Chris Schmidt (crschmidt) wrote,
Chris Schmidt
crschmidt

Reviewing in Zilla

Lately, xb95 keeps saying "Can someone review this for me?" Every time he does, I feel a bit more frustrated, because it feels to me like he's requesting that people do something that they shouldn't be forced to do, that they review patches when they don't feel comfortable. In fact, he YELLED (talked in caps) and swore in IRC about the fact that instead of reviewing perl code, people discussed (okay, I'll admit - argued) about the changes he wanted to make. (Note that I'm using the "person" username for him and not the employee username. This is deliberate.)

So, I felt like he was picking on all devs with this comment - "you guys don't review enough!" "Why doesn't someone review this?" is the feeling I got about it.

This morning, I realized that this might be because I feel like I'm the only volunteer in zilla who reviews.

To figure out whether or not this is actually the case, I went through and did a little bit of digging. I went through the most recent 50 bugs closed with "reviewed" keywords. (Raw data at http://lj.crschmidt.net/reviews.txt.) Some stats:

Patches in the last 50 "reviewed" patches touched by me (either patch written by me, or reviewed by me): 25
Patches reviewed by me: 19
Patches reviewed by all other volunteers combined: 10
Patches reviewed by employee: 25 (about half jesse, half david.)

It seems that I review more than david, more than jesse, and more than all other volunteers combined.

I'll admit that these statistics aren't that great: I compiled them by hand, anything that's reviewed by an employee typically doesn't get marked with the "reviewed" keyword, just commited, etc. But it seems to be a rather accurate representation of the number of patches that I review compared to other volunteers, even if it's not compared to other employees (which isn't my main concern anyway - if you're working 40 hours a week on LJ, i should hope you review more patches than me).

So, I'd say that my angst at that particular complaint is largely my own. because I feel like I'm the only person in Zilla who reviews, when Mark asks for reviews, I read it as a request for me to review something, which then puts me on the spot, which is bad in times like now, when I'm busy as hell with exams and abuse and tens of other things that I'm forgetting about.

Of course, I don't know the solution to this. The reason that people don't review in Zilla is their own, but a lot of it is obviously due to lack of environment. There's never BEEN a development environment since I started developing, so I'm not sure if that's actually the cause or not. People insulting volunteers doens't help the cause, obviously, since that hardly makes me want to develop more. Nor is this an isolated incident: there's examples of it in tons of bugs that are out there.

http://zilla.livejournal.org/show_bug.cgi?id=1230#c1
User gets yelled at for adding CC list per request of support manager.

http://zilla.livejournal.org/show_bug.cgi?id=550#c10
Employee yells at me for not testing a patch. (I still maintain I tested that patch, I'm not sure how it failed so miserably)

http://zilla.livejournal.org/show_bug.cgi?id=1257#c13
I get yelled at for doing what I was told to do on a bug. (Since the bug is not regularly reproducable, there's no way to take it and test it thoroughly in the situation - it's a "may fix it, may not", which is what I was told the "patch only" keyword was for on IRC.)

http://zilla.livejournal.org/show_bug.cgi?id=1544#c12
Brad gets upset over a patch that goes in, and says "remove this guy from CVS", also points out to whitaker code that is needed that (I suppose?) he was supposed to finish and hadn't.

Now, all these are typical developer things that wouldn't be a big deal in a close knit developer community. If you're friends with the people you're working with, yelling at them when they screw up occasionally is no big deal, because everyone plays nice in the end. However, with a community as large as LiveJournal, you have a large disjoint between paid developers and volunteer developers. Lack of communication creates this to a certain extent, but it also just comes from the fact that you're working with people with many VERY different points of view on how things do and should work.

So, I see why people don't review. I also see why I feel like I do all the non-employee reviewing. So, what's the solution out there? Any one of you seen this kind of problem before, where lots of people code patches but nobody tests them? Anyone have any possible suggestions for how to change that behavior? Any suggestions as to what to do to encourage more and better testing from volunteers?

I'm kind of at the end of my rope with what to do here. If no one else reviews, I feel bad when I don't do it, but if I get upset, then I'm less likely to do reviews, etc. etc.
Subscribe
  • Post a new comment

    Error

    Anonymous comments are disabled in this journal

    default userpic

    Your reply will be screened

    Your IP address will be recorded 

  • 6 comments