On Feb 2, 2008, at 2:21 PM, Mike Jagdis wrote:

> On Sat, Feb 02, 2008 at 02:48:39AM -0500, Daniel Melnechuk wrote:
>> I downloaded 20080201 snapshot and tried to build and when i run
>> configure i get these messages:
>>
>> checking vale.h usability... no
>> checking vale.h presence... no
>> checking for vale.h... no
>> configure: error: Vale version 0.0.1 or newer is required to compile
>> Callweaver.
>>
>> [...]
>>
>> My set of commands to build vale are:
>>
>> $ curl -O http://www.soft-switch.org/downloads/vale/vale-0.0.2.tgz
>> $ tar zxf vale-0.0.2.tgz
>> $ cd vale-0.0.1
>> $ CFLAGS="-I/usr/local/callweaver/include" LDFLAGS="-L/usr/local/
>> callweaver/lib" \
>>   ./configure --prefix=/usr/local/callweaver
>> $ make
>> $ sudo make install
>>
>> The vale lib and the vale headers are in the correct place. The  
>> 1.2.0-
>> RC5 configure worked fine but it didn't need vale.
>
> You should drop the --prefix from the configure command for vale  
> and let
> it put it in /usr/local (not /usr/local/callweaver).
>
> trunk needs vale, 1.2 doesn't
>
> Mike
>
HI Mike!

Thanks so much for the pointers! I reread the InstallGuide.txt file  
again and I realized my problem. I was running ./bootstrap.sh after  
expanding callweaver-RC-1.1.99.20080202.tar.gz package (downloaded  
from devs.callweaver.org/1.2_snapshots) when all i should do is "./ 
configure; make; make install". Presumably ./bootstrap.sh was  
changing ./configure to look for vale. I knew not to run bootstrap on  
snapshots at some point but had forgot! I hate when i do that. So  
much code, so little time.

./configure ran fine after downloading and installing pkg-config  
into /usr/local and speex into /usr/local/callweaver -- both  
apparently new requirements for 1.2 since RC5. Callweaver built fine  
and i have it installed and running.

Good news, after only 1/2 day of use audio quality of uLaw sip calls  
(all i use is sip) appear to be MUCH better -- really much cleaner  
and without beeps or drops!

Bad news, DTMF after connection to PSTN is still not 100%. It does  
work better as i was able to call into a conference call and checking  
voicemail, but then checking the same voicemail system didn't work a  
second time. So i would hazard that rfc2833 still needs some work.

What is strange is that callweaver never has a problem with what is  
dialed to it to start a call, just what is dialed out to the PSTN via  
my sip provider. There is some evidence to suggest that when it  
fails, it is every other digit that is being dropped. For instance,  
after i called a voicemail system out on PSTN, when it asked for the  
mailbox number and i put in 123-456-7890, the voicemail comes back  
and says "2 4 6 8 0" is not a valid mailbox. Maybe this is a clue?

What can i do to help solve this problem? Is there some trace, or  
log, or something i can do on my end to help this along. I am a  
systems, software, network engineer and would like to do my part to  
help solve this for callweaver seeing as i am having the problem.

Peace,
Dan

_______________________________________________
Callweaver-users mailing list
[email protected]
http://lists.callweaver.org/mailman/listinfo/callweaver-users

Reply via email to