lubkin@cs.rochester.edu (Saul Lubkin) (06/17/91)
support@isc.com wrote a response to a question of mine: >The job control in the c shell is broken. This is a known bug and it will >be fixed in the next release. If you try ctrl-Z in vi from a borne shell >you'll see it also works. The problem is in the cshell job control. >Sorry not to give you any new information, just confirmation. I don't think that that's entirely correct. When I run /bin/sh, as a login shell, and run vi under it, the resulting vi job cannot be stopped by "Cntrl-Z". Vi jobs under csh can be stopped, as expected. My original question was: >If you are running a job like >pg /etc/profile >under the C shell, with ISC 2.2, then "Control Z" (or whatever is the stop >character) won't stop it. Similarly for "more". (Vi jobs stop fine.) >I downloaded the PD program "less", compiled it, and found that "less" >also couldn't be stopped with "Control Z". >"Less" jobs under the Sun OS are stoppable with "Control Z". >Is there any fix? It seems to me that the problem may be in the terminal handler. It looks like, when running, e.g., pg -- or even vi in input mode -- the "Control-Z" doesn't get intercepted by the tty driver, but just passes through to the program. Perhaps some stty aetting might fix this? I should note that, I've compiled a POSIX job-controlled version of bash under ISC 2.2 that works quite well -- but it has the same problems as csh, when you try to stop a "pg", "more", or "less" job with "Control-Z" (or a "vi" job that is not in control mode.) Anyone have any ideas? Sincerely yours, Saul Lubkin
shj@login.dkuug.dk (Stig Jacobsen) (06/17/91)
lubkin@cs.rochester.edu (Saul Lubkin) writes: >I should note that, I've compiled a POSIX job-controlled version of bash >under ISC 2.2 that works quite well -- but it has the same problems as >csh, when you try to stop a "pg", "more", or "less" job with "Control-Z" >(or a "vi" job that is not in control mode.) I'm using tcsh (got the binary from the net) and ^Z does work with pg, more and less (also vi of course). The funny thing is that with our own applications the ^Z is passed right through to the application - I was under the impression that applications did not _need_ to do anything in particular to support job control? -- Stig Jacobsen Internet: shj@login.dkuug.dk FidoNet: 2:231/16 BIX: stigjac
steve@nuchat.sccsi.com (Steve Nuchia) (06/18/91)
In article <shj.677155825@dkuugin> shj@login.dkuug.dk (Stig Jacobsen) writes: >the application - I was under the impression that applications >did not _need_ to do anything in particular to support job >control? True if the application does not do anything destructive to the state of its parent. Like change the ioctl state of the controlling terminal. -- Steve Nuchia South Coast Computing Services (713) 964-2462 "Innocence is a splendid thing, only it has the misfortune not to keep very well and to be easily misled." --- Immanuel Kant, Groundwork of the Metaphysic of Morals