HDB UUCP doesn't switch speeds??

[Dave Hanna]

In article <1991Mar8.212145.5459 at ceilidh.beartrack.com> dnichols at ceilidh.beartrack.com (DoN Nichols) writes:
>In article <1991Mar8.003509.500 at wa8tzg.mi.org> wwm at wa8tzg.mi.org (Bill Meahan) writes:
>>Sometimes, the usual 2400-baud route to my newsfeed becomes temporarily
>>unavailable for various reasons.  No sweat, there is a 1200-baud route
>>on a different phone number so I've just simply added a SECOND entry for
>>my newsfeed with a 1200-baud entry to my /usr/lib/uucp/Systems file.

>	[ ... ]

>>That's not quite true.  The 3b1 HDB **does** TRY to execute the N+1st
>>entry, but if it is a different speed, the serial port doesn't get
>>switched!!
 
>	Well, I guess that I'm 'anybody'.  This is just a guess, I'm not an
>expert on HDB - far from it - but is it possible that DTR isn't being
>dropped between trys, since they're on the same port, and the modem, not
>being reset, is not looking to re-determine the baudrate coming from the
>serial port?  I haven't checked the modem I have here, but I think that it
>might not expect to have its baud rate changed unless it is awaiting an 'AT'
>after reset.

No, I remember this being discussed a couple of years ago.  (I remember
because, though I didn't participate in the discussion, I had the same
problem.)  Unfortunately, as I recall the answer, there is a bug in
the 3b1 implementation of HDB.  There is an ioctl call to set the
speed that ought to be _inside_ the loop that tries successive Systems
entries, and, instead, it is _outside_ the loop, i.e., done just once
at the top.  The result is that it will execute multiple Systems entries
for a single system, but it does them all at the speed of the first one.

Without access to the source, there's no easy fix.



-- 
Dave Hanna,  Infotouch Systems, Inc. |  "Do or do not -- There is no try"
P.O. Box 584, Bedford, TX 76095      |                        - Yoda
(214) 358-4534   (817) 540-1524      |
UUCP:  ...!letni!dms3b1!dave         |