- From: James Graham <james@hoppipolla.co.uk>
- Date: Thu, 21 Nov 2013 10:45:20 +0000
- To: public-test-infra@w3.org
On 21/11/13 09:42, Dominique Hazael-Massieux wrote: > * I wonder how much of the stylistics checks could be automated or > partially automated; is there anywhere one could find a list of frequent > such errors that have been raised in previous test reviews? "Some". I started looking at this during TPAC since several people suggested it. It's obviously possible to automate checks like "does this file contain trailing whitespace", but obviously impossible to automate checks like "is this test actually doing something useful". For the former kind of thing, actually doing the checking is rather easy. What I'm not sure about is how it should integrate with the code review e.g should it try to post one comment per issue (would be super-annoying if there is a 1000 line patch with trailing whitespace everywhere), or one comment overall, or something else? Should it post to github (easy) or to critic (harder, but more useful when it is being used, since one can track changes).
Received on Thursday, 21 November 2013 10:45:50 UTC