Please take another look. On Mon, Jun 4, 2012 at 11:41 AM, Richard Smith <[email protected]>wrote:
> On Mon, Jun 4, 2012 at 10:51 AM, David Blaikie <[email protected]> wrote: > >> several threads on this list discussed that fixits on errors and > warnings > >> have to be very likely correct. "did you mean" fixits on the other hand > >> should go on a separate note instead. The attached patch adds this to > the > >> internals manual. > > > > > > Looks reasonable (I think you can take my opinion as approval in this > case - > > but I guess you sent it out for review looking for other feedback too, > > perhaps). If it interests you, it's probably also worth highlighting the > > "recover as written" requirement which isn't entirely clear/explicit in > > those docs (and the implication that has on fixits on warnings: that they > > cannot change semantics - they have to be the "suppression" fix at most > > (which means there are very few cases where fixits on warnings are valid > > (since if you're confident the warning is good then the chances are that > the > > suppression is not the right fix))) > > +1. I think that the statement in the patch that "[...] warnings > should only have fix-its that very likely match the user's intent" is > likely to mislead people if we don't also document the "recover as > written" requirement. >
fixit-docs.diff
Description: Binary data
_______________________________________________ cfe-commits mailing list [email protected] http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
