A New Timing API for Asterisk, Silencing Digium Critics
Russell Bryant has post details of a new timing API for Asterisk:
Starting with Asterisk 1.6.1, you will no longer need DAHDI installed at all to get proper timing in Asterisk. There is a new timing API, and there are already two implementations. There is a DAHDI timing interface (res_timing_dahdi) and another (res_timing_pthread) that has no special kernel requirements.
I hope this will shine some much needed light on the folks that are clamoring that Digium is keeping Asterisk proprietary by requiring a DAHDI timing source. One particularly annoying blog post states that the Asterisk fork(s) and other related VoIP/PBX software are gaining speed, because they are not tied to ‘proprietary timers needed for the Digium version of Asterisk.’
Need I inform you (usken.no) that:
- The DAHDI interface has never been proprietary, the code is open and licensed under the GPL.
- Timing is only utilized for MeetMe, the IAX2 Trunking feature and optionally during file playback.
- app_conference is an alternative to MeetMe.
- IAX2 Trunking feature is an optional feature that is not friendly to voice quality in high packet loss situations, so the loss of a timing device does not cripple or otherwise hinder Asterisk whatsoever.
- File playback has a ’software timing’ method, if no DAHDI timing is available. This software timing method very well may also be available to IAX2 Trunking, but I am unsure if that functions or not.
I do not see any proprietary timers that are hindering Asterisk.
I see this post from usken.no as yet another attempt to slant the reasoning why people positioned themselves against Digium’s efforts.
Granted I have not agreed with every decision Digium has made (who does?) however I do see that Digium’s interest is very much with keeping Asterisk open source, along with developing a viable business model and an ever evolving ecosystem around its open source creation.
|
|
|
Related Posts:
Digium Launches New Headquarters
You Can’t Make This Stuff Up
Digium Innovation Awards
Digium Acquires SwitchVox
Digium Launches Embedded Platform
Asterisk is so behind of others now… I hope by 1.6 they will catch up with FreeSWITCH, and also fix their core problems, bugs… and get some new features that FS has.
VoIP Enthusiast: Please explain what you mean. Exactly where is Asterisk behind, what core problems, what bugs… Digium folks do read this blog, so don’t hold back - you can help fix Asterisk, if you put in at least some effort.
These kinds of general comments really boil my blood.
hey… don’t take me wrong, I love Asterisk and I want it to be the best… I report bugs from time to time on Mantis.
I’m not a coder, but a sysadmin and I try to do what I can… I understand that they are doing what they can too… and I also know that they are doing a lot with 1.6, and I’m really excited about it, but in the past I been having some problems…
For example, I tried to deploy an asterisk server on a XEN environment, and I needed conference, so I tried to use MeetMe, but when I tried to use zaptel/ztdummy… it didn’t work well with XEN, so I had to find another way to do conferencing, I found app_conference, but it lacked some features that MeetMe had… I got app_conference from the SF repository… I had no choice than to go with a dedicated (real machine) and use asterisk there.
This finally fixed some of my problems… but in the long run I experienced more, DTMF problems, etc. Now the server is working great, but I still would like to see my bugs fixed and improvements overall.
I think Asterisk is great, and no… I’m not going to use FreeSWITCH, I refuse to use XML for my configs, but still… they have some nice features that I would like to see in asterisk as well, that’s what I mean when I say that Asterisk is a bit behind of them.
I prefer to see Asterisk improving it over time… because it’s what got me into the voip world.
I think Asterisk folks are doing an amazing work with Asterisk and I would like to continue seeing Asterisk progressing.
Voip Enthusiast: Perhaps your issues are with the Xen environment? Last night I installed Asterisk 1.4 with Zaptel + ztdummy into a CentOS 5.1 x86_64 minimal image (plus dependencies for Asterisk/Zaptel) into a VMware Server virtual machine. I then configured a SIP peer, and the following minimal dialplan:
exten => service,1,NoOp()
exten => service,n,Answer()
exten => service,n,Playback(tt-monkeys)
exten => service,n,Hangup()
exten => 555,1,NoOp()
exten => 555,n,Answer()
exten => 555,n,Milliwatt()
I then load tested the machine with SIPp (http://sipp.sourceforge.net) while my Polycom was playing Milliwatt() (I dialed extension 555). Then I started generating calls with the default uac script and was getting around 60-70 simultaneous calls before audio was interupted on the Milliwatt() application.
I find this test to be extremely simple to implement, and is a pretty good indicator of approximate simultaneous call ceiling.
So in my opinion, your perceivied issues of short-comings in Asterisk on your Xen implementation seem to be more related to your selected distrubution model, and not of Asterisk itself.
Leif.