It seems we've deleted the only implementation - nsGlobalHistory - for many years. Event comm-central doesn't seem to have any implementation: https://searchfox.org/comm-central/search?q=nsIGlobalHistory2&path= We should remove the idl and clean up the dead code.
nsDownloadHistory.cpp that is the generic downloadhistory implementation (not used in firefox, but likely used in other software?) seems to use it yet. Though, if there is no implementation of nsIGlobalHistory2, it's worthless to keep it around, so let's also remove nsDownloadHistory at this point. From a quick glance most callpoints are properly checking for success when instantiating an nsIDownloadHistory handler. Fwiw, the only consumer I am worried about is Android, but it should just do nothing, as before, since nsIGlobalHistory2 is not implemented.
Marco recently worked on excluding nsDownloadHistory from Firefox, maybe he could be interested into removing everything here, so cc-ing him.
Assignee: nobody → standard8
Status: NEW → ASSIGNED
Summary: Remove nsIGlobalHistory2.idl → Remove nsIGlobalHistory2.idl & docshell implementation of nsDownloadHistory
nsIGlobalHistory2 is not implemented by any mozilla-central nor comm-central applications, and hence can be removed. As a result, nsDownloadHistory's implementation can also be removed. MozReview-Commit-ID: 5bQ2OHsV1Ky
Comment on attachment 8989524 [details] Bug 1419704 - Remove nsIGlobalHistory2 and docshell's nsDownloadHistory as they are redundant. Boris Zbarsky [:bz] (no decent commit message means r-) has approved the revision. https://phabricator.services.mozilla.com/D1929
Comment on attachment 8989524 [details] Bug 1419704 - Remove nsIGlobalHistory2 and docshell's nsDownloadHistory as they are redundant. Marco Bonardo [::mak] has approved the revision. https://phabricator.services.mozilla.com/D1929
Pushed by firstname.lastname@example.org: https://hg.mozilla.org/integration/autoland/rev/2291d4737e17 Remove nsIGlobalHistory2 and docshell's nsDownloadHistory as they are redundant. r=mak,bzbarsky
You need to log in before you can comment on or make changes to this bug.