Open mjeveritt opened 4 years ago
I hadn't thought of this, but there's no reason not to!
Something like, as a worker itself (we only need 1 worker here really):
Ah is that how it's done .. sounds good.
Nattka also has a [better?] commit script for this function than tatt.
The simplest way seems to be:
First approximation here!
From IRC (#gentoo-at-sam @ Freenode):
[3:29 pm] <kveremitz> sam_: thinking maybe it would be helpful to have a Good & Bad list of bugs .. the 'good' list can be auto-parsed, but the Bad list can be worked through for issues
[3:29 pm] <kveremitz> so possibly subtly different output
Had another thought, if we use a pickle file for the 'good' bugs, we can directly call nattka functions from python. In fact, we might not even need a file .. but that might fit into the 'future development' basket. Or it would be a good way to test/decouple testing from 'passing' vs. 'failing' functions in separate scripts, before integrating...
Useful bookmarks:
We could implement a queue on the output side for 'good' packages, that picked up on the tatt 'commit.sh' script? And you ran a similar worker to actually perform the final 'ack' which spun up repoman so you could commit the finished work....