[19:03:13] <AaronSchulz>	 TimStarling: https://gerrit.wikimedia.org/r/c/mediawiki/core/+/841392/2 (part of restoring connection pool patch)
[19:05:27] <AaronSchulz>	 Krinkle: still a few patches based on https://gerrit.wikimedia.org/r/c/mediawiki/core/+/784756
[19:08:52] <Krinkle>	 AaronSchulz: Regarding the DBConRef refactor, with the destructor removed (as per Tim, identifying it a no-op) does this mean that effectively any connection once opened will remain opened and involved in the lbfactory-glboal commit/waitFor/etc rounds? It seems like previously these would sometimes end up ignored or closed, but now remain active. Probalby not an issue, if anything perhaps for long-running scripts but evne in prod we 
[19:08:53] <Krinkle>	 don't have all that many different sections, but just wondering if I observed it correct and that it is intentional. 
[19:12:03] <AaronSchulz>	 Krinkle: more or less
[22:38:27] <TimStarling>	 reuseConnection() never closed the connection anyway, it just returned it to a pool for future use