[Adium-devl] Some disclosure and some ideas.

[Augie Fackler]

On Dec 7, 2006, at 12:15 PM, Colin Barrett wrote:

> <snip colin is making big $$$ and should buy me a Wii />
>
> Anyway, on to the meat of the email, which is mostly just a bunch of
> lists. Firstly: Adium development has slowed down a bit over the past
> couple of months, with school and what not, but things tend to pick up
> during the holidays, so I just wanted to throw out some things I think
> we should be getting done in the next month or two:
>
> - Finally release 1.0.
Amen - I think about half of the remaining blockers are things I  
should try to handle, so we'll see how that goes in between all my  
thesis work over break. Just as a heads up to the team, I've got a  
week in the middle of break where I'll be mostly on dialup and  
therefore hard to reach except via email, and then I'll also be  
spending a couple of weeks worth of time doing research for my honors  
thesis.

> - Finish our plugin architecture. I think we are most of the way to
> getting something we can support at least some initial development on.

I'd like to see this too - I'd really like to see some nice way for  
us to support 3rd party protocol plugins that we wouldn't need  to  
worry about maintaining.

>
> In the more long term, I think we should try to get some of these
> things done, for the health of the project:
>
> - Codify our procedures that are in place, and document more of our
> community structure on our wiki. The PatchMaster idea was a good one,
> but I think it should be retooled to reflect reality, where we have
> reviewers who check things in. For example, I think we should make
> sure that anything major that gets checked in to a release branch or
> trunk gets looked at by at least one set of eyes. We don't need to
> institute a formal code review process (yet), but what we've been
> doing with branches seems to work well, and I think that definitely
> should be written down and codified somewhere.

I'd be game for being a patchmaster - I found this summer that one of  
my better skills seems to be reviewing code.

> - After doing the above, we should look at our practices and see what
> we can do better, and how we can keep things involved. I've noticed
> that the Mozilla projects tend to have weekly meetings on IRC, with an
> agenda on the wiki people can add things to. Perhaps we ought to give
> that a shot? There are a bunch of other things we can try, too.

Weekly meetings might be nice, but I suspect they're somewhat  
infeasible given the fact that we all have very busy lives outside of  
Adium.

> - Finish the unit testing work. This is so important, it will allow us
> to track regressions and keep us from doing the two steps forward, one
> step back dance.

If someone could give me a better rundown on exactly how unit testing  
works, I'd love that. I keep hearing wonderful raves (Shipley aside)  
about how unit testing is the shit and we should all unit test the  
daylights out of all our code, but I'm still a bit confused as to how  
unit testing actually *works*, and no site I've found has an  
explanation of the how, just the why.

> - Perhaps investigate a different SCM system. Subversion is fantastic,
> but I think we have already started to see some limitations of it's
> merge and branch support. That said, svn 1.5 has improvements in this
> area, so we'll see. That said, it wouldn't hurt to do a test import
> into various other SCMs (darcs? Mercurial? git? Bazaar?), if anyone is
> interested. Perhaps we can recruit someone on IRC or something.

I'm very opposed to using only some kind of distributed system, and  
honestly, of the systems I've been able to examine, subversion is the  
cleanest and easiest to use system I've seen. When they get offline  
commit support added in, I'll be happy as a clam.

>
> Some thoughts on remaining relevant as a product:
>
> - Jabber. Turning Adium into a top notch Jabber client is how we're
> gonna remain relevant in the future, IMO, and we're doing a lot
> already, and I think we should continue on this.

*nod* but I'd like to discontinue the trend of making  
$J_RANDOM_JABBER account types (see LiveJournal account types.  
Honestly, I'm starting to wonder about the legitimacy of having a  
separate GTalk account type. Maybe creating a Jabber account could  
support having a pop up menu of already-known services?)

> - MSN. Our MSN support sucks, and it's our most popular protocol.
> There isn't a whole lot we can do to libgaim, but we should definitely
> keep our eyes peeled.

MSN in general sucks - I'm not entirely convinced we'll be able to  
inflate[1] it.

> - Events. I think one way to compete with iChat 4 (which has many of
> our features), is to add much more robust events system with a great
> UI (seee Augie's email thread). Since Leopard is coming out in the
> spring, I really think this is something we should definitely think
> about spending a lot of time on in the coming months.
> - Quicksilver. Adium and QS seem to attract a lot of the same users.
> Eventually I'll get around to writing one, but I encourage someone to
> either continually poke me or to beat me to it :)
> - Audio/Video. We're making (slow) progress in this area, which is
> good (and not so good). We're eventually going to need it to compete,
> there's just no question. If anyone has any ideas about how to speed
> this up ($$$?), please, speak!



> - Committers/Community. We've got a great community of triage and
> forum users set up, which is awesome. Now we need to start building up
> our developer community. The Map of Adium Peter has been working on
> (and I've been neglecting — sorry!) is a great start, and I think it's
> something we should focus on. Our numbers are only going to dwindle,
> not grow.

Once 1.0 is out the door, I plan on working on some kind of  
doxygenation project. At that time, I plan on taking on the title of  
doxynazi.

>
> Some updates on things I work on:
> - Shirts have stalled for the moment while I think about
> "distribution" (the term used by companies in that business is
> fulfillment). I may just end up shipping things out of my place in
> California after I move in January. We aren't going to be selling THAT
> many units this time around, at least I don't think.
> - Scrollbar improvements branch: Haven't worked on it in nearly a
> month. I'm fairly sure I can get everything working though, I just
> need to sit down and *do* it.
>
> I think that's most of what I wanted to get down. I'm a bit jetlagged,
> so please excuse any incoherentness.
>
> I've been pretty much off the radar for about a month. What's been
> happening?
>
> Long windedly,
> -Colin
>
>
> _______________________________________________
> Adium-devl mailing list
> Adium-devl at adiumx.com
> http://adiumx.com/mailman/listinfo/adium-devl_adiumx.com
[1] This seemed enough like an opposite to suck for me...