license I used for years and years when I worked at a university?
Fair enough, I'd forgotten about that - that's the MSDN Academic Alliance programme. It's a special case, there's no equivalent in the commercial world. I don't think it's supposed to be unaudited though.
If you're an OEM reseller and you had a site-licence then you'd just be installing Windows on PCs and sending them out of the door. If you buy individual OEM licences then you can sell on CDs, activation keys and the basic licence / warranty docs.
Ever heard of a "site license?"
For XP? There's no such thing. The best you can do is Microsoft's Volume Licence programme but you still need to pay per number of installs.
But as the other guy said this guy said he was a reseller and needs to sell the licences on to third parties.
NCSoft was publically adamant that Tabula Rasa development would continue too, right up until they weren't. MMO companies are always like that, they need to maintain the illusion that everything is fine even while the ship sinks.
But they are still developing even though they've admitted the ship is sinking. They pushed a new release to the test server this week.
Now, since you started with GPL, then entire project is GPL. It doesn't matter what the university or company states they owned. GPL overrules...always as previous court cases have shown.
Uh, citation needed.
If you extend a GPL work but your changes belong to someone else then you are unable to licence your changes under the GPL (because you don't own them). GPL and copyright are separate. See the GPL FAQ, item "How do I get a copyright on my program in order to release it under the GPL?" or the copyright disclaimer section at the bottom of the GPL itself.
Rubbish. If that were true, then it would be really easy to sink any GPL project.
Yes - which is why you can't contribute to GNU projects without signing over copyright to the FSF. They're paranoid about that and for good reason. (Though I suspect it's you that would get nailed to the wall for it, not the project.)
That *is* what liquidpete's talking about; he's saying if "I start with a GPL project then it's all GPL and so I've screwed them". And he's wrong.
so I will go on and say that this isn't 1994, pretty much every single OS/Programming language/browsers supports lots of characters(ok, so maybe Windows doesn't, but that hardly qualifies as an OS.)
Oh ahahahahaha. Windows NT was fully Unicode as far back as 1994. Win95 had Unicode APIs (if not good support).
Windows has always been ahead of the game on multilanguage support (at the OS level at least).
OK, fair enough. Yes, it's the 6502 and Z80 that I know best. Variable length instructions hadn't occured to me, although the 6502 was generally opcode, one-byte-argument except for a few specific cases such as long jump - same as most RISCs I know e.g. MIPS.
Hmmm, and here I thought that old-school 8-bit computing is the future, since more than half of all CPUs sold are 8-bit processors.
But if you had to categorise them as RISC or CISC it'd have to be RISC. None of them are particularly complex, there's no microcode in them, etc.