[Asterisk-Dev] A few driver ideas...

[Christian Richter]

kpj at junghanns.net wrote:

>>A few idea's for drivers have been bouncing around in my brain, and I'd
>>be interested in some feedback from the asterisk dev community...
>>
>>1. zaptel wrapper for mISDN. Would this add any value to anyone? I've
>>had a look around in the code for both drivers and it doesn't seem too
>>impossible...
>>    
>>
>
>What should that look like and what should it do? mISDN handles the
>complete protocol stack in the kernel but zaptel is just a stupid layer 1
>device...
>
>  
>
>>1a. A hypothetical question. ISDN adapter A has a perfectly functional
>>mISDN (and therefore CAPI) driver available for it. ISDN adapter B has a
>>perfectly functional zaptel driver available for it. Both adapters are
>>reasonably priced and are functionally identical. For use under
>>Asterisk, which would you prefer?
>>    
>>
>
>Well, zaptel ISDN drivers (as in BRIstuff) have the advantage of
>delivering a timing source to asterisk. But they also reduce latency a lot
>when you are bridging from an ISDN BRI card to another zaptel device (may
>that be a PRI card or some FXS/FXO thingies...) because it is bridged
>inside the kernel. Also debugging signalling problems in userspace is much
>quicker than in kernelspace.
>An additional benefit would be that you get layer 1,2,3 certified (TR3,
>TRB3-A1) drivers (and protocol stack) with our zaptel approach.
>
>  
>
* mISDN  bridges in Hardware if possible and in Kernel Space else, which 
has at least as low latency as zaptel has
* the comunity of mISDN is big
* mISDN is not a patch to asterisk/zaptel

>>2. TDMoE driver for existing ATA's (eg sipura/linksys). At a guess this
>>would involve either getting Linux to run and writing a driver for any
>>'foreign' hardware... including the F[SX]O ports, or some insane black
>>magic in wedging a driver into the existing O/S. Unless someone has
>>Linux already running on one (I've looked and can't see any evidence of
>>it having been done), then this one probably goes in the 'too hard'
>>bucket. Maybe TDMoE doesn't really make sense when you are talking about
>>one channel...
>>
>>    
>>
you won't get anyone make this i think. It would probably be easier 
developeing a new ata for that.

>>3. zaptel driver for the 'netjet' PCI BRI adapter (one of the few
>>available in Australia, and by far the cheapest). There is an mISDN/CAPI
>>driver for it now, but it is a bit unreliable for me at the moment.
>>Maybe my time would be better spent fixing bugs in that...
>>    
>>
post that bug at www.isdn4linux.de/mantis they have a bug tracker, like 
digiums.

>
>Can you tell me which chipset they are using? I already have some
>winbondish card laying on my desk that is waiting for a zaptel driver
>(unfortunately it only supports TE mode but the chipset is suitable for a
>good zaptel driver compared to the HFC-S PCI A chipset).
>
>  
>
>>James
>>_______________________________________________
>>    
>>
>
>best regards
>
>Klaus
>--
>Klaus-Peter Junghanns
>
>
>
>_______________________________________________
>--Bandwidth and Colocation provided by Easynews.com --
>
>Asterisk-Dev mailing list
>To UNSUBSCRIBE or update options visit:
>   http://lists.digium.com/mailman/listinfo/asterisk-dev
>
>  
>