pulkit added inline comments. INLINE COMMENTS
> martinvonz wrote in narrowcommands.py:149 > Also need to check old capability here Only need to check the old capability here. Will update. > martinvonz wrote in narrowcommands.py:282 > Should keep this in the "old" case We don't need this in old cases. This is only required when widening non-ellipses cases as that will end us in a situation with an empty changegroup. > martinvonz wrote in narrowcommands.py:303-305 > Is the `known` set needed when not using ellipses? Conversely, `commonheads` > shouldn't be needed when using ellipses, but perhaps it's still useful to > have it there (it's usually way smaller, so it's much less of a concern). > These things can be fixed in a separate patch, of course. No, the known set is not needed in non-ellipses cases. We should not compute that also as in non-ellipses repos, computing known just adding all local commits which can be in millions. The server logic completely rely on common in non-ellipses cases. REPOSITORY rHG Mercurial REVISION DETAIL https://phab.mercurial-scm.org/D6436 To: pulkit, durin42, martinvonz, #hg-reviewers Cc: rdamazio, mercurial-devel _______________________________________________ Mercurial-devel mailing list Mercurial-devel@mercurial-scm.org https://www.mercurial-scm.org/mailman/listinfo/mercurial-devel
