[DBIx-Class-Devel] Let's review the review process
Peter Rabbitson
rabbit+dbic at rabbit.us
Wed Nov 21 11:25:14 GMT 2012
I figured this is something that needs collective brainstorming, as it
affects commit velocity in general.
The situation (just a timeline):
* Abraxxa added a branch with some POD fixes - in essence just a rewrap
of existing text
* I noticed it in the commit log[1] which I frequently peruse, and
contacted him pointing out that the *original* text that he rewrapped
was entirely wrong (this happened on IRC)
* He did an iteration which I looked at an hour later, which addressed
oly half of what I pointed out.
* Anther iteration took place a day later, which I looked at but didn't
get a chance to comment on the fact it is still incomplete. Then things
fell through for a week.
The end result of this is: Alex' work is sitting seemingly warnocked,
while I sort-of "took" the review baton for this particular branch. From
my POV the most effective way to point out what is wrong with the text
as-is is to actually submit a correct version of the text, which however
will negate the original contribution. I don't want this to happen.
On the other hand if I do not say anything and just let the commit go
through, my OCD will compel me to fix it in earnest (as in [2]), again
negating the contribution.
What is even ore disturbing is that this takes place for a silly
doc-patch. If anyone has thoughts on what do I/we do so we stop ending
up in this situation - do share!
[1] http://git.shadowcat.co.uk/gitweb/gitweb.cgi?o=age
[2] https://www.xkcd.com/386/
More information about the DBIx-Class-Devel
mailing list