[App_rpt-users] [PATCH] Zaptel modified for newer kernels & IOCTL issues.

[James Le Cuirot]

Hello all,

I'm not a radio user but I've been helping Ken out and I made the
necessary changes to get Zaptel working with Linux 3.1.0. I know it is
customary to send patches directly to the list so here it is.

Ken didn't mention that I did not fix all the drivers. I was primarily
just trying to get wctdm working. If you blank out the SUBDIR_MODULES
variable in the Makefile then the build will complete, albeit with some
drivers missing.

No specialist knowledge is needed to complete this task as I merely
applied the same changes that have already been applied to DAHDI. Both
the name change and the other numerous modifications since then mean
that existing patches cannot simply be applied.

This brings me to a controversial point. As I won't be staying here
long, I won't mince my words. Why on earth are you still dragging your
heels with Zaptel? All the work that went into making this patch had
already been done before by somebody else. This made the effort feel
pointless, hence why I wasn't compelled to finish the job. There have
no doubt been countless other feature additions and bug fixes to DAHDI
that you are missing out on. Jim complains that DAHDI has stupid bugs
that he had to fix in his Zaptel fork. I'm new here so I don't know if
he's ever mentioned exactly what these bugs are. Why not fix them in
DAHDI? If upstream doesn't agree, you can still continue to apply these
patches against future versions. You can't cling to the past forever in
the hope that things won't break because they inevitably do.

Jim also stated that DAHDI would not work for radio setups. Is this
really true? I gather that if you don't have a telephony card, only the
ztdummy module is required for timing. I've seen the code for ztdummy
and dahdi_dummy. It's relatively short. Do the aforementioned bugs
apply to this code? Would they really be that hard to fix? I wish I had
the radio hardware just so I could try it for myself.

As for Asterisk, I don't know whether Jim insists on this old version
because later versions don't support Zaptel or because he also has
other issues with it. If it's the latter then I agree with the
sentiment but not the solution. You only have to look as far as the
publicly posted rationale for FreeSWITCH to learn that Asterisk has
some fundamental flaws. This was written by someone who invested an
awful lot of time and code in Asterisk but ultimately decided that a
clean break was the only way to go. If Asterisk is so bad, why have you
stuck with it and how does clinging to an older version help? One would
hope that it has at least gone some way to address the flaws in more
recent versions. I have just begun developing a new VoIP product and
will be evaluating Yate in favour of Asterisk very soon. I don't know
much about it yet but I like what I've seen so far. Maybe you should
also consider the alternatives?

Zaptel wasn't the only thing I had to modify for Linux 3.1.0.
chan_usbradio itself was also broken. I need to go into some detail
about that so I'll save it for another post.

Regards,
James
-------------- next part --------------
A non-text attachment was scrubbed...
Name: zaptel-newer-kernels.patch
Type: text/x-patch
Size: 20314 bytes
Desc: not available
URL: <http://lists.keekles.org/pipermail/app_rpt-users/attachments/20111126/b9ea9799/attachment.bin>