Summary:ASTERISK-00263: Latest CVS checkout for zaptel breaks WCFXS
Reporter:jmbrown (jmbrown)Labels:
Date Opened:2003-09-13 00:11:22Date Closed:2004-09-25 02:46:18
Versions:Frequency of
Description:Latest CVS checkout for zaptel breaks the loading of wcfxs and maybe wcfxo.  Compiles clean, installs clean, when doing a modprobe wcfxs or  modprobe wcfxo  the driver complains that the FXO and FXS settings need to be reversed. The config file that is being used was a working config, so the signalling reversal was already set.  Ergo not the problem.  ztcfg -vvvv  shows that all the cards are now FXO, not the 2 FXO and 1 FXS that is physically in the system
Comments:By: John Todd (jtodd) 2003-09-14 15:56:23

John - is this fixed now?  I recall seeing some stuff on the list about things working for you.

By: jmbrown (jmbrown) 2003-09-14 16:12:17

John - I believe that the drivers should load independantly and not have a ordering issue.  Previous releases of the coded didn't have this problem.  So I think there is a code issue. At least its probably an issue with how the kernel probes and what registers are set when.  Not having any docs for the chip set I can't dive to deep.
I'd say we are in "work around" mode and that there still is an issue with the code.  Mark would be better able to comment..

By: John Todd (jtodd) 2003-09-14 16:20:07

OK.  Two tasks for you then: 1) please comment in this note as to how you fixed the symptoms, and 2) when you talk to Mark or someone at Digium on Monday, please have them update this ticket with whatever the outcome is, so we can move towards closure of the ticket.  Otherwise, tickets will remain open "forever" which is not a good situation.

By: Mark Spencer (markster) 2003-09-19 09:31:45

Is this ticket ready to close?

By: jmbrown (jmbrown) 2003-09-19 11:40:02

No I do not believe this ticket is ready to close.!
1. No one has actually addressed the issues in this ticket
2. The problem still exists

Ticket systems and such are designed so that the responsible parties (in this case the code geeks) actually put data into the system, fix the problems, etc.  That way future searches of the ticket system will find (ANSWERS).  Heck this ticket hasn't even been assigned to someone yet.  If this was a bay area company, I as the dev team manager would really  be wondering why this 6 day old ticket hadn't been assigned and that the only entries was "can we close this".

Whats broken here is.

1. That someone from Digium hasn't actually had this ticket assigned to them.
2. That the only reply from Digium was a IM chat for a quick work around
3. That the problem still exists.
4. The assumtion that i was going to talk to somone at Digium on monday

So I suggest that the maintainers of the code base (ergo Digium) have this ticket assigned to a person able to actually resolve and respond with a proposed resolution.  That means, to me at least, that the code geek setup a
test system, determin why the loading of the drivers fail and require a specific order, that the code (wcfxs I belive) gets patched to remove this
problem, and that this ticket be updated with the correct CVS tag when that code is released.  Then the reporter of the problem can retrieve that CVS version, install, build, test and provide feedback.  If the problem is resolved then the customer (bug reporter) can request that the ticket be closed. If the customer blows off requesting the ticket to be closed, then the
bug manager can see that
a) a response with resolution has been filed
b) the customer was notified
c) lacking response from the customer, can close the ticket.

Given this customers quick response to bugnote adds, I don't think you will find "the customer didnt' respond" as an issue.

Guys, I've done this for a living, hence my comments.

By: Brian West (bkw918) 2003-09-19 17:51:07

This is the same issue isn't it?


By: Mark Spencer (markster) 2003-09-23 13:49:37

I cannot duplicate this problem.  If you want to find me on IRC I can ssh in and look at your setup and see what is different.

By: Mark Spencer (markster) 2003-09-23 13:52:11

Also, keep in mind this is the *bug tracker* and not the *request tracker* that we use at Digium for maintaining support requests.  The purpose of the bug tracker is for the community at large in general.  If you need a quick response, and you've purchased a product from us at Digium, then contact us directly at support@digium.com, where we *do* have people taking and monitoring requests at a rapid pace.

By: Mark Spencer (markster) 2003-09-25 08:17:46

Were you able to place a ticket at Digium?  If so I'd like to close this one out.

By: Mark Spencer (markster) 2003-09-26 19:37:59

Unfortunately I don't seem to be able to locate the customer :(