On Tue, 2004-03-09 at 13:08, Leonard den Ottolander wrote:
Hi Alexander,
> It seems its always the fault of the developer when something isn't
> ideal. The developer should just write more devel docs, should just
> write more docs, should just fix more bugs, should just communicate with
> the community more, should just add that new important feature.
Actually I was blaming management not to manage the available resources
well enough. This could mean letting developers spend more time on
fixing bugs and communicating with/educating volunteers, or by setting
up structures that address these issues (package managers besides
package developers). I am well aware of the fact that individual
developers are pressed for time, but that is because management expects
certain things of them. So what I am saying is that management lacks
vision.
Management tells me very very little what to do in details like this. I
make almost all decisions about what to do in the small scale myself. Of
course, there is a larger plan that controls how I schedule my time, but
that doesn't say whether I should spend time reading this mail or if I
should fix a bug.
Also, there is all this talk about "Management" with I just don't
understand. Redhat is a company with managers, but Fedora is about free
software development. People are supposed to work on it because they
like to, to scratch itches, to learn stuff, to have fun. If all you want
is to have a Redhat fix bugs for you the best way do to that is to spend
money on Redhat Enterprise Linux, that means we'll get more paid
developers and will be able to fix more bugs (in fedora too, fedora is
what will become the next RHEL version).
> Why isn't it never the fault of the person who wants to fix
a difficult
> bug that he didn't spend enough time trying to understand the code,
> instead of the developer not spending enough time writing docs for
> something that probably only that person needs.
Obviously trying to understand the code yourself is the approach often
taken. But that also often means a duplication of effort. On the one
hand I like this do it yourself approach, but on the other hand it can
be rather time consuming and diverging. I as a user have to manage
dozens if not hundreds of packages, and you can't expect me to
understand the internals of all of these. So I as a user need a place to
be able to take up my beef. And although many developers are very
responsive their lack of time often makes that they only half answer
your questions.
You have to manage dozens of packages. I have to manage several
thousands of bugs, and hundreds of people taking up their beef with me.
It just doesn't scale well.
I guess i just don't understand what you'd like developers to do. It
sounds like you just want them to do more work in less time.
> We do already spend time fixing bugs, and it is an essential
part of QA.
> However you seem to want us to spend more time on that than we currently
> do. The hard question is then: What would you want us then to not do
> instead?
You have vast amounts of possible volunteers on these lists, which if
managed well could take some of the burden of your shoulders. I can see
that individual developers are taking part in managing these resources,
but I don't see an overall strategy. But then maybe I am expecting too
much of a BSDish approach...
All the possible volunteers on this lists are great, but the fact is
that if I have to personally hand-hold them I won't have time for
anything else, which might be a net loss to the project. Its very much a
fact that volunteers in free software projects have to be very driven
self-starters. It might be an unfortunate fact, but its been true for as
long as I've done free software, and I don't see it changing.
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Alexander Larsson Red Hat, Inc
alexl(a)redhat.com alla(a)lysator.liu.se
He's a time-tossed coffee-fuelled librarian on his last day in the job. She's
a blind antique-collecting mechanic with her own daytime radio talk show. They
fight crime!