lamy@cs.toronto.edu (Jean-Francois Lamy) (07/29/89)
rcp *copying from a remote machine* where the login shell is bash1.01 does not work; i.e. if the login shell is bash on host neat, then rcp neat:file . will transfer most of file, but will abort near the end with rcp: protocol screwup: expected control record The machine running bash (neat) is a Sun 4 running SunOS 4.0.3; the local machine is either a Sun or an SGI using the 4.3BSD rcp -- the same version of the protocol used in 4.0.3. Removing all bash configuration files (.bash*, .profile, .inputrc) does not help. Changing the shell on neat to /bin/tcsh solves the problem immediately. Has anyone else seen this? Is bash doing some oddity that would cause characters to be sent, confusing rcp? Is this fixed in a current release? Jean-Francois Lamy lamy@ai.utoronto.ca, uunet!ai.utoronto.ca!lamy AI Group, Department of Computer Science, University of Toronto, Canada M5S 1A4
lamy@ai.utoronto.ca (Jean-Francois Lamy) (07/29/89)
lamy@cs.toronto.edu (Jean-Francois Lamy) writes: >The machine running bash (neat) is a Sun 4 running SunOS 4.0.3; the local >machine is either a Sun or an SGI using the 4.3BSD rcp -- the same version >of the protocol used in 4.0.3. Same occurs running bash 1.02 + various patches; the person I got it from does not report the problem under a SunOS4.0 Sun 3, so its either a Sun 4ism, a 4.0.3ism, or a local rcp-ism. Mutter. Jean-Francois Lamy lamy@ai.utoronto.ca, uunet!ai.utoronto.ca!lamy AI Group, Department of Computer Science, University of Toronto, Canada M5S 1A4
bet@ORION.MC.DUKE.EDU (Bennett Todd) (07/30/89)
>>The machine running bash (neat) is a Sun 4 running SunOS 4.0.3; the local >>machine is either a Sun or an SGI using the 4.3BSD rcp -- the same version >>of the protocol used in 4.0.3. > >Same occurs running bash 1.02 + various patches; the person I got it from >does not report the problem under a SunOS4.0 Sun 3, so its either a Sun 4ism, >a 4.0.3ism, or a local rcp-ism. Mutter. Well, I hadn't noticed the problem before; I seem to use rcp less and less these days, as NFS and NNTP take over most of my regular data movement. When I do use rcp these days I am copying to the other machine, rather than from it -- and the problem doesn't occur then. Perhaps that is why this wasn't reported sooner. In any case, this problem shows up with Bash 1.01, between every permutation I tried (a sample, *not* exhaustive!) of Sun 3/50, 3/60, 3/160, 3/260, running SunOS 3.2, and 3.5. Freeyow! I just remembered that I have a couple of SunOS 4.0 machines to test with on the net as well, and went off to try it against and between them, and the problem went away. *THEN I TRIED IT BETWEEN 3.x MACHINES AND IT WAS GONE AS WELL!!!* What is going on??? This seems to be a really bizarre one; for a while it persists between all pairs of machines I can find; 5 minutes later I can't reproduce it between any pair! -Bennett bet@orion.mc.duke.edu
lamy@ai.utoronto.ca (Jean-Francois Lamy) (07/30/89)
I recompiled with -Bstatic -g and the problem went away; or maybe i'm just in the 5 minute time slot where it works :-( But boy is it reassuring to see someone else having seen the phenomenon. We need gnu.bash.ufo :-)
jjd@BBN.COM (James J Dempsey) (07/30/89)
>> lamy@cs.toronto.edu (Jean-Francois Lamy) writes: >> >> >The machine running bash (neat) is a Sun 4 running SunOS 4.0.3; the local >> >machine is either a Sun or an SGI using the 4.3BSD rcp -- the same version >> >of the protocol used in 4.0.3. >> >> Same occurs running bash 1.02 + various patches; the person I got it from >> does not report the problem under a SunOS4.0 Sun 3, so its either a >> Sun 4ism, a 4.0.3ism, or a local rcp-ism. Mutter. I have a Sun 3/60 running SunOS 4.0.3 and a Vax 8350 running Ultrix 3.0. I have bash 1.02 as my login shell on both machines. I was successful in rcp'ing /usr/dict words from the sun to the vax and from the vax to the sun. I can't reproduce this rcp problem. Perhaps that narrows it down to a Sun 4 problem? --Jim--
johani@nada.kth.se (Johan Ihren) (08/03/89)
In article <8907292025.AA13274@life.ai.mit.edu> jjd@BBN.COM (James J Dempsey) writes: >Perhaps that narrows it down to a Sun 4 problem? > > --Jim-- We've just noticed it on a VAXstation 3100... Johan Ihren