Re: Making better use of multiple maintainers
How are things going to get better if new things like this aren't tried?
It'll certainly knock out some issues in sponsorship and NM... If your
only objections are those of the improbability of getting it to work, let
Martin run with it and find out for hisself.
On Sun, 2 Sep 2001, Marcelo E. Magallon wrote:
>>> Martin Michlmayr <firstname.lastname@example.org> writes:
> > - Bugs could get fixed sooner since more people are working on the
> > package. This would help us meeting our release goals.
> The Mythical Man Month, of which Debian is exemplary. "Given enough
> eyes, all bugs are shallow", of which Debian is also exemplary, is a
> another, prettier, way of saying "men and months are interchangeable
> commodities only when a task can be partitioned among many workers
> *with no communication among them*". It doesn't mean "throw more
> people at a problem and it will get solved faster".
> Your idea is practicable iff the group of maintainers actually
> communicate with each other, that is, the package is actually jointly
> maintained. Let me take two of my own packages as an example:
> * wmaker. It's not a complex package, but it operates under several
> different conditions (or enviroments) which are different enough
> from each other that makes it imposible for me to gain expertise
> in all of them (IOW, I don't use wmaker with GNOME, and it's very
> time consuming for me to try to figure out ways of reproducing
> bugs under those conditions, given the *absolutely* *wonderful*
> and *marvelous* documentation GNOME and its packages have)
> * celestia. Simple. Very limited scope. Not many configuration
> Would I like to find a co-maintainer for wmaker? Yes. Celestia? No.
> My criteria for considering co-maintainership is simple: "Can I cover
> all the possible ways of using this package myself?"
>  We never finished that conversation at Linux Tag.
EMACS == Eight Megabytes And Constantly Swapping
Who is John Galt? email@example.com, that's who!