Re: How to volunteer for something on debian's todo list?
Thanks for the detailed replies. Not much to add, really -- just a few
random comments below. To keep things concise, I munged the responses.
On Tue, Aug 10, 2004 at 07:52:26PM -0500, Peter Samuelson wrote:
>
> Anyway, I don't know if you got the Debian source package (apt-get
> source ifupdown) or just the upstream source - the Debian source
> includes the generated .c files etc.
Actually, I did fetch the upstream source. However, to really do
something beyond just looking at the source, I'm afraid I won't get
around noweb, anyway.
>
> > Anyway, before I take the trouble to install noweb,
>
> Well, that step should be trivial: "apt-get build-dep ifupdown".
I guess I meant "...install _and figure out how to use_ noweb" :)
>
> Oh, no. You don't need to be in NM to hack on source code or do
> packaging work, either one. Your first step should be to check
> http://bugs.debian.org/ifupdown to see if there are any open bugs on
> this subject.
>
> I checked it and it seems there are not - while 'ifstate' seems to be a
> source of a few other bugs, it's not considered a bug itself.
> (...)
>
> There is no formal process for this TODO list. It's mainly just
> informative. Since this item is confined to a single package,
> ifupdown, it should be enough to just coordinate with the maintainers
> of that package. Most developers should not care at all whether the
> person doing the work is a developer or not.
sounds promising so far...
> Just explain to <ifupdown@packages.qa.debian.org> that you are
> interested in looking into this, and that you'd like whatever
> information about it anyone can provide. You don't need to make any
> commitment to actually *doing* the work. As you hack on this, you
> can make status reports to that same address - or just keep your
> mouth shut until you have finished, if you prefer.
...will do, if I should choose to go for it at all. I guess I'll first
take a closer look at the code, at noweb, at outstanding bugs, and
whatever else... (sounds very much like my "work it all out completely
before..." strategy, doesn't it? :)
On Tue, Aug 10, 2004 at 06:06:54PM -0700, Steve Langasek wrote:
>
> No, but your grabbing it doesn't say anything one way or the other about
> whether other people will also be working on it.
> (...) The risk that you'll step on anyone else's feet by working on
> this problem is actually quite low. ;)
Sure. All I was thinking of was that it might be a good idea to
coordinate efforts -- well, presuming there actually is anything to
coordinate besides the processes in my own head ;)
>
> That I didn't remember this particular todo list even existed, so I
> don't know it would be something I'd recommend it as a starting point
> for getting involved in Debian.
May I ask what you would recommend? In addition to the mentioned todo
list, I had also already taken a look at the pages linked off from
http://www.debian.org/devel/wnpp/, but with many of the packages there
it seemed to me there's good reason they're up for adoption, etc.
At least there wasn't much I'd consider particularly appealing for my
personal taste.
> You at least seem to have picked one of the less
> interesting/important tasks listed there, which accounts for why no
> one has tackled it yet. ;)
heh, want to discourage me!? :) Some people like to play a game of
chess or whatever to keep their brain cells active. I occasionally like
to tackle a little programming riddle or debug something for exactly
the same purpose... That's why I was originally looking for something
cute and small, not being too interwoven with a whole bunch of other
packages and projects. Yet challenging enough to provide incentive
beyond simply "doing something useful".
Actually, I kind of prefer tasks that can be completed within a
reasonable period of time, to then be able to get them out of my head
and time schedule. I'm generally averse to projects that require
continued long-term dedication. This makes me think I probably
wouldn't be a good maintainer, BTW. Although I sometimes wish I was
different, more than 30 years of firsthand practical experience with
my self have convinced me that I'm generally rather unreliable when it
comes to long-term commitments ;( (hey, there are so many interesting
things to do, and life's too short to be doing the same thing all the
time :)
Okay, I'll better just stop, before this is gonna drift off entirely
into off-topicness...
On Wed, Aug 11, 2004 at 10:52:35AM +1000, Matthew Palmer wrote:
>
> Note that some of the items on that list are marked "Taken By". That'll be
> how others know someone's working on it. You'll also note at the bottom of
> that page that it says "To add or remove items in this list, or to inform
> you have taken over a task, please contact Josselin Mouette".
Thanks for pointing me to the latter info. I must confess I completely
overlooked it -- apparently, the position led me to dismiss it as some
copyright boilerplate.
>
> Half the reason this one's got three stars is probably the political issues
> involved. There was a large flamefest on d-devel some time ago about
> ifupdown and whether aj was going to do anything with it and whether he was
> going to let someone else take it over and blah blah blah. As a
> longstanding and fairly deeply ingrained developer, Anthony Towns is one of
> the "demigods" of Debian -- you don't go hijacking his packages. I think
> the ifupdown SF project was a fork by rebel elements, and so it goes...
>
> But if you go through the PTS and BTS for ifupdown, you'll notice that the
> last maintainer upload was made over 3 years ago, and there's a very, very
> large pile of bugs on it. That, by itself, is enough to scare most people
> away.
>
> Your mission, should you choose to accept it, is mostly going to be getting
> the patch integrated into the package. It's not the sort of thing you do in
> an NMU, but there effectively hasn't been a maintainer to do an upload for
> several years.
>
> For all that, the patch is definitely worth writing, if only for the pyhrric
> victory of having done it. And you never know, someone may get the courage
> up to wrestle the package from AJ's unmoving fingers and integrate all those
> NMUs and outstanding patches...
Thanks for sharing those background infos. It definitely gives me a
clearer picture of what I'd be in for. Unfortunately, I'm not a big
fan of "political issues" -- too bad about all the time and energy
wasted.
For a moment I was positive I'd found a nice little task for a rainy
weekend, but with somnambulistic certainty I seem to have picked a
rather tricky one -- as usual ;)
Looks like there might be a few hills to climb... Well, before I
decide on whether I'm going go for it, I'm going to climb a few real
mountains in the French Alpes (me is going on vacation in a few days :)
Thanks for all the help!
Almut
Reply to: