[Asterisk-Dev] UK Caller ID patch and new CVS

[Rich Adamson]

> > On Thursday 22 July 2004 11:25, matt.riddell at sineapps.com wrote:
> > > It's still pretty crap that we all bought X100P cards that Digium
> > > must have known would not work in our market, only to be told that
> > > we will have to buy new cards, and that the old ones are only
> > > usefull as paperweights...
> > 
> > "Must have known"?  Considering that every country seems to have its
> > own standard for telephone service, it's a pretty broad statement to
> > suggest that Digium could have or should have known how CallerID
> > worked in the UK when they designed the X100P, much less when they
> > began to get demand for the cards in multiple foreign countries.
> 
> :-) No sorry I meant the impedence not being alterable from software, 
> and therefore only legally usable in the USA.
 
FWIW, if you look back over history you can easily see there was no
effort by digium to establish x100p resellers outside the US, and its
really only been the last twelve months or so that * has picked up
significant interests in other countries.

The impedance matching issue is somewhat important, but not to the 
degree that one could say the card won't work in non-US countries. The
facts now show it does work just fine in many countries, however the 
echo cancellation routines are not sufficiently resilient to remove 
that echo in many cases. That limitation is not only affecting 
non-US implementations, but also many US systems.

Its also fairly common knowledge the x100p was not designed/built by
digium, but rather they choose to use an existing modem card that had
the chipsets (etc) that could be used for entry-level systems at a
very low cost, and those cards _were_ being manufactured in hugh
volumes making the cost per card very reasonable. (Compare that 
cost to what you'd pay Nortel for the equivalent as one example.) 
Its certainly not difficult to understand the economics of that and 
its certain that most understand modem sales (worldwide) have 
dropped very significantly. That suggests the x100p-type (and the
follow on x101p) cards will become more extinct over the next 
months/years.

The bottom line is that given our perfect 20-20 hind-sight (now),
is there truly anyone that would have made a different design decision
two or three years ago? If you think the answer really is yes, then
why didn't that individual step up and either object to it, or
program the equivalent of * at that time (instead of Mark)?

Complaining _now_ about a design decision that was made years ago
is both unreasonable and unfair, particularly given that no one
at that time could have even guessed at the popularity of * today.

Now that we're all older and wiser, lets get on with the tasks of
advancing the code, including making the echo cancellation more
resilient (which is really the source of your complaint, not the
card itself).

Rich