->On 07-27-97 13:25 Rick Was discussing problem with Kenneth<-
RC> On Jul 24 17:35, 1997, Kenneth Burgess of 1:2216/77 wrote:
->>On 07-22-97 14:57 Dennis Was discussing problem with Rick<-
RC->> You can generally receive data with no flow control enabled (because
->> the CPU is so much faster than the modem, the PC never has to tell the
->> modem to "wait
KB> Although most often true, not always true. Take for example these
KB> system configurations. Although rare there are still some in use
oday.
KB> System (1) is a bbs user, system (2) is the bbs being accessed.
KB> (1) - 2mhz CPU, 33.6k modem. Max buss at 19,200
KB> (2) - 33mhz CPU, 33.6k modem. Max buss at max modem through put.
KB> Modem 1 & 2 can still negotiate a 33.6k link. However cpu #1 can not
KB> accept data from it's modem at that rate. If modems #1 and #2 are not
KB> BOTH configured for (Modem-Modem Xon/Xoff) flow control. There is a
KB> real possibility that buffer overruns WILL occur.
RC> CPU 1 would possibly require flow control in that situation, but with
RC> a USR modem it will not connect at 33.6 with a 19.2 port rate. The
RC> receiving modem would assert modem to modem flow control with the
RC> sending modem, regardless of the flow control setting used in the PC,
RC> assuming an ARQ connection.
Yes you are correct. The max attainable link would be the max buss
speed of the slower cpu, or the slower modem. Which ever is slowest.
I should have looked my msg over, but was in a hurry. just didn't
notice I had said a link of 33.6 (Wrong) instead of 19.2. :)
And yes in that instance cpu #1 does require flow control. It must
dedicate some of the processor time to screen & memory refresh. Which
does not allow it enough time for a continous stream from the modem.
___ Blue Wave/DOS v2.30
--- TriToss (tm) Professional 10.0 - #246
---------------
* Origin: The Other Side * Massillon, OH * (330) 832-0565 (1:2216/77.0)
|