On Wed, Mar 21, 2012 at 9:00 PM, Mike Mestnik
<cheako@mikemestnik.net> wrote:
I say one can easily split technical
and legal responsibility without the need for any gray lines.
While I am certainly not opposed to your idea in principle - that everyone has something to contribute (including non-programmers) to Debian's continued success - I think that for most packages, the problem would be logistical.
From my experience working with the Debian Perl Group as a contributor but not a Debian Developer, our workflow works something like this:
1. An interested party commits desired changes with corresponding debian/changelog updates to the team repository
2. The Package Entropy Tracker notices the change, and flags it as Pending Upload
3. A Debian Developer reviews the package and provides sponsorship (uploading the work on behalf of the original contributor) if applicable, or requests further changes
When it comes to copyright and licensing information, which is typically a matter of looking through the accompanying documentation and leaving appropriate notes in debian/copyright, it is typically a small job that is done along with the rest of the packaging process. One nice way of doing a quick spot check is using "grep -ir copyright ." to find all instances of the word "copyright" in the source files. Logistically, requiring developers to wait for an external party to work on copyright information (which typically doesn't take too long in my experience) would significantly slow down at least the Debian Perl Group's ability to process and upload packages.
When it comes to translations, which I think is an area that recieves much more non-developer attention than debian/copyright files, the logistical issue still arises - but since we don't all write all of the languages in existence, we often have no choice but to seek the assistance of interested parties.
However, all that being said, I think that Debian can benefit from interested parties assisting with copyright audits. We certainly have a lot of metadata and a lot of code in the various Debian repositories - but how accurately does that metadata (e.g. license and copyright information) reflect the reality?
Moreover, there are a lot of open bug reports where we are blocked on an ITP due to incomplete or missing copyright/licensing information - it would be nice to have more eyes to look over these bugs, forward the information upstream where appropriate, and follow up on open bug reports (unfortunately, of which, there are many).
To sum up, the two places I see non-developer assistance being beneficial to the Debian project (in the context of copyright and licensing information) are:
1. Auditing of copyright/licensing information: ensure that the metadata stored in debian/copyright is correct. This can be very difficult to do as sometimes code is taken from other sources by upstream developers without attribution.
2. Following up on bugs related to copyright/licensing information: for cases where an ITP/RFP has been filed, but where copyright information is not clear from the source data, file a bug report with the upstream developers, or alternatively, ping the upstream developers in case the bug has been overlooked. Possibly spend some time investigating alternative bug trackers that the upstream developers may use instead, or their personal e-mail addresses.
Cheers,
Jonathan