On Mon, Nov 14, 2011 at 1:58 PM, Jay Ashworth jra@baylink.com wrote:
On Mon, Nov 14, 2011 at 11:51 AM, Jay Ashworth jra@baylink.com wrote:
And sometimes, that means assigning general responsiblity about that issue to one specific person. Do we have that person?
Depends on the issue. Are you asking about a particular one from this thread? (i18n/l10n issues should probably start by being assigned to Siebrand Mazeland if there's no specific candidate you know of offhand.)
No, I meant more "one specific person whose situational awareness scope specifically includes the collective status of tickets, software or operations" (probably one of each, depending on the size of the problem).
As a general escalation path one should start with the bugmeister (that's hexmode / Mark Hershberger), whose job it is to prioritize issues and distribute them to folks who can work on them. Since figuring out just which part of what goes where sometimes is much easier with more historical context, old-timers like me can also help direct these to the right people (or be the right person).
Don't be afraid of being unsure who to send it to at first; we're much more likely to *get* it to the right person if the communication channels are open, even if our first response is a "what the heck is this????". :)
An example of a site JS issue; just today I wandered across < https://bugzilla.wikimedia.org/show_bug.cgi?id=32366%3E which on inspection turned out to be a continuance of an earlier incompatibility between site JS on als.wikipedia.org and the newer, jQuery-oriented MediaWiki front-end code https://bugzilla.wikimedia.org/show_bug.cgi?id=29595.
One developer had made an attempt at helping out and fixing the site JS back in July, but it turned out to be an incomplete fix and caused additional scripts to break, so got reverted by the locals (as the things it broke were more important than the things it fixed).
That thread of conversation unfortunately dropped off, and the rest of the fixes didn't get completed in July...
The followup bug in November got a slightly snippy initial reply from the first person to check it out, but since it came back on the radar I was able to pull up the earlier fixes and figure out what was wrong (additional scripts didn't get updated to match the new MediaWiki:Common.js) and fix them -- it seems to now be resolved.
So that shows us doing some things right (investigating and helping out with the site JS problems when a problem shows up, especially in concert with changes to MediaWiki itself such as jQuery coming in) and some things wrong (an in-progress fix & test cycle got disrupted and wasn't continued for a few months; some sub-ideal responses).
-- brion