Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!tut.cis.ohio-state.edu!cs.utexas.edu!sun-barr!sun!imagen!atari!portal!cup.portal.com!JerryK From: JerryK@cup.portal.com (Jerry E Kindall) Newsgroups: comp.sys.apple Subject: Re: ZMODEM/EASYLINK Message-ID: <19638@cup.portal.com> Date: 19 Jun 89 00:41:29 GMT References: <8906180052.AA20316@obsolete.UUCP> Organization: The Portal System (TM) Lines: 24 Zmodem is a very fast protocol, which doesn't require acknowledgement of each block sent. Blocks that aren't correctly sent are NAKed and sent again. Actually, the protocol isn't quite that simple (look at Chuck Forsberg's specs for Zmodem sometime for a real thrill) but that's about the gist of it. The protocol will work fine with Apples because it has a convention for indicating the "buffer size" on the receiving machine. In other words, if ProTERM can only receive 15K of data before it has to save it to disk, the sending computer will only send 15K at a time, then wait for the go-ahead before sending the next chunk. The problem seems to be in ProTERM's implementation of Zmodem. ProTERM's Ymodem is flakey on GEnie as well, but I've personally never had problems with it. In any case, Greg Schaeffer is working with GEnie to get the problems ironed out. By the way, I've used Zmodem here on Portal, and on a local MS-DOS BBS, and ProTERM handles both of those hosts fine. I suspect that the Zmodem Portal and this MS-DOS BBS use differs in some subtle way from Forsberg's specifications. ProTERM probably is off-base in the same direction, which would explain why its Zmodem works with those hosts but not on GEnie. /\ Jerry Kindall JerryK@cup.portal.com \/ Death to COBOL GEnie: A2.JERRY ALink: A2 Jerry