Ticket #32682 (closed update: fixed)
glib2-devel: Update to 2.31.8
| Reported by: | blender3dartist@… | Owned by: | ryandesign@… |
|---|---|---|---|
| Priority: | Normal | Milestone: | |
| Component: | ports | Version: | 2.0.3 |
| Keywords: | Cc: | ||
| Port: | glib2-devel |
Description
glib2-devel is out-of-date, the latest version is 2.31.6 and the current macports version is 2.29.92. The source release can be found at it's usual location at: http://ftp.gnome.org/pub/gnome/sources/glib/2.31/
Change History
comment:2 Changed 17 months ago by ryandesign@…
- Owner changed from macports-tickets@… to ryandesign@…
- Cc ryandesign@…, openmaintainer@…, blender3dartist@… removed
- Status changed from new to assigned
This update is more difficult than usual as they have excised some code upstream that we were patching. Currently our ed script is failing and I need to spend some time investigating how we need to modify it to make it work again.
comment:3 Changed 17 months ago by blender3dartist@…
What is the purpose of the ed script? Is it required?
comment:4 Changed 17 months ago by ryandesign@…
- Status changed from assigned to closed
- Resolution set to fixed
- Summary changed from Update glib2-devel to dev version 2.31.6 to glib2-devel: Update to 2.31.8
Updated in r88770. I hope it works. Upstream has removed their previous mutex code, so I removed the corresponding sections from the configure patch and ed script, but I don't know if whatever they've implemented as a replacement needs any new patches.
Yes, the ed script is required. It is there to edit the config.h file the configure script generates, to correct the assumptions made by the glib developers, that one can determine the size and endianness of integers and pointers at configure time and then use that information at build time; this assumption is incorrect on OS X when you build for multiple architectures simultaneously -- architectures which differ in size and/or endianness.
Cc Me!