[systemd-devel] [PATCH] hwdb: ship ids-update.pl & sdio.ids in the release tarballs.

Zbigniew Jędrzejewski-Szmek zbyszek at in.waw.pl
Sun Mar 22 12:49:58 PDT 2015

On Fri, Mar 20, 2015 at 09:21:33AM -0700, Marcel Holtmann wrote:
> Hi Zybszek,
> >>>>>>>>>>> On 16 March 2015 at 23:15, Marcel Holtmann <marcel at holtmann.org> wrote:
> >>>>>>>>>>>> Hi Dimitri,
> >>>> 
> >>>>>>>> Just tell patch or git to skip the hunks modifying ids-update.pl and sdio.ids. Problem solved.
> >>>>>>> 
> >>>>>>> I'll apply the patch, but with a slightly different motivation.
> >>>>>>> 
> >>>>>>> [L]GPL requires commercial entities distributing a modified version of
> >>>>>>> the program to provide full source in the preferred form for modification,
> >>>>>>> including all scripts used for building. This includes sdio.ids and
> >>>>>>> ids-update.pl. We should make it easy to follow the our licensing, so
> >>>>>>> we should include those files in our tarball to make it directly
> >>>>>>> redistributable.
> >>>>>> 
> >>>>>> that is just making stuff up.
> >>>>> Where was I wrong?
> >>>>> 
> >>>>>> Are you now also including usb.ids?
> >>>>> We probably should
> >>>>> 
> >>>> 
> >>>> Right, we should include all source files, not only the generated files.
> >>> 
> >>> is this turning in Freaky Friday now? Can we please be real for a bit.
> >>> 
> >>> You do realize that I did not put a copyright header on top of sdio.ids. In addition this information will be most likely considered a not copyrightable piece of source anyway. If we assume that this is non-copyrightable, then how can you apply a copyleft license to it?
> > 
> > It's not about applying a license to sdio.ids. If may just as well be
> > non-copyrightable. It's just distributed because in the tarball because
> > it is one of the things necessary to generate a patch.
> > Now, in practice, any sane person would clone the repository and do
> > modifications there, but the terms of our license demand that
> > redistributors redistribute the source. I think it makes sense to
> > set things up so that if they just use the tarball, they actually
> > are doing that.
> So where do you think the Bluetooth IDs are coming from? Or the USB IDs or PCI IDs.
> For me this sounds that someone want to be really lazy with back porting and not add the needed extra switch to exclude these two files while generating patches and/or applying patches. And out of a sudden some bogus license argument gets dragged into it.

Please keep things in perspective. This is a two line patch which
increases the size of the tarball by some insignificant amount. So
actually the fact that it makes the life of some downstream users
easier would be a good-enough justification for me.

> please get your story straight. Now it is because of generating a patch. You can tell git to leave certain files or directories out when generating patches. It is plain simple.

Dimitri was talking about applying patches. I was talking about generating
patches. I hope you can see the difference.

To make an analogy: a common upstream error is to say "This work is
licensed under BSD license found in file LICENSE", but then there's
actually no file named LICENSE in the tarball. For all intents and
purposes you can assume that the project is available under BSD
license, but as a packager for a distribution you'll file a bug for
the project to include the file and fix the discrepancy.

This is similar: we are saying "you can distribute copies, binary or
otherwise, as long as you provide the source code", but then *we*
don't actually provide full source code in the tarball.x


"The source code for a work means the preferred form of the work for
making modifications to it." — gpl-2.0.txt

More information about the systemd-devel mailing list