From: SMTP%"CMU-OpenVMS-IP@sacto.mp.usbr.gov" 25-APR-1995 14:11:12.87 To: EVERHART CC: Subj: Re: Compressed Slip driver and RLOGIN - NEW for CMU ... X-ListName: Discussion list for the CMUIP (aka CMU-TEK) TCP Product Warnings-To: <> Errors-To: owner-cmu-openvms-ip@sacto.mp.usbr.gov Sender: owner-cmu-openvms-ip@sacto.mp.usbr.gov Date: Tue, 25 Apr 1995 14:53:03 +0000 (GMT) From: don@rata.vuw.ac.nz (Don Stokes) Reply-To: CMU-OpenVMS-IP@sacto.mp.usbr.gov Subject: Re: Compressed Slip driver and RLOGIN - NEW for CMU ... To: CMU-OpenVMS-IP@DRYCAS.CLUB.CC.CMU.EDU Message-ID: <3nj2cf$mif@st-james.comp.vuw.ac.nz> Organization: Victoria University of Wellington Content-Transfer-Encoding: 7BIT X-Newsgroups: vmsnet.networks.tcp-ip.cmu-tek X-Gateway-Source-Info: USENET Lines: 46 NNTP-posting-host: rata.vuw.ac.nz In article <1995Apr24.193941.7449@bay.cc.kcl.ac.uk>, Andy Harper, KCL Systems Manager wrote: >Thanks to Don Stokes, a compressed SLIP driver and an RLOGIN/RSH are now >available for CMU users. > >I've added it to the collection on my ftp server: > > ftp://ftp.kcl.ac.uk/cmu-tcpip/csdrv.bck > ftp://ftp.kcl.ac.uk/cmu-tcpip/rlogin.bck Ah, there's nothing like releasing software for making bugs leap out end get you. New versions of both are up on toyvax.zl2tnm.gen.nz. The only change to CSDRV is in CMUIP-VMS.H where I've specified the include files' paths incorrectly. When I restored the relevant source saveset on my system I flattened it and used the flattened paths to find include files. This only affects those wanting to build fom source -- there are no actual code changes. To update your copy without sucking the whole saveset down, simply change all occurences of #include "/CMU_SRC/central/" to "#include /CMU_SRC/central/include/netinet/". There's a fix to RSH to deal with weirdness in command files relating to the use of async GETs. For example, the command file: $ rsh some-host some-simple-command $ write sys$output "Hello World" would actually abort silently after the rsh command. The fix is simply to turn off the ASY bit in the RAB before closing the "file". There is still a remaining related weirdness, and that's that after an RSH with input from the terminal which is terminated by the remote end, control T gets disabled. SET CONTROL=T gets it back. It's not terminal modes, it's some strangeness in the async $GETs and PPFs -- it's fine if the input "file" is closed sanely without an I/O outstanding. Could someone with a post-VMS V5.4-2 system try it out and tell me if the weirdness is still present? It may be a VMS bug. It looks like a VMS bug. I *hope* it's a VMS bug coz it's got me stumped. I'd be interested to know if the command file weirdness is still present in more recent versions of VMS too. -- Don Stokes, Network Manager, Victoria University of Wellington, New Zealand. don@vuw.ac.nz(work) don@zl2tnm.gen.nz(home) +64 4 495-5052 Fax+64 4 471-5386