Hello everybody, Although it should never happen, sometimes a new package we submit to our archive managers is rejected because the description of the copyright status of its files is either incorrect or lacunar. This is waste of precious time for everybody. In order to ameliorate the quality of our submissions, I propose to introduce a dose of optional pre-submission peer review.
One may wonder why not reviewing all QA aspects of a package? One reason is that this requires skills that are not evenly distributed, and for some aspects is simply a matter of taste. Another reason is that such kind of review is already taking place, is scaterred on many different teams, mailing lists, and social networks, and would be difficult to centralise. In contrary, all package maintainers contributing to Debian must be able to write a good copyright file, and follow the same guide: exhaustivity. I therefore think that it should be possible to centralise the effort of reviewing debian/copyright files of new packages even if it mixes people with various backgounds. I have drafted a page on the wiki that summarises the motivations and proposes a mode of operation. The key principle is peer review: the ones who review the work of others are the ones who need a review for themselves. Such a system should be self-sustainable and will not require the goodwill of persons who are not using the system themselves. http://wiki.debian.org/CopyrightReview I welcome everybody interested by the concept to help me to bootstrap the system, by writing reviews and submitting their own new packages. Would it be sucessful, the system could be extended to new upstream releases where the upstream diff is really big, via the use of RFH bugs. Have a nice day. PS: and of course, consider using the machine-readable format if you have not tried yet: http://wiki.debian.org/Proposals/CopyrightFormat PPS: If there is interest this proposal could be turned into a DEP. -- Charles Plessy Tsurumi, Kanagawa, Japan You can take your time if you want to answer: I am going to sleep -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
