Nathan C. Smith wrote: > Tony did give a demonstration of Freeswitch and it is the next generation of > Open Source telephony -- If they can maintain their high standards and > development momentum many of us may be using it next year. Of course the > 80/20 rule could come in to play And progress could slow drastically. I > hope many of you will support this project through testing and code review. > There could be some excellent cross-pollination.
Firstly, what I'm about to write is in no way meant to troll. I've been using OpenPBX pretty much since it started, and I've been impressed. However, it seems to be kinda dead at the moment, as it has been on and off since about February this year. I suspect that the OpenPBX group is quiet because a lot of people have either taken more interest in FreeSWITCH, or turned back to Asterisk (as I fear I may have to for my own projects). I like the fact that OpenPBX brings together many modules that one would normally have to compile separately for Asterisk (I'm referring to all the cool out-of-tree modules available for Asterisk). But a lot of the code in OpenPBX is very old now (hasn't been touched for eight months). This same code has been updated multiple times in Asterisk, both in terms of enhancements, and bug fixes. I wonder how many bugs are still lurking in OpenPBX, but have since been fixed in Asterisk. I don't see a lot of point in writing new modules for OpenPBX when the core is still inherently inferior. With Asterisk/OpenPBX, it seems that the core depends on the modules, not vice versa. Since an alternate core was always one of the plans for OpenPBX, I propose the idea that what remains of the OpenPBX dev team is refocused on developing PBX modules to wrap around the FreeSWITCH core. Any modules we write for the current Asterisk-derived core will not be easy to migrate, due to the deep interdependency of Asterisk core and modules. On the other hand, Anthony has worked hard to ensure that FreeSWITCH is an easy platform to write modules for. In terms of the Asterisk core... well, you can't really paint over rust. At a quick glance, the features I see as lacking in FreeSWITCH (purely because of its early stage of development), that would be required of an average PBX, are: - completed SIP & IAX registrars, which support auth (waiting on finalisation of "user model" concept) - voicemail (also waiting on user model) - call parking - queuing (I'm told that FreeSWITCH has support for this built in the core, by way of its very design - something like an enhanced ICD) Certain modules, such as voicemail, could be written in JavaScript, due to FreeSWITCH's support of spidermonkey libs. I don't mean to disrespect the hard work that has gone into OpenPBX so far, but I think we're beating a dead horse. I'm watching FreeSWITCH with great interest, as I think it has the potential to be a far more suitable platform for an open source VoIP solution. _______________________________________________ Openpbx-dev mailing list [email protected] http://lists.openpbx.org/mailman/listinfo/openpbx-dev
