hudgens@fsu.scri.fsu.edu (Jim Hudgens) (04/08/90)
Anyone ever noticed this? We're running Ultrix-32 V3.1B on a VAX. The use of double quotes and back quotes seems to cause octal 377's to be interspersed with regular output. I seem to recall seeing this mentioned somewhere (perhaps in this newsgroup). Anyone know why csh does this? Neither bash nor sh seems to do this, nor does any other UNIX machine I have tried. csh> echo "$$" `date` 2 4 9 8 6 Sat Apr 7 20:13:52 EDT 1990 csh> echo $$ `date` 24986 Sat Apr 7 20:17:50 EDT 1990 csh> echo "$$" 24986 csh> echo "$$" `date` | od -c 0000000 377 2 377 4 377 9 377 8 377 6 S a t A 0000020 p r 7 2 0 : 1 4 : 3 0 E D 0000040 T 1 9 9 0 \n 0000047 csh> echo $$ `date` | od -c 0000000 2 4 9 8 6 S a t A p r 7 0000020 2 0 : 1 7 : 0 4 E D T 1 9 9 0000040 0 \n 0000042 csh> cat << EoF > x "$$" `date` test EoF csh> od -c x 0000000 " 2 4 9 8 6 " 377 S 377 a 377 t 377 0000020 377 A 377 p 377 r 377 377 377 7 377 377 2 0000040 377 0 377 : 377 2 377 1 377 : 377 1 377 7 377 0000060 377 E 377 D 377 T 377 377 1 377 9 377 9 377 0 0000100 \n t e s t \n 0000106 Just curious. void *standard_disclaimer() { /* The above represent my personal views */ } Jim Hudgens Supercomputer Computations Research Institute hudgens@vsserv.scri.fsu.edu
grr@cbmvax.commodore.com (George Robbins) (04/08/90)
In article <642@fsu.scri.fsu.edu> hudgens@fsu.scri.fsu.edu (Jim Hudgens) writes: > > Anyone ever noticed this? We're running Ultrix-32 V3.1B > on a VAX. The use of double quotes and back quotes seems to > cause octal 377's to be interspersed with regular output. > I seem to recall seeing this mentioned somewhere (perhaps in this > newsgroup). Anyone know why csh does this? Neither bash nor > sh seems to do this, nor does any other UNIX machine I have tried. This is a feature that DEC introduced in trying to make the shell 8-bit transparent - they had to change the quoting mechanism from 8-th bit to "escape" characters. If you are running VAX stuff, use an either use /usr/new/csh or csh from one of the older releases, I don't remember if 3.0 is ok, or you have to go back to 2.2. For RISC stuff, you need the version from the 3.1C mandatory patch tape. -- George Robbins - now working for, uucp: {uunet|pyramid|rutgers}!cbmvax!grr but no way officially representing: domain: grr@cbmvax.commodore.com Commodore, Engineering Department phone: 215-431-9349 (only by moonlite)
hans@umd5.umd.edu (Hans Breitenlohner) (04/14/90)
In article <642@fsu.scri.fsu.edu> hudgens@fsu.scri.fsu.edu (Jim Hudgens) > describes some strange behavior of csh. I have fixed most of those (all except the last one -- i'll have to take another look at that) as well as related problems with Segmentation fault and Illegal instruction when strings get long (> 512 chars or thereabouts). If you have sources I can send you the fixes. If not, I'm afraid I can't help you.