TextIndexNG or the problems of group maintainance

[Andreas Tille]

Hi,

I was the original maintainer of zope-textindexng2 package and handed
it over to group maintainance because I did not really used it any more.
At this time I liked this group maintainance because it took some
burden from me.  The change to group maintainance was done in times
we were under heavy critics about having very old and in fact unmaintained
Zope products in Debian (I might be able to research the actual e-mail
that did a comparison between product versions in Debian and latest
upstream versions).  I do not have the time to do a research on the
current status but my impression is that the group maintainance has
a weak part because we do not have a real person who is responsible
for a certain package and thus some packages get out of focus.  For
instance this is the case with zope-textindexng2.  I personally
trusted that "anybody" of the group will care and lost interest
(perhaps it is my fault - I do not want to blame somebody else for
this).  Today the website

    http://www.zope.org/Members/ajung/TextIndexNG

says: TextIndexNG2 is no longer maintainted and the successor is
TextIndexNG3 at

    http://opensource.zopyx.com/software/textindexng3

in version 3.2.1 - so we are way behind upstream.  Another
package I handed over to group maintainance is zope-testcase.
This package was completely droped - well, this might be better than
unmaintained, but it was a nice tool for quality checking.

So the question is: Are we able to avoid the situation that some
packages come out of our focus and become way outdated?  How can we
avoid that single maintainers blindly trust a group that has no
generic interest in a single package?

IMHO a first step would be to do an automatic uscan and publish
the result on a web page or even better send a monthly uscan report
to the mailing list.  Any other thoughs about this?

Kind regards

         Andreas.

-- 
http://fam-tille.de