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

Re: RH and GNOME



Scott McDermott <vaxerdec@frontiernet.net> wrote:
> Ok, then perhaps I am sheltered.  My opinion is that if they want to
> defect, let them, they are kruft.

NO!

They are not kruft.

They have made a decision, based on their own value system, that they're
going to do something else for a while. You don't have to despise them
for doing that.

And they still might have something worthwhile to say (or maybe not).

There are some serious problems with debian: the biggest is that
we don't tell enough people about it, or what it can do for them.
[For that matter, we haven't done much research into what people need
out of an os distribution.]

I think we can solve these problems while improving what's good about us,
but I don't think it's necessary to create a big gap between us and
Red Hat to do so.

> If you can't figure out how to do it then use a distribution that
> caters to you.  It's not exactly hard to learn of alien doing a simple
> web search or reading through package descriptions.  Just how much of a
> handhold would you suggest?

When I'm working on a project, writing the code and the documentation,
I sometimes find that it's easier to change the code than it is to
document the current behavior.

If we can't write better documentation, I think we ought to write better
code... [and dpkg, unfortunately, has had neither for quite a while --
actually that's not fair: things like the policy manual and the packaging
manual count.  But when I read its code I realize that I'm still missing
some big clues about how it functions.]  It's the combination of code +
documentation that we should be improving.

> it's the hacker distribution, put together by hackers, for hackers, in a
> way to encourage proficient hacking.  So really I don't see anything
> contradictory with not caring about newbies.

I don't think it's necessary to not care about newbies.  But we do need
to make sure we don't trash the system trying to adapt to them.

Another group to think about is the group in transition from newbie-hood
to hacker-excellence.  For these folks (and for hackers who have
specialized) it's important to give them a road-map to the system,
letting them know where the documentation is, what's [rougly] next most
important to read, etc.

> It isn't already? Each little bit of "easier" (divorce from UNIX
> paradigm) involved a geometric increase in the effort involved to bring
> it to that point.

Sometimes this is true, sometimes it isn't: sometimes it's simpler to
make things easier, and you don't lose anything in the process.  Then
again, we're often pretty eager to make these kinds of improvements.

> Debian's "course" has inertia.  Deviations require increasing amounts of
> energy.

That is too true.  And, right now, I don't like the trend where dpkg
is mostly being ignored, people are doing "quick-expedient fixes" on it
and more serious bugs are not getting fixed.  Staying on this course is
asking for trouble.

-- 
Raul


--  
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org


Reply to: