Rockwell problems.

• Messages sorted by:[ date ][ thread ][ subject ][ author ]

Hi,

Marc Eberhard wrote:
> > 02/04 21:13:49 yS1 send: \dATQ0V1H0[0d]
> > 02/04 21:13:50 yS1 waiting for ``OK''
> > 02/04 21:13:50 yS1 got: [0d][0a]OK ** found **
> > 02/04 21:13:50 yS1 send: ATS0=0Q0&D3&C1[0d]
> > 02/04 21:13:50 yS1 waiting for ``OK''
> > 02/04 21:13:50 yS1 got: [0d]
> > 02/04 21:13:50 yS1 CND: OK[0a]ATQ0V1H0[0d]
> > 02/04 21:13:50 yS1 CND: ATQ0V1H0[0d][0a]OK ** found **
> 
> Huh? Here we are out of sync. It since the mgetty lib does not check
> the command echo it seems to read the OK of the previous command...
> (Gert: Now you know, why I used to check this)! The rest of the
> initialization must fail...

I agree. I find this spurious "OK" very strange (it should be caught by
the various "wait for line to clear" and "flush buffer" calls). I have
seen this occasionally if the modem init sequence wasn't terminated with a
final "expect OK", but never in the very beginning of the chat sequence.
Bad, vary bad. :-(

My problem with the approach of checking the command echo is that you
can't use a generic expect-send-expect-send loop for that. So I would
have to change that to a "modem-specific" function, that has a very clear
idea about AT commands (and echo), and responses, and works on a
line-by-line basis instead on char-by-char as do_chat does now.

I will have to think about that. Maybe it's the only way to go, 
because the "distinctive RING" handling needs a radical change in the
inner workings of mgetty anyway.

Promise, I'll fix this. But not this week (faxrunqd work to do).

gert
-- 
USENET is *not* the non-clickable part of WWW!
           //www.muc.de/~gert/
Gert Doering - Munich, Germany      gert@greenie.muc.de
fax: +49-89-35655025    gert.doering@physik.tu-muenchen.de
.