> On 2021-08-31-T, at 07:14, Christopher Nielsen <[email protected]> > wrote: > >> On 2021-08-31-T, at 06:28, Ryan Schmidt <[email protected] >> <mailto:[email protected]>> wrote: >> >> I haven't been following everything closely and wasn't aware harfbuzz devel >> ports were going to be introduced. But if they are, then you can make the >> required dependency change to all ports in a single commit. I would not >> consider it necessary to ask permission of the port maintainers since it's >> not a matter of opinion; it's just a necessary change so that MacPorts can >> continue to function as intended. > > Sounds good; changes committed via the following: > > https://github.com/macports/macports-ports/commit/f685ca73e40027b414af643d067ea402e754479c > > <https://github.com/macports/macports-ports/commit/f685ca73e40027b414af643d067ea402e754479c> > https://github.com/macports/macports-ports/commit/77e3d9ee214f3c4826c4320a7ed3d7de1a00310d > > <https://github.com/macports/macports-ports/commit/77e3d9ee214f3c4826c4320a7ed3d7de1a00310d> Folks, now that this change is done, we can test dependent ports with the Devel versions of harfbuzz/harfbuzz-icu, along with pango.
Does anyone know of any particularly fragile dependents on these - particularly in terms of runtime behavior - that we can target? Meanwhile, I’ll try to rebuild as many dependent ports as possible locally, to identify obvious build issues. And if any maintainers are interested in replacing their non-Devel versions of pango/harfbuzz/harfbuzz-icu with the Devel versions, and testing locally, we’d be grateful! Any thoughts on the best approach to validate all of these, prior to updating the non-Devel versions? Thanks, -Chris
