On Wednesday 21 March 2007, Simon Budig wrote:
> Shlomi Fish ([EMAIL PROTECTED]) wrote:
> > Actually I also think it was too rude. Let's analyze it:
> > 1. You didn't say "hi".
> Oh, come on. A lot of the mails on mailinglists don't include a greeting
> if quoting something. I've never perceived that as *rude*.
It isn't normally, but it was in the context of that laconic email.
> > 2. You phrased it as a question that implied the original poster should
> > have thought abuot it himself, instead of giving an answer.
> Oh, come on again. It is not unusual to formulate a suggestion in terms
> of a question. To pose this as rude is IMHO blown way out of proportion.
Yes, but by phrasing it as a question, the OP was essentially accused of not
thinking about it before.
> I really don't get why this discussion sparked from this specific mail.
> Please also keep in mind that most of us are *not* native speakers and
> discussing stuff in this linguistic nuances is probably a wasted effort.
As said, it is a general trend we noticed here. This email was just an
instance of it and we found it necessary to discuss it.
BTW, I believe there had been an improvement here, but it is still not enough.
> > However, I believe Alpár's and Joao commentary was induced by the general
> > trend of treating people on this list (and potential future contributors)
> > rudely or impatiently. I've noticed this general trend here too after
> > someone made me more aware of it.
> Funny, my impression is, that it got better in the last few months.
> > I believe the GIMP could have been much better off today, if it weren't
> > for all the antagonism that the developers' have created. I mean, sure
> > after Spencer Kimball and Peter Mattis left the project and most of the
> > other original developers left to work on GNOME, very few developers were
> > left. But since then the community of FOSS developers grew by leaps and
> > bounds, and there shouldn't have been any problem finding much more
> > potential developers than we have today. There are plenty of fish in the
> > sea.
> I am concerned that this might not be exactly true. I don't have
> specific numbers, but my impression is, that it is way harder to sort
> the wheat from the chaff nowadays. In the beginning of the GIMP the
> people who a) found the GIMP, b) bothered to subscribe to the
> mailinglist, c) expressed interest in contributing generally already
> were enthusiasts about free software. They used stuff like e.g.
> slackware to convince their boxes to boot into linux. They were editing
> textfiles to get an internet connection. They had a thorough
> understanding of their system. This is no longer the case today. I mean,
> I am teaching stuff at the CS department of our universities and there
> really are students here who never did any programming and/or are
> annoyed if they have to. I mean, WTF?
> Granted, the pool became bigger, but the amount of fish in it did not
> increase in the same proportion. And there are a lot more projects
> fishing around in there.
> I am not sure if this is a fair description, I probably sound a little
> bit like a grumpy old man. Not sure what to do about this though... :)
Well, you're somewhat right. But other projects do not seem to have a problem
finding developers. Inkscape has them. Subversion has many of them. While
the "hackerdom" level of the open source people has decreased somewhat, we
still have many potential B.Sc./B.A/B.E. graduates who contribute to FOSS who
can program in C on Unix. Or there are many enthusiastic high school or
undergraduate students who learn a free Unix clone on their own, and can do
enough programming. So not all are bad.
Many projects are propsepering because they can find many developers. Granted,
a lot of them are easier to program for than the GIMP, but most of them are
And I think even they had it pretty easy. I started programming and computing
in 1987 with the PC XT ROM BASIC and DOS, when I was 10 years old. I think
modern kids who start out today with Linux, still haven't suffered enough.
However, I talked with a woman on the IRC, who started programming with IBM
1401 ( http://en.wikipedia.org/wiki/IBM_1401 ) assembly, and has used UNIX
since AT&T UNIX Revision 6. (When it still printed on paper.) Now this is
> > Let's look at Inkscape for a counter example. They have been around
> > for a much shorter time than the GIMP, but have made remarkable
> > progress, because the atomosphere they have is much better. If they
> > could do it, why can't we? Only because we reject potential
> > developers.
> I believe that "reject" is the wrong word here, because it implies that
> we'd do this on purpose which is in general not true. We do have a habit
> of expecting a lot from patches (look at the number of iterations some
> patches go in bugzilla)
That's not exclusive to the GIMP. I sent some patches to Subversion, too and
it took me several iterations. They were rejected for such trifles as
incorrect style, or trailing whitespace. I had to write a detailed commit
log, and write the system tests in Python (where the initial test cases
posted to the bug tracker were in Bash).
When patching the core perl5 documentation, I had to remove trailing
whitespace, correct typos or punctuation, and fix phrasing according to the
input of the more experienced developers. I also had to handle some "This
patch sucks! The documentation is good enough." attacks.
> but in general I believe that even the original
> patch author will agree that the result is way better than what was
> originally submitted. We also have the habit of expecting more than
> half-baked ideas when someone comes up with an idea and sometimes ideas
> are probed a lot before a developer admits that it might be a good idea.
It is good to have several such iterations if the patch is not good enough.
However, we should do it in a friendly and supporting manner.
> I guess that these intentions do not always make it through to the
> person presenting the idea and they take it as a hostile attitude of the
> gimp developers (sometimes even "threatening" us with switching to other
> applications). This is the area where we probably can improve things
> a lot by being more careful with the language. However, the probing in
> itself is necessary and important for the quality of the Gimp.
> [and regarding inkscape - I like the project a lot and I don't mean them
> any harm, but they lost me as a contributor even before I tried to
> communicate with them - 430 source files in a single directory is a good
> way to do this, hopefully they will improve there.]
Inkscape also lost me as a contributor (at least temporarily) because when
trying to fix a problem I encountered in the bugzilla, I could not understand
where in the code it happened. Then, after a while, a different developer
(probably more experienced) wrote a patch and applied for it himself.
> About your suggestions:
> > 1. Create a page with some FAQs and canned responses in both HTML and
> > plain text formats. I volunteer to prepare and maintain this page. I also
> > get sick of reading more "You should change the name of the GIMP, it is
> > an insult in English", "When is CMYK/16-bit/whatever support coming?",
> > etc. questions, but I believe we can at least answer them politely.
> Please add "I will switch to [whatever]" to that list...
> > 2. Have a system of self-moderation. Let the messages of developers with
> > a tendency to be rude and untactful pass through a small forward
> > of "friendliness experts" for approval and correction. IF the experts are
> > unhappy, they'll tell the senders to edit them and re-send them.
> > Note that I'm not necessarily suggesting we force it down the developers'
> > throat, although it may not be a bad idea either.
> I think it is a bad idea. We do have to be careful that Gimp stays fun
> for ourselves. Putting ourselves in a self-surveillance-corset is a
> great way to kill the fun IMHO.
OK. I suppose we can send such notices, in private mail, after such an
insulting message is sent, to have the poster of it be more aware of it, and
to pay more attention next time.
> > 3.
> [snipped - that was IMO not necessary]
> > 4. Otherwise have the developers pay attention and try to be as
> > hopsitable, friendly and tactful as possible. Reading the Producing OSS
> > book would be a good start.
> I might give it a shot.
Shlomi Fish [EMAIL PROTECTED]
Chuck Norris wrote a complete Perl 6 implementation in a day but then
destroyed all evidence with his bare hands, so no one will know his secrets.
Gimp-developer mailing list