Sometimes programmers waste their time for implementing non-useful/non-important features, but sometimes such features aren't just not useful, but harmful.[1; you are more than welcome to convince me otherwise]
Both developers, designers and project managers should always ask themselves why do we need such feature before going to implement it and should be able to convince others on the motivation for it.
I think that sometimes we fail… (and sometimes even in the "postmortem" step[1]) The question is how can Wikimedia foundation improve the process?
For large changes there are already reviews and RFC on mediawiki wiki, but for smaller ones it is sometimes missed (and even phabricator ticket). Should any feature should be associated with phab task?[2]
Eranroz
-----
[1] https://phabricator.wikimedia.org/T100691https https://phabricator.wikimedia.org/T100691:// https://phabricator.wikimedia.org/T100691phabricator.wikimedia.org https://phabricator.wikimedia.org/T100691/T100691 https://phabricator.wikimedia.org/T100691 / https://gerrit.wikimedia.org/r/#/c/95723/ defeature __NOEDITSECTION__ from VE
[2] https://www.mediawiki.org/wiki/VisualEditor/2015_Process_Review#Recommendati... .