Hi Pedro,
On Wed, 2012-05-23 at 10:36 -0500, Pedro Giffuni wrote:
> There are no details of the relicensing process but doesn't look
> very clean to me:
> - They are basically assuming that OOo 3.3.x (where they started
> has been relicensed under ALv2, which is not true.
This is emphatically not our assumption.
> - They are ignoring the changes done during incubation, claiming
> they are only 6%, which happens to be critical for licensing purposes.
My analysis suggests that (apart from license header changes, and hdu's
cleanup of permissions - you know 'chmod -x *cxx') only around one in
twenty of the files (either all, or C++) have any change at all. Quite
possibly I got that wrong :-) Perhaps I don't count entirely new files,
there is some error bar there - but I suspect if you re-do the count,
you'll discover it is roughly right.
Please notice the 'incremental' nature of the change; ie. doing it file
by file.
> If they are basically counting files and switching headers then I guess
> that simplistic view will leave a lot of holes
The counting of files is merely a helpful curio that falls out of the
heavy lifting.
> An example, just to show it's not FUD: Apache
> OpenOffice doesn't support webdav through Neon: support for
> Neon was not relicensed so the either take our changes to use serf
> or they should drop completely that code.
Sure - so we'll have to review both bits of code and decide which to
use; plenty of time for that - this is not a one-shot jump, but an
ongoing careful unwinding process.
> The other thing is that it doesn't look like they want to build community:
> when you take someone else's code you should at least say thank you.
> And they do own a lot of beers to Oracle and to this project.
If by "taking others code" you mean importing lots of files totally
untouched by any Apache project commit (beyond arist's license header
change) ie. pristine OO.o code, and then re-basing our often substantial
changes on top of it, then updating that header and checking it back
into out repository ? That is the case for ~90%+ of the files.
It is clearly a happy thing that this is possible; however it's unclear
that thanking anyone for the very substantial amount of pure busy-work
and the horribly un-necessary project duplication that the Oracle /
IBM / Apache crew have created together is really that appropriate :-)
you're not asking for that surely ?
It might also interest you (though sadly now the commit history is
comprehensively lost) that some considerable chunk of the code that you
have in Apache OpenOffice was created by members of the LibreOffice team
over the past decade+; if we all start thanking each other left & right
it might go on for quite a long time, to the point of insincerity.
All the best,
Michael.
--
michael.me...@suse.com <><, Pseudo Engineer, itinerant idiot
