Re: [hylafax-users] Huge Problem

[Lee Howard <faxguy@deanox.com>]

At 10:29 AM 1/26/01 +0100, Dieter Kluenter wrote:
>
>Hello,
>Lee Howard <faxguy@deanox.com> writes:
>
>> At 10:27 AM 1/25/01 +0100, Dieter Kluenter wrote:
>> 
>> >Changing KillTime in any config, that is hfaxd.conf, sendfax.conf or
>> >~/.hylarc, as recommended in man sendfax didn'd work. If you have a
>> >look at the source code, you will realise why that would not work.  Only
>> sendfax -k
>> >changes killtime for a single job. 
>> >After all I changed the settings in the source code and recompiled
>> >it. That works fine now.
>> 
>> Could you post the patch?
>
> In hfaxd/Jobs.c++ I changed:
>
>  defJob.tts		= 0;	    // ``NOW''
>    defJob.killtime	= 72*60*60;  // FAX_TIMEOUT
>    defJob.retrytime	= 0;
>
>In config.h I changed;
>
>#define	FAX_REDIALS     3  	/* number times to dial phone */
>#define	FAX_TIMEOUT   "now + 72 hours"  /* default job timeout(at syntax) */

In reviewing these alterations, I don't see how it fixes anything that was
"broken".  The default KillTime is was three hours, and these changes
simply change the default KillTime to be 72 hours.  Surely the three-hour
default KillTime worked beforehand if these simple changes make a 72-hour
KillTime function.

However, as Giulio has mentioned, I believe... there are much easier ways
to accomodate this.

Following the sendfax manpage a KillTime setting can be added to hyla.conf
or sendfax.conf for a system-wide configuration, or ~/.hylarc can be given
KillTime for per-user configurations - and all this without changing the
code or recompiling.  You have said that this didn't work, however... yet
in tests I ran a few months ago it seemed to work fine.

Now, perhaps you're not using sendfax.(?)  It appears that hfaxd refers by
default to settings in hfaxd.conf and that the default kill-time can be
assigned there via KillTimeMap (which is different in syntax than
KillTime).  I would assume that hfaxd will revert to these settings unless
otherwise specified by the client.  (However, I can't easily test this as I
use sendfax.)  If I am misinterpreting the functionality of KillTimeMap,
and if hfaxd.conf does not provide this functionality - then I'd say that
the client is broken (since clients are made to function with the server -
and not the other way around).  But, I'll agree if that's the case, that
hfaxd could use a patch.  However, this is not what your changes do; they
simply change the default settings.

Lee.



____________________ HylaFAX(tm) Users Mailing List _______________________
 To unsub: mail -s unsubscribe hylafax-users-request@hylafax.org < /dev/null