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

Re: Bug#466550: Pristine source from upstream VCS repository



On Thu, Mar 12 2009, Russ Allbery wrote:

> Manoj Srivastava <srivasta@debian.org> writes:
>
>>  a) Run a upstream version check from cron, which mails me if there are
>>     new upstream versions of something I have.
>>  b) If there is a new upstream version, cd checked out dir
>>     1. No munging required: use uscan --rename --verbose to get the
>>        latest source.
>>     2. Munging needed. Run get-orig-source to get the latest upstream
>>        source via uscan; and munge it as needed to create the
>>        orig.tar.gz file
>
> Oh, okay, so your get-orig-source target would internally use uscan.

        It _could_ use uscan. it does not have to be limited to it.

> How do you tell from that what tarball it downloaded for an automated
> target?  Would you parse the output of uscan somehow?

        I just glob for the same pattern as in the watch file, and use
 the last in the lexical sorting, I suppose one could use dpkg
 --compare-versions if one were paranoid enough, and heck, use shell
 sort on the orig tar balls discovered :P

>>  c) Proceed as per:
>>     http://www.golden-gryphon.com/blog/manoj/blog/2009/02/25/A_day_in_the_life_of_a_Debian_hacker/>
>>         Is this so very different from what people do? Some times I do
>>  not package every upstream version, if they are coming in rapid
>>  succession, or if I find some version unfit for Debian -- but in any
>>  case, the majority of the time I want to package the very latest
>>  upstream version.
>
> I never use uscan --download; I always download the new upstream source
> myself using wget or a web browser or FTP client.

        But this is not about our individual work-flows -- it is about
 policy trying hard not to proscrive the work flows _any_ of us use. If
 it turns out there are two sets of processes people follow, I would
 much rather have two mechanisms, with two different sets of semantics,
 rather than have us select one in policy.

        I am beginning to think this whole target is too immature to
 actually be in policy; we are still doing design discussions of this
 feature.

        manoj
-- 
The real problem with hunting elephants is carrying the decoys.
Manoj Srivastava <srivasta@debian.org> <http://www.debian.org/~srivasta/>  
1024D/BF24424C print 4966 F272 D093 B493 410B  924B 21BA DABB BF24 424C


Reply to: