Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!think.com!zaphod.mps.ohio-state.edu!cis.ohio-state.edu!ucbvax!ucsfcgl!babar.mmwb.ucsf.edu!srp From: srp@babar.mmwb.ucsf.edu (Scott R. Presnell) Newsgroups: comp.sys.sgi Subject: Re: stty/sgtty in IRIX, lpd, and transcript Message-ID: Date: 28 May 91 03:59:27 GMT References: <9105220304.AA12959@nazgul.physics.mcgill.ca> <82588@bu.edu> Sender: daemon@cgl.ucsf.edu Lines: 32 eap@bu-it.bu.edu (Eric A. Pearce) writes: >I managed to get Transcript 2.1 compiled under IRIX 3.x without much >trouble, I found it preferable to one SGI sells since it's the 'BSD' >instead of the 'SYSV' version. There was one substantial set of changes that I made to the transcript 2.1 software - the BSD version of pscomm expects to use the BSD signal semantics where a system call interrupted by a signal will be restarted. This is currently not available under IRIX, so I blocked signals around these critical regions. That stopped the core dumps from pscomm. >We had a locally hacked version of the Tahoe lpd, but only for printing >remotely. It looks like IRIX 4.0 has a Berkeley version. We still >don't have any directly connected postscript printers. IRIX 3.3 seems to have /usr/etc/lpd (it's on the dist tapes) - but I haven't tried it for directly connected printers. I also have patches to lpd to use termio stuff (I think I got them from BU :-). But, we run a modified version of lpd that handles printing via terminal servers - one of our iri operates in this manner. This particular setup is sort of between printing to a directly connected port and sending to another machine - it was different enough to expose problems in SIGCHLD handling, wait() timing, and a few random BSDisms in lpd, and the pscomm problems. If anyone can use these changes, let me know. - Scott -- Scott Presnell +1 (415) 476-9890 Pharm. Chem., S-926 Internet: srp@cgl.ucsf.edu University of California UUCP: ...ucbvax!ucsfcgl!srp San Francisco, CA. 94143-0446 Bitnet: srp@ucsfcgl.bitnet