'Twas brillig, and Tanu Kaskinen at 13/06/11 17:46 did gyre and gimble: > That said, I think both the configuration methods that I provided > originally should be left as is. The environment variable allows the > user to run just one application with the version limit without having > to worry about other apps. Similarly, the setting in daemon.conf might > be too coarse for some use cases: the user may want to load the tcp > module with some limit, but leave the unix module using the normal > protocol version. Yeah, I reckoned that the env var would be useful too, and I'm relatively indifferent about the modarg, but I guess there is not specific reason to remove it now that it's there. Your example about it being per-transport (e.g. native-socket == latest, native-tcp == older) is a good one, even if the primary use case for this is paprefs's tcp loading which could still trip us up. Perhaps the modarg version could take -1 as the protocol version which means "the latest" that way, default.pa could be modified (by the user - I don't suggest shipping it like this), to have -1 when loading the native protcol, but daemon.conf could have the limited version... the modarg taking priority would mean that socket protocol would be the latest but the paprefs loaded -tcp would be limited. I've probably described that terribly, but in contrast to this morning's ramblings, I'm not tied because of a long day :p > I'll post a couple of patches in the near future that add support for > daemon.conf and client.conf configuration. Or do you want that I merge > those features into the two patches that I already sent, and resend the > patches? What ever you fancy :) Either way is good for me! Col -- Colin Guthrie gmane(at)colin.guthr.ie http://colin.guthr.ie/ Day Job: Tribalogic Limited [http://www.tribalogic.net/] Open Source: Mageia Contributor [http://www.mageia.org/] PulseAudio Hacker [http://www.pulseaudio.org/] Trac Hacker [http://trac.edgewall.org/]