diff options
| author | Wilmer van der Gaast <wilmer@gaast.net> | 2010-12-07 23:41:49 +0000 |
|---|---|---|
| committer | Wilmer van der Gaast <wilmer@gaast.net> | 2010-12-07 23:41:49 +0000 |
| commit | c775a58faa7d5905b06e2f8900db7337082d5165 (patch) | |
| tree | 92bbeb0c5316ec6df1b2c3988e78801a698af268 | |
| parent | 83f1bd1d195b21617bbc58b3c36e9d23d7905661 (diff) | |
At least don't disable Twitter anymore for libpurple builds. libpurple
doesn't have Twitter support by default and it's unlikely that someone will
prefer libpurple-twitter over BitlBee's Twitter code anyway.
Also automatically disable libpurple protocol modules if a native variant is
already set up. The linking conflict works the other way around (libpurple
modules may break if native modules for the same protocol are compiled in)
so this should mostly resolve the linking conflict problem.
| -rwxr-xr-x | configure | 3 | ||||
| -rw-r--r-- | protocols/purple/purple.c | 5 |
2 files changed, 6 insertions, 2 deletions
@@ -600,13 +600,12 @@ EOF jabber=0 oscar=0 yahoo=0 - twitter=0 if [ "$events" = "libevent" ]; then - echo echo 'Warning: Some libpurple modules (including msn-pecan) do their event handling' echo 'outside libpurple, talking to GLib directly. At least for now the combination' echo 'libpurple + libevent is *not* recommended!' + echo fi fi diff --git a/protocols/purple/purple.c b/protocols/purple/purple.c index 07f385ff..d2e85554 100644 --- a/protocols/purple/purple.c +++ b/protocols/purple/purple.c @@ -1286,6 +1286,11 @@ void purple_initmodule() PurplePlugin *prot = prots->data; struct prpl *ret; + /* If we already have this one (as a native module), don't + add a libpurple duplicate. */ + if( find_protocol( prot->info->id ) ) + continue; + ret = g_memdup( &funcs, sizeof( funcs ) ); ret->name = ret->data = prot->info->id; if( strncmp( ret->name, "prpl-", 5 ) == 0 ) |