home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.sys.apollo
- Path: sparky!uunet!gatech!destroyer!terminator!pisa.citi.umich.edu!rees
- From: rees@pisa.citi.umich.edu (Jim Rees)
- Subject: Re: SLIP & SR10.3
- Message-ID: <5aaca51e.1bc5b@pisa.citi.umich.edu>
- Keywords: slip domainOS 10.3
- Sender: news@terminator.cc.umich.edu (Usenet Owner)
- Reply-To: Jim.Rees@umich.edu
- Organization: University of Michigan IFS Project
- References: <1541@lysator.liu.se>
- Date: Thu, 20 Aug 1992 15:11:16 GMT
- Lines: 21
-
- In article <1541@lysator.liu.se>, pen@lysator.liu.se (Peter Eriksson) writes:
-
- I have some slight memories of someone saying that the hardware
- flow control for the serial ports was broken under SR10.3. Is that
- so? I have had very big problems with running a SLIP link between our
- Sun 4/380 and our four old Apollo (DN300, DN550, DN560 and DN660) machines.
-
- The sr10.2 tcp slip had some problems. It had a fixed mtu of 1000, it
- turned off hardware flow control on "ifconfig up," and, worst of all, the
- sl_output routine just did a plain term_$write. So the whole tcp process
- would hang waiting for the tty to drain. This is fatal if you're using your
- Apollo as a slip gateway to your home ethernet and token ring, as I do (my
- ISDN line isn't set up for ip yet).
-
- There is nothing wrong with the tty hardware flow control (below 19.2
- anyway), but you need to turn it on with tctl after you bring the interface
- up.
-
- I have fixes for these problems, and vj header compression too, but I don't
- think they've been incorporated into later releases of Domain/OS. I am
- unable to redistribute these fixes outside of UM, so don't ask.
-