Re: [gnu.misc.discuss,gnu.emacs.gnus] Free software: Packagers vs Developers
On Friday 2 July 1999, at 15 h 20, the keyboard of Per Abrahamsen
<abraham@dina.kvl.dk> wrote:
> For their own problems, maybe. For the users problems, the users are the
> experts.
As a developer, I cannot disagree more. I'm fed up with useless bug reports
from people saying "Hey, I tried to run your program and it fails". I prefer
pre-digested reports by someone knowledgeable and trustable. Judging by the
reactions of upstream authors, when I forward them bug reports, I'm not
isolated.
> > Adam di Carlo tried, on that same mailing list, to remind Debian
> > developers to be nice with upstream authors, since they are at the
> > origin of free software, after all
>
> A rather condenscending formulation.
If you don't like this wording (which is mine, not Adam's) please stop using
"middleman" as an insult. We are packagers, and proud of it.
> Most developers just curse you,
Again, I interact with many upstream authors and my experience is not this
one. I would say that 60-70 % are kind and responsive, 20-30 % are kind but
unresponsive (overload of work?) and the rest are probably retired in a desert
island.
> > No need to do so. Everything is public
> > <http://www.debian.org/distrib/packages> .
>
> Yeah, right . That is called " figure out what code he has received
> from Debian".
The Debian source package is made of the original tarball (pristine) plus a
'.diff'. Exclude from the '.diff' everything which is in the "debian/"
directory (which is meta-data) and you'll have "what code he has received from
Debian".
> > Do you know this specific package and its author? (BTW, my opinion
> > is that he is not a maniac refusing any help from Debian as a matter
> > of principle, he simply has other things to do).
>
> I have never heard of it. If he is too busy, couldn't you join as a
> co-developer of the project, or maybe even take it over?
It's more work and responsabilities. For instance, it means testing on other
operating systems. You can have a slot time to work on a Debian-specific
version and not to work on the upstream one.
> Because bugs and feature requests and don't get reported to the
> developers, if they are already addressed by the middlemen.
If they are Debian-specific, it's reasonable.
> First step is to acknowledge that there is a problem.
I don't think so. There is an issue, cooperating with other people. It is not
easy and will never be. But it's not a problem, rather an opportunity to meet
new and interesting people.
> Second step is to realise that the problem can be lessened.
Yes, but I disagree with the measures you suggested.
> > (BTW, for queso, I didn't wrote them, someone send me the patches :-)
>
> This actually makes it worse. Getting patches from someone who didn't
> wrote them is _really_ useless.
I don't think so. If you think about security, it's true that I accept patches which are not PGP-signed, but I never integrate them without reading them and testing them.
Reply to: