GNOME.org
 

Re: [gnome-love] Revitalising gnome-love ... some history

First of all, glad to help start up some activity ;) I would love to contribute to GNOME some how, but I have no clue where to start. I really would like to contribute some code, but I guess I will need to teach myself GTK+, etc before I am knowledgeble enough to write anything useful. I don't know that I really have any other special skills. I know a little C and that's about it as of now. I guess for now, I will be busy reading all the little tutorials and other docs that are available on developer.gnome.org 

Aaron


Malcolm Tredinnick wrote:

[Warning: very long, possibly tedious.... If you already know the
history of this list, do not read. There is nothing that is both new and
correct in here.]

I had a real feeling of deja-vu yesterday after Aaron Walker's post,
because I have been thinking a lot lately about whether this list is
really worth keeping. I had sort of convinced myself that is was, since
a quiet list causes no harm and can periodically be useful.

Now there seems to be a few people who have subscribed recently and are
perhaps wondering why they bothered. So what follows is the peanut
gallery version of gnome-love's history and what we can do moving
forwards. All of this is from my recollections, notes and a quick rustle
through the mailing list archives. I was not present when the list was
formed, although I did get involved fairly quickly afterwards.
Conseuqently, some of the following may be slightly incorrect, but not
wilfully so...

Back at the GUADEC conference in 2001 (Copenhagen, Denmark) some of the
discussions centered around how to get more people involved in making
direct contributions to GNOME.

To put things into perspective, this was at a time when we were
nominally working towards GNOME 2.0, but there was little in the way of
concrete code to show for the wishes. People had plans about what needed
to be done and others were sort of spinning their wheels waiting for
GTK+ 2.0 to be released and the new versions of gnome-libs. In other
areas, band-aids were being applied to GNOME 1.4 to try and keep it up
to scratch. In all, it was a very frustrating time for many people, I
suspect (certainly was for me).

So, true to form, Miguel de Icaza, Chema Celorio and Federico Mena
Quintero -- all Ximian employees and key GNOME motivators (amongst many
other achievements) -- decided to start a mailing list. This list
(*this* list -- gnome-love gnome org)  was aimed towards helping anybody
who wants to start contributing to GNOME itself. We sent general "how do
I do XX" style questions off to the more appropriate lists, general
development questions went to gnome-devel and we concentrated on helping
people find and solve small problems.

The mailing list membership has always consisted of a mixture of
experienced people and people looking for things to do. The purpose has
always been to encourage people to help and provide some mentoring if
required.

This was a solution looking for an audience. Very quickly a couple of
hundred people signed up to the list and it was a success. As I hinted
yesterday, I consider a large part of this success to be due to the
cheer-leading efforts that people like Miguel and Chema put in in the
early days. Go back and read this post from the archives:

        http://mail.gnome.org/archives/gnome-love/2001-May/msg00105.html

The original (and continuing!) purpose of this list is well summed in
this post that Miguel made:

        http://mail.gnome.org/archives/gnome-love/2001-May/msg00020.html

(note that the link to the TODO list is no longer valid in that mail,
for reasons I will get to later).

So, did this really work? Absolutely! If you read through the archives,
you can see a few names of people who are now contributing hours of time
every week improving GNOME. One of the real early success stories that I
think the gnome-love group can take some credit for is procman; if you
are running the system monitor applet on your panel, right click and
select "open System Monitor" (ignore, for the moment, the problem that
the applet is called "system monitor" and there is a menu option to open
the "system monitor"). That box of all your processes is procman. Kevin
Vandersloot wrote it. His original mail talking about it is here

        http://mail.gnome.org/archives/gnome-love/2001-June/msg00113.html

Kevin was already a programmer when he started to enter the GNOME world,
but he wasn't an experienced GNOME developer at the time. If you keep
reading the mail archives for June 2001, you can see the progress of
procman: a couple more discussions, getting a CVS account, making a
GNOME release, etc. These days, people don't give it a second thought
and you have to really think to remember what gtop looked like. I only
pick this example because it is so visible. There are a number of
similar cases, but where people worked on the supporting infrastructure,
which you do not see directly on your desktop.

The year 2001 was a very good year to be involved in GNOME. It was a
very difficult year and a lot of concentration and flexibility was
required as things changed and changed again. However, that sort of made
it easier to keep up people's enthusiasm -- we were all partly confused.
Nowadays things are different. Not worse, just different. Many people
are more comfortable with the GNOME platform and I suspect this often
gives new developers the feeling that they cannot contribute
succesfully, since the learning curve seems so steep and yet there are
all these people who have mastered it without trouble.

News flash: there are very few people who will claim they have
"mastered" the whole GNOME platform. There are some, for sure, but you
always get some people with brains the size of planets in projects like
this. Most of us have certain areas we are comfortable in, other areas
where we bluff our way through in times of trouble and other areas where
we run away screaming. And it is absolutely not necessary to have an
understanding of the whole picture before you can contribute. I thought
I understood GTK+ at one point, but I learnt probably five times as much
about it by debugging somebody else's application -- and it wasn't
because that person knew so much either. I had to work out why their
method was wrong and what a better to fix it was.

Over the past couple of years this list has slowly stagnated and my
guess is because of what I try to outline in the past couple of
paragraphs. The bar to newcomers contributing has been raised and we
have forgotten to maintain a ladder to help people get there.

If you read the Miguel email I posted earlier, you would see a reference
to a TODO list. You might be thinking "what a good idea, that is exactly
what is required". You would be right ... and wrong. The problem with a
TODO list like that is that it absolutely, definitely *must* be kept up
to date. There is nothing more frustrating than working on a problem
that turns out to have been fixed six months ago. As it transpired, we
(as a group) sucked at keeping the list up to date. It did not help that
the cgi scripts supporting it were broken for much of 2001, but with
sufficient motivation that hurdle could have jumped.

At the same time, we need to have something like a TODO list. When
somebody comes along who may have experience ranging from "strange
ability to crash applications" to "likes writing documentation" to
"experienced programmer looking for something to do in spare time", we
need to be able to point out low hanging fruit for them. Pointing
somebody in the direction of bugzilla and saying "go for it" is, in a
sense, professionally irresponsible. That is why my mail yesterday said
that it is not possible to provide the required inspiration without
planning here.

It is fairly easy to come up with ideas about how we can fix this.
Without even trying, here we go... recreate the TODO list, create a Wiki
to record projects, have a project of the week / month, have a
suggestion list of the week posted here regularly. I could probably go
on.

Somewhat harder is to actually *do* something. Some problems that need
to be overcome: Firstly, finding a person with the discipline and time
to regularly update the information. Ideally, we would have a group of
people so that between them something happens each week, but each
individual only needs to find time every three or four weeks.

Secondly, tracking down and estimating the difficulty of the tasks that
can be done. In the past (again, mostly back in the glory days of 2001),
big-time contributors like Michael Meeks, Jody Goldberg, Martin Baulig,
and Owen Taylor would periodically drop a note with things that might be
appropriate. Sometimes this still happens (Jody tries to persuade people
to do little Gnumeric jobs fairly regularly, for example). However, as a
rule, we have to hunt down the tasks ourselves. It is fairly difficult
for maintainers to try and track information from multiple sources. So
they will mostly focus on their application's mailing list and bugzilla.
*We* have to go to the other projects and ferret out the little tasks.
Again, this is not a single person job. And the person who produces a
list of candidate tasks may not feel they are able to prioritise them.
But, that is little fish in the big pool of things, since there are
others who can quickly whack a reasonably accurate 'trivial', 'easy',
'medium', or 'the rest of your life' label on tasks.

If we could solve these two problems reliably, even if just for a few
months, then we can probably help some more people get involved.

I am not offering solutions here; you may have noticed that already. I
am just laying out what has happened and why I think it may have stopped
happening. I would love to see gnome-love return to the mentoring style
of things that happened in the early days of this list. It was good fun.
So I am mostly posting this to throw open the doors for people to
discuss what could happen (or just to flame me). I cannot play the role
of Chema or Miguel and do the "ra, ra, have at 'em" posts -- I really
suck at those (I think you have to Mexican for that). But I have both a
carrot and a stick here somewhere that I can use if we get going with
high volume again.

Cheers,
Malcolm

_______________________________________________
gnome-love mailing list
gnome-love gnome org
http://mail.gnome.org/mailman/listinfo/gnome-love
[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]