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

Re: NEW handling ...



On Thu, Mar 17, 2005 at 07:31:39PM +0100, Remi Vanicat wrote:
> Sven Luther <sven.luther@wanadoo.fr> writes:
> 
> > Now the idea was to find some way to help them along, and this may be the
> > solution to it. Notice that they still have veto right so nothing can get past
> > them if thet don't want.
> >
> > Having them take positive action to counter the NEW review team or the
> > automated scripts may speed things up somewhat though.
> >
> > /me wonders what the ratio of really-new over not-new NEW packages are anyway.
> 
> I've tried to find some information about it:
> 
> $ grep '^<td valign="top" class="[se][ix][dp]">[a-zA-Z0-9+.-]*</td>' new.html | sed 's/.*>\(.*\)<.*/\1/'  | wc
>     396     396    4927
> $ for i in $(grep '^<td valign="top" class="[se][ix][dp]">[a-zA-Z0-9+.-]*</td>' new.html | sed 's/.*>\(.*\)<.*/\1/' ); do grep-dctrl -eq -P "^$i$" ftp.fr.debian.org_debian_dists_sid_main_source_Sources && echo $i does exist || echo $i does not exist; done | grep "does exist" | wc
>      69     207    1415
> 
> so it seem that 69 of 396 packages are not really new in the queue the
> 17.03.2005 at 18:07:03(UTC)

A patch to lisa[1] that makes it order the 'new binaries' packages above
the 'full new sources' would, I think, be appreciated.

I know it might be hard to test the patch, but SELECT queries on the
database work on merkel, so partially testing at least the 'does it
indeed detect correctly whether the source is NEW or not' should be
possible. Or you can install dak[2] and really try it out yourself.

--Jeroen

[1] http://cvs.debian.org/dak/lisa?cvsroot=dak
[2] http://packages.debian.org/unstable/devel/dak

-- 
Jeroen van Wolffelaar
Jeroen@wolffelaar.nl (also for Jabber & MSN; ICQ: 33944357)
http://Jeroen.A-Eskwadraat.nl



Reply to: