[hylafax-devel] Re: [hylafax-users] Re: config/usr-2.0

[Jay R. Ashworth <jra@baylink.com>]

On Wednesday, August 23, 2000, at 3:32:03 AM, Dmitry Bely wrote:
> Lee Howard <faxguy@deanox.com> writes:
>> >I don't see any difference with the stock config (99% of your Modem* and
>> >Class* parameters are redundant and so should be wiped out).
>> 
>> Yes, they should be.  I've left them there for possible future
>> alterations.

> No way! Most of them should *never* be changed. By the way, some parameters 
> may be autodected, e.g. ModemType.

I can understand both viewpoints, Dmitry.  Yes, 95% of them shouldn't
be changed... but most people won't be poking around in that file
anyway; and it *can* be useful to have all the config parameters in a
template file; even geeks misplace the manuals.

>>  When diagnosing a problem it's difficult to know what the possible
>> problems are if all of the settings are not listed in one place. That's
>> why I left everything there.  True that the non-default settings are the
>> only ones needed.

> Again, doing so you encourage user to change some setting, that cannot be
> changed. Once he/she will change them, and get completely broken system.

"Stupidty is supposed to be expensive."

>> >Alas, no config file can make all USR modems reliably work in Class
>> >2.0. What's why almost all commercial fax software switches them to Class1
>> >mode.
>> 
>> I can't change USR,

> BTW, why? Your time is cheaper than the good reliable modem? How much time
> have you spent fighting USR? Just multiply this value by your hour rate ...

Because he's only one person, and there are *lots* of Couriers out
there.  It's a good investment, if he's willing to make it; more power
to him.

>>                      but I'm just trying to help in what I can. If I could
>> get my hands on one of these unreliable modems I'd like to see the problem
>> first-hand.  I've noticed that commercial software doesn't always make an
>> effort to go beyond the necessary, so it's no surprise that they simply
>> disable Class 2.0 altogether.

> IMHO wise choice. It's better to use Class1 which is very simple and
> error-free for most modems, than try to work around all those buggy
> Class2.* firmwares. Also note, that some fax protocol features can be only
> implemented using Class1: error-correction mode (ECM), binary file transfer
> (BFT), gray-scale mode and many others. The only problem is that you need
> *good* Class1 software.

Well, yeah, Dmitry, but this can get dicey when you have 400 modems on
the same machine, no?  :-)

Dmitry; you sound like you've groveled in the code enough; you might
have an idea on this.  Have you been near the DestControls parsing
code?

There have been several requests in the past for the ability to
decide which modem to use to send a job based on the destination phone
number -- this is useful, for example, in situation where you have a
foreign exchange line to a different toll area.

It seems to me that if the list of things that the destctrls file can
change could be extended to include ModemGroup (formerly ModemClass),
you could redefine "Any" to be only the modems that are appropriate
for that destination.

Do you know that code, does this sound sane, and does that processing
happen early enough (ie: before modem selection) to make it possible?

(PS: if anyone else has played with the destctrls code, feel free to
jump in, here :-)

Cheers,
-- jra
-- 
Jay R. Ashworth                                                jra@baylink.com
Member of the Technical Staff     
The Suncoast Freenet
Tampa Bay, Florida     http://baylink.pitas.com                +1 727 804 5015


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