bitmap: some other free software licenses do have automatic "or later" clauses, the creative commons license also have a limited for of built-in "or later". But not GPL.
2014-01-07 00750, 2014
bitmap
warp: oh, good point. I'll make sure to update the header
2014-01-07 00732, 2014
warp
bitmap: great, thanks :)
2014-01-07 00753, 2014
zas__
perhaps MetaBrainz Foundation should publish some "official" licensing recommadations, so all projects have coherent licensing and headers/files
2014-01-07 00732, 2014
zas__
Copyright (C) 2013 MetaBrainz Foundation and contributors (see AUTHORS.txt)
2014-01-07 00709, 2014
warp
I'm not a fan of Gilligan's Island copyright notices.
zas__: so for individual files I think it is important that the copyright notice mentions the actual developer(s) who worked on the file (or who paid for it in a work-for-hire situation).
2014-01-07 00712, 2014
warp
for the website a gilligan's island copyright is perhaps appropriate, if it links to a full list of contributors in some way.
2014-01-07 00715, 2014
zas__
warp: yes, but as you can see, this isn't done
2014-01-07 00710, 2014
bitmap
warp: does this look okay for a single-line notice? "Licensed under the GPL version 2, or (at your option) any later version: http://www.gnu.org/licenses/gpl-2.0.txt"
I personally do not like having contributors mentioned by name/individually in each file they've touched. If that history is needed, the commit (message, if not straight up author) information should be able to disclose that if needed.
Freso: that doesn't help when a single file is re-used in a different open source project.
2014-01-07 00720, 2014
Freso
warp: Sure. But that different project should note where the code comes from, exactly so it's possible to track down.
2014-01-07 00754, 2014
ruaok joined the channel
2014-01-07 00757, 2014
Freso
(I've personally taken great care to import revision histories of files I've imported into my own things from other repositories, exactly so that attribution can be tracked.)
2014-01-07 00702, 2014
warp
Freso: right, and a real project will do that.
2014-01-07 00700, 2014
warp
Freso: but we live in a world where people throw shit on github without any consideration of licenses and stuff like that. The easiest way to make sure your work is attributed and licenses correctly is to keep short attribution and license information in the files themselves.
2014-01-07 00728, 2014
Freso shrugs
2014-01-07 00731, 2014
Freso
I still don't like it. :)
2014-01-07 00707, 2014
warp
Freso: because it's too much work to maintain and people forget to maintain it?
2014-01-07 00711, 2014
Freso
Because it can infuse some unhealthy feeling of ownership on the file.
2014-01-07 00745, 2014
Freso
Doesn't seem to much of an issue with MusicBrainz (but that's possibly just because there are so few devs involved in the first place...).
2014-01-07 00719, 2014
warp
hm, where I think it is important that authors are rewarded for their work by at least having their name attached to it, you think that is arrogant boasting? :)
2014-01-07 00702, 2014
Freso
Note that I'm not saying that they shouldn't be added to e.g. an AUTHORS file or some such, nor did I anywhere mention anything about boasting... ?
2014-01-07 00704, 2014
Freso
"Feeling of ownership" may just as well be from some other person: "Oh, I shouldn't try and fix bugs in the release editor since bitmap 'owns' that."
2014-01-07 00717, 2014
warp
hm
2014-01-07 00751, 2014
Freso
There's also the question of *when* someone should be credited in a file.
2014-01-07 00703, 2014
Freso
And what happens when that file is split up into several files?
2014-01-07 00705, 2014
warp
it can also be helpful, "Oh I want to fix this thing in this file, but I'm not sure how this works, I'll ask bitmap since he wrote most of it"
2014-01-07 00714, 2014
warp
although a "git blame" is probably more useful for that
2014-01-07 00723, 2014
Freso
Sure. But that can also be ach... yeah.
2014-01-07 00749, 2014
warp
Freso: the "when" is easy, when they make changes to it.
2014-01-07 00741, 2014
Freso
warp: So when I correct "theyre" to "they're", I get an entry the same weight as when you rewrote 86% of the code?
2014-01-07 00742, 2014
Wixy joined the channel
2014-01-07 00747, 2014
Wixy joined the channel
2014-01-07 00717, 2014
warp
I can imagine the "unhealthy feeling of ownership", but I don't think I've seen that in any real projects. perhaps you have, in which case I can see it may be useful for that project to have a different policy.
Freso: no, you only add a copyright notice if your contribution is eligable for copyright
2014-01-07 00749, 2014
warp
the weight doesn't really matter, every contributor who has copyright in the file should have a copyright notice in the file.
2014-01-07 00744, 2014
ruaok
:wq
2014-01-07 00754, 2014
Mineo
good bye ruaok
2014-01-07 00755, 2014
ruaok looks sheepishly
2014-01-07 00703, 2014
ruaok
heh
2014-01-07 00729, 2014
Freso
warp: And so, when is something eligible for copyright? Is changing an if clause in a file to mirror changes to an identical if clause elsewhere in the project (made/thought up by a third person)?
2014-01-07 00703, 2014
Freso
If not, should the change be attributed to that third person who otherwise hadn't touched that file previously?
2014-01-07 00734, 2014
warp
Freso: if you're not sure whether your contribution is large enouh to be copyrightable, it probably isn't :)
2014-01-07 00746, 2014
Freso wonders what the w in ruaok's :wq is supposed to write
2014-01-07 00719, 2014
ruaok
a python hellow world script.
2014-01-07 00733, 2014
Freso
warp: Or maybe it is. I personally wouldn't want to judge that, because I think copyright sucks and I'd just not have to make the judgement call.
2014-01-07 00753, 2014
ruaok is being geeky with his woman
2014-01-07 00757, 2014
Freso
:)
2014-01-07 00741, 2014
warp
Freso: there is more harm in overusing copyright notices than underusing them, so what I said is a good guideline nonetheless.
she's working on her own version. this one's mine.
2014-01-07 00717, 2014
ruaok
well, ours, actually. :)
2014-01-07 00729, 2014
reosarevok
Learning python together?
2014-01-07 00756, 2014
reosarevok has no idea if you're already fluent in python
2014-01-07 00744, 2014
Freso
warp: I'll look at it. Does it answer my question about the split file you never replied to?
2014-01-07 00750, 2014
ruaok
she's just aced her coursera python course.
2014-01-07 00717, 2014
ruaok
I've yet to take that class. I bet I could learn a lot. ;-)
2014-01-07 00719, 2014
warp
Freso: most projects don't keep track of which parts of a file each contributor has a copyright on, so I would consider the file as a whole to be the copyright of all contributors to the file.
2014-01-07 00749, 2014
warp
Freso: so if you split the file, both sides of the split get all the contributors as copyright "owners".
2014-01-07 00721, 2014
warp
Freso: ofcourse with a well-maintained git repository you can be more accurate if you want.
2014-01-07 00742, 2014
Freso
So what about the if clause example?
2014-01-07 00753, 2014
warp
Freso: if you change a single if clause, that is not a copyrightable change.
2014-01-07 00707, 2014
Leftmost
So many committers don't set authors. Grumble grumble.
2014-01-07 00758, 2014
Freso
warp: I'd say that heavily depends on the complexity of the if clause.
2014-01-07 00759, 2014
warp
Freso: if you copy/paste a large chunk from a different file, obviously that chunk will have a license and most floss licenses require you to keep copyright notices, so usually you'd copy those along as well.
2014-01-07 00700, 2014
warp
Freso: that must be some seriously massive if clause though :)
2014-01-07 00719, 2014
Freso
How much does it require to be copyrightable? If there's a definite change of logic (ie., it's a different idea), I would likely say it's copyrightable. No matter the size.
2014-01-07 00720, 2014
reosarevok
Leftmost: I never set author because the file already says MetaBrainz Foundation, which is good enough for me
2014-01-07 00733, 2014
reosarevok
(well, when doing MB stuff that is)
2014-01-07 00747, 2014
Leftmost
reosarevok, I think it's useful for a lot of stuff, not just copyright.
2014-01-07 00702, 2014
reosarevok
For the only other stuff it's useful for we have git blame
2014-01-07 00729, 2014
Leftmost
Correct and simple patch attribution through git, Ohloh narcissists...
2014-01-07 00713, 2014
warp
Freso: it cannot be a mere functional change, creativity is a requirement for copyrightability.
2014-01-07 00720, 2014
Freso
ocharles gave me slack at some point for using MetaBrainz instead of my own name for adding a file to MBS that was basically a copy of a ton of other files, just with a single string changed. (Ie., nothing I would claim was copyrightable.)
2014-01-07 00727, 2014
reosarevok
502...
2014-01-07 00732, 2014
warp
Freso: ocharles could have been wrong :)
2014-01-07 00724, 2014
Freso
warp: "A different idea". Ie., a different way of approaching a problem. I don't see how that's not creativity.
2014-01-07 00732, 2014
reosarevok wonders what's with http://musicbrainz.org/edit/14790517
2014-01-07 00745, 2014
Freso
warp: Yeah. One of the reasons I didn't comply with his "request" to replace MetaBrainz with my name. :)
2014-01-07 00754, 2014
warp
"If the creative component of a work cannot be separated from its functional component, it cannot be copyrighted."
2014-01-07 00756, 2014
reosarevok
Not that it matters, since I also wonder why the hell we didn't just drop all PUID edits like we did with the TRM ones
2014-01-07 00713, 2014
reosarevok
warp: isn't that true of almost all code?
2014-01-07 00727, 2014
reosarevok
The work *is* the functionality after all..
2014-01-07 00759, 2014
warp
reosarevok: I think the idea is that you cannot copyright code which can only be implemented in one way.
2014-01-07 00710, 2014
reosarevok
Hmm
2014-01-07 00727, 2014
reosarevok
But then, if you approach the problem in a different way, it can obviously be implemented in two ways :p
2014-01-07 00726, 2014
Freso
warp: And there we have it: the if clause of the example can obviously be implemented in two different ways, since two different ideas can be the base of solving the same problem. Thus, there's not just one way to implement the solution to the problem.
2014-01-07 00721, 2014
Wixy joined the channel
2014-01-07 00730, 2014
reosarevok
And there we have it, Freso proving there are also two way of saying the same, one more pedantic than the other
2014-01-07 00731, 2014
reosarevok
:D
2014-01-07 00736, 2014
reosarevok
*ways even
2014-01-07 00736, 2014
Freso
And the new if clause, based on a different idea/approach to the problem but solving the same problem in the end, must thus be copyrightable.
2014-01-07 00756, 2014
Freso
reosarevok: I didn't read yours until after mine was sent though.
2014-01-07 00757, 2014
warp
Freso: I'm still not convinced the contribution is substantial enough.
2014-01-07 00747, 2014
Freso
warp: Which brings me back to my original issue with "when is something appropriate to include in a file's copyrightee's list?" as people do not necessarily agree on the answer to that. :)
2014-01-07 00727, 2014
warp
Freso: the GNU project maintains 15 lines, which seems like a reasonably ballpark, but I have no idea if it is based on any real case law.
legal matters aren't code, it is never precise -- it is all about managing risk.
2014-01-07 00733, 2014
warp
different judges will rule differently, different countries have different ideas about copyright, different projects have different policies, etc..
2014-01-07 00735, 2014
Freso
Which is why it's far easier to do all or none. Between the two is a grey zone that depends too much on individual developer attitude (e.g., I'm unlikely to ever add myself to a list, other people might want to be added for trivialities), cultural differences, judicial system differences, etc.
2014-01-07 00737, 2014
warp
Freso: where you keep track of copyright is orthogonal to this question though
2014-01-07 00703, 2014
Freso looks up orthogonal
2014-01-07 00712, 2014
Freso
In a sec when I've got noms.
2014-01-07 00721, 2014
warp
Freso: if you keep copyright notices in a central location or in each file, you still need to figure out whether a contribution is significant enough to add a copyright notice somewhere.
2014-01-07 00723, 2014
warp
central location vs in each file are both valid approaches, I prefer the each file approach obviously.
2014-01-07 00735, 2014
Freso
I prefer the "in commit history" approach. :)
2014-01-07 00757, 2014
warp
but then you don't have any useful copyright information if your project is distributed as a .zip or .tar.gz file
2014-01-07 00723, 2014
warp
(granted, I think that is a bad idea anyway, commit history is important and should always travel with the code :)
2014-01-07 00713, 2014
Freso
For Git, you include the .git :p
2014-01-07 00706, 2014
Freso
Man.
2014-01-07 00711, 2014
Freso
y'know.
2014-01-07 00717, 2014
Freso
I *don't* miss CVS.
2014-01-07 00757, 2014
warp
Freso: in a way, the linux "Developer Certificate of Origin" is the "in commit history" approach.
I like point (d) in particular, since I know at least a few people who have been surprised their e-mail was suddenly "in the wild" after having used GitHub etc.
2014-01-07 00723, 2014
warp
somewhat related, the GPL says:
2014-01-07 00734, 2014
warp
"The “source code” for a work means the preferred form of the work for making modifications to it. "
2014-01-07 00706, 2014
warp
I think nowadays "the preferred form of the work for making modifications to it" is a git repository
2014-01-07 00710, 2014
Freso
2, 3, or both?
2014-01-07 00744, 2014
Freso
(Mostly because I'm curious if that definition has changed over time. :))
2014-01-07 00747, 2014
warp
and I think larger projects should publish some kind of policy statement to that effect along with their GPL license :)
2014-01-07 00705, 2014
warp
Freso: this is from GPLv3
2014-01-07 00711, 2014
warp
GPLv2 is almost identical, "The source code for a work means the preferred form of the work for making modifications to it."
2014-01-07 00758, 2014
warp
(although it adds some more detail in subsequent sentences which are quite different from GPLv3)
2014-01-07 00742, 2014
Freso
I would imagine, yeah. I think this is one of the areas GPLv3 is supposedly improved over v2. :)
2014-01-07 00720, 2014
warp
enough rambling on this topic, I'm going to play outside for a bit and then sleep :)
