Package Details: freeswitch 1.10.12-1

Git Clone URL: https://aur.archlinux.org/freeswitch.git (read-only, click to copy)
Package Base: freeswitch
Description: An opensource and free (libre, price) telephony system, similar to Asterisk.
Upstream URL: http://freeswitch.org/
Licenses: MPL
Conflicts: freeswitch-fixed, freeswitch-git
Provides: freeswitch
Submitter: None
Maintainer: Korynkai
Last Packager: Korynkai
Votes: 17
Popularity: 0.001234
First Submitted: 2009-03-02 18:40 (UTC)
Last Updated: 2024-11-07 19:00 (UTC)

Required by (0)

Sources (25)

Latest Comments

« First ‹ Previous 1 .. 7 8 9 10 11 12 13 14 Next › Last »

PrimusNZ commented on 2015-01-30 00:14 (UTC)

Is anyone else getting a core dump on running lua scripts ? EXECUTE sofia/external/2002@**.***.**.** lua(app.lua ring_groups) *** Error in `/usr/bin/freeswitch': free(): invalid pointer: 0x00007f5f38103338 ***

andriesinfoserv commented on 2015-01-20 06:15 (UTC)

@sanerb: That was quick. thanks for maintaining this.

sanerb commented on 2015-01-18 04:47 (UTC)

1.4.15 is now pushed. i'm testing a makepkg now to confirm it builds, but i don't see any big differences between versions so it should be okay. if not i'll rel-bump it.

sanerb commented on 2014-11-25 18:28 (UTC)

or poor timing, lol. updated to 1.4.14

philwo commented on 2014-11-25 08:22 (UTC)

@sanerb: Thank you! I noticed that the Freeswitch maintainers just released 1.4.14. Talk about good timing. ;)

sanerb commented on 2014-11-24 19:01 (UTC)

@philwo- thanks! package has been updated to 1.4.13 per upstream's version of production.

sanerb commented on 2014-10-05 00:41 (UTC)

hrm. I can try but were I in their position, I'd probably close as WONTFIX or INVALID. This is in part due to Arch using a newer build toolchain than upstream typically tests against (so the method may not be obsolete in their implementation), as well as their intended behaviour (normally it should compile; note they're warnings, not actual errors- they enable warnings as failing as errors during the make process to aid in package maintainers if my recollection of reading the source is correct). But I'll go ahead and file a bug so they're at least aware of it in case they aren't. :)

chetwisniewski commented on 2014-10-04 21:04 (UTC)

Thanks sanerb. Maybe this should get reported upstream?

sanerb commented on 2014-09-30 16:25 (UTC)

Okay. New version uploaded. I had three possible ways of fixing this: 1.) Simply disable the addition of -Werror to CFLAGS ('if test "$ac_cv_gcc_supports_w_no_unused_result" = yes; then' +2 lines in configure.ac) 2.) Strip out references to _BSD_SOURCE in src/include/switch.h (there are only two) 3.) Like 2., but replace with references to _DEFAULT_SOURCE I had planned on 3., but I'm not a developer so I'm not experienced or knowledgeable enough about gcc/make/etc. about the repercussions of such. I could have gone with 2, but that would have removed support for BSDs using the PKGBUILD (and we should keep that in there lest we forget our ArchBSD friends). So I went with 1. It's a bit more verbose as it flashes a LOT of warnings and alters the distributed source in a bit of an unexpected way, but it doesn't kill the compile now. Why only the apr-util libs actually die, I have no idea, but this is fixed now. Many thanks for reporting this, chetwisniewski!

sanerb commented on 2014-09-30 03:37 (UTC)

Thanks- I'm working on the 1.4.9 release. And I seem to get the same error and even have apr-util installed separately so it looks like it prefers to compile and use its own bundled version during compile. The "error" that causes the build to die isn't even an error so much as a warning, but because their default build options treat it as an error. There's nothing on their Jira, so this is probably something in makepkg.conf and I'll probably need to package a custom one. I should have this figured out shortly. Thanks for bringing it to my attention!

« First ‹ Previous 1 .. 7 8 9 10 11 12 13 14 Next › Last »