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

Re: RFC: Rules for distro-friendly packages

* Ian Jackson <ijackson@chiark.greenend.org.uk> schrieb:

> "Rules" is just wrong.  "Guidelines" might be appropriate, but I would
> suggest "Recommendations".

In my case it are rules, which are required to get an qm stamp in
my company. Actually, it's just the tip of an iceberg of processes,
I can't describe in detail right now.

Of course, nobody forces other people to follow these rule, but it
would be a great help of upstreams would do so.

> > That might come from the backgroud that it's actually meant to
> > form a set of QM requirements. I'm going to implement an checklist
> > system for the OSS-QM project [1] where all the releases are
> > undergoing an approval process.
> Undefined footnote error in "[1]".  So I'll guess what "OSS-QM" is.


> In general, if you recast your document as "we will take as input
> upstream packages in general, and as output will will produce packages
> which conform to these rules", then that's fine, because _your_
> propject's output is something that you are legitimately in charge of.

Yes, but that's just a part of that. These rules are necessary to
run through our internal qm processes and build engines. (more of
it will be published soon).

> Although I don't think that many Debian maintainers would prefer to
> add an additional layer of intermediary between themselves and
> upstream.

They don't need to. OSS-QM provides canonical repositories [2],
where everybody can fetch from easily (especially built for 
automated systems). Debian changesets are also imported automatically.
(as far as possible)

> Also, I guess "Q" stands for Quality.  If you're going to be talking
> to upstreams I would avoid using the word Quality to describe your
> recommendations.  Upstreams will have their own ideas about what
> constitutes quality for their software (ie, depending on their own
> goals), so you are at best going to cause confusion. 

I'm currently writing a paper which describes the whole issue
more in detail - that should clear up the confusion.

 Enrico Weigelt, metux IT service -- http://www.metux.de/

 phone:  +49 36207 519931  email: weigelt@metux.de
 mobile: +49 151 27565287  icq:   210169427         skype: nekrad666
 Embedded-Linux / Portierung / Opensource-QM / Verteilte Systeme

Reply to: