[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: CDBS documentation



Coin,

Don Armstrong <don@debian.org> writes:

> A changelog is not a copyright notice. A changelog doesn't indicate
> who holds the copyright on the work, when the work was created, under
> which license the work has been released under, nor does it disclaim
> warranty. [The latter two aren't part of a copyright notice either,
> but are other bits covered by this section of the GPL.]

I do think you're right it is not a copyright notice. But as long as we
stop maintaining copyright information in a per-file basis to agregate
information in a changelog file, i do think this should be considered as
a replacement for the missing copyright notices which is comonly
accepted for mere convinience.

> That's not what it says at all; it may be an intention,[1] but it's
> definetly not what the verbiage says. Indeed, a mere changelog can't
> even satisfy your interpretation of this requirement, because the
> changelog isn't present within the modified file.

So where should we find all these copyright information then ?

It is said in FSF texts that we should record information of changes so
as to determine if a sufficient amount of contribution has been reached
to know if the person can claim a copyright. How is that possible if
files stop holding information about changes when a new contributor
comes or when the central changelog omit all the details about these
changes ?

You have to fulfill licence's conditions, but you also have to fulfill
law's conditions first, and it is clearly impossible if you destroy
proofs of who changed what to determine validity of any copyright claim.
So i guess according to jurisprudence this is not an accetable thing.

> That being said, a complete changelog is a good thing from a social
> perspective, as it enables people to figure out why different people
> have changed things in the work... but it's clearly not a requirement
> of the license. I'd suggest making sure the changelogs are as complete
> as possible, but that's just because it's the right thing to do, not
> because the license requires it.

I may be wrong, but i don't think so, even if my knowledge of the law is
not as good as needed to find the best arguments. But if i'm wrong, i do
think too this is a good thing to keep changelogs intact, as this is a
good track of what happened and respectful to previous authors's work.

Seems the start of it was Peter considering viewing this changelog
getting longer and longer at the begining of the document is ugly and
unecessary to most users. I suggested either merging changelogs
externaly or commenting in XML so it would not be displayed, but Peter
clearly do not care about such things...

-- 
Marc Dequènes (Duck)

Attachment: pgpO5J5Upzpoc.pgp
Description: PGP signature


Reply to: