Fostering developers (was Re: Feature Request)

[Jouni Heikniemi]

My non-organized thought flow worth less than 2 euro cents:


Shane H. W. Travis wrote:

>>Able reviewers who can take the time to sit down and stare at patches for
>>hours are rare, and it's been hard nurturing more to come along.
> Do you (does anyone) have any thoughts on *why* it has been hard? Is it a
> lack of volunteers, or the internal culture of the existing group, or an
> absence of guidelines to facilitate the process, or the attitude of one or
> more individuals, or a lack of time/desire?

Personally, I find reviewing a very fun way to contribute to a project. 
In a sense, it gives you much better view of the project in whole. 
However, the problems are clear: Reviewing is very time-consuming, and 
the meticulous approach it takes is rare to find in people (at least in 
their spare time). Bugzilla's codebase is an issue as mentioned, as is 
the fact that it's somewhat difficult to stay up to date with the 
thoughts of the core developers unless you're active on IRC during 
American evening hours (often fairly unfeasible for Europeans).

I still believe the key problem is the rarity of the people who actually 
want to delve deep enough to the code to be able to review properly. 
Bugzilla seems to emit an interesting characteristic: While the quality 
in some parts of the codebase is far from superior, the requirements for 
new code checkins are pretty harsh. This probably makes Bugzilla a bit 
less attractive than many other OS projects with more lax standards for 
participating.

For me, personally, lack of time is an issue as it's probably for 
everybody else on the current reviewer list. That said, my lack of time 
usually makes me do something else than search for things to do - but if 
requests land on my plate, I'm likely to get rid of them in a timely 
manner.


I believe better organization of the development community would help. 
Dave, while being intimately familiar with the project and a leader with 
lots of vision for the project as a whole, has little time for practical 
every day management / driving of the patches, even the minor ones. This 
is the effort Bugzilla could use, but candidates for such a position are 
really few - many realistic alternatives don't spring to mind atm.

Our formal documentation lays this responsibility onto the shoulders of 
the whole reviewer team, but it's becoming obvious none of us have the 
combination of authority and time to do it. Here's one tidbit of 
experience: When I had just joined the reviewing team 2,5 years ago, I 
had a lot of enthusiasm for all sorts of things. Back then, I _made_ 
time (I really didn't have it available even then, but the motivation 
was great) for reviews and helping people.

However, at that point I also often lacked the authority and the 
knowledge to really help people through the hard decisions. Deferring to 
the senior team was costly in terms of time - and this problem has grown 
worse ever since. It's this particular scenario I believe to be the 
chief culprit for making Bugzilla development harder than it perhaps 
should be.


Jouni