[Asterisk-Users] Digium should develop and sell just Dummy card. For timing...

Andrew Kohlsmith akohlsmith-asterisk at benshaw.com
Wed Oct 15 05:06:25 MST 2003


> > If the software needs a specialcard to keep time then the
> > software is broken or poorly designed.

> Don't complain so loudly unless you're willing to contribute the
> fixes.  Opinions are like assholes, and you know where that's going.
> Takes something else entirely to fix a perceived problem.

He is correct though; I am willing to put money on the idea that the 
Tiger320 chip is NOT a stratum 1 time device (or even stratum 3 for that 
matter) -- Chances are that Digium was trying to tie in their hardware and 
make money to support * that way, and you can't really fault them for that.

There's quite a difference between an uninformed opinion and the opinion of 
someone who's actually written software and designed hardware that meet 
more stringent requirements than what is being discussed, do you not think?

I agree though -- there is _no_ reason why the USB, RTC or even regular PIT 
timer can't do this job just as easily...  In fact, both USB and RTC 
options have been created and seem to work just fine for most people.  
Providing hardware to do this also gives an advantage that * can be ported 
to multiple platforms with minimal software shuffling, and since drivers 
would have to be written for the card anyway, you've got your time source 
with you where ever you go.

There are a number of questions that keep coming up that Digium seems 
unwilling to answer.  It'd be nice if they just came right out and said 
"Guys we need to make money to support * development, this is how we are 
proposing to do it..." but instead we get closed lips and speculation.

It does seem that there are a number of systems out there that have 
marginally-compliant PCI busses or some other slightly odd aspect to them 
that make jitter and echo on * worse than other systems.  It'd be really 
nice to get a database of known-bad hardware together.

Regards,
Andrew



More information about the asterisk-users mailing list