nor1675@dsac.dla.mil (Michael Figg) (01/13/90)
I've tried to post this a couple times, from two different systems, but it doesn't look like it has taken. Sorry if you have seen this before. I've had a couple of problems with running shell scripts under Csh v2.07. I don't use scripts often enough (although I can't seem to live without s:startup-sequence!) to have looked into the problem before. The latest was a message of something along the line of "missing k directive" when trying to execute the Lattice v5.04 patch against v5.02. I was able to get around this by using WShell, but am curious about what the problem is with Csh. I vaguely remember (now that I'm away from the machine) something about 'run execute'. Is this the problem and does one of the newer versions of Csh address this? What is the current version? A slightly related (maybe more than I realize) question. I also tried using the 'protect' command to set the script bit, but it seemed like under Csh it also had no effect and the 's' didn't show up when running 'list' as I expected.I didn't check the bit under WShell, but the script ran without using 'execute'. Any enlightenment on this would be welcome. Thanks. -- "Could we be the bellwether | Michael Figg DSAC-FSD of major societal shifts?" | DLA Systems Automation Center - Columbus,Oh mfigg@dsacg2.dsac.dla.mil
hull@hao.ucar.edu (Howard Hull) (01/15/90)
In Article <1643@dsac.dla.mil> Y'all said... >a message of something along the line of "missing k directive" when trying to >execute the Lattice v5.04 patch against v5.02. I was able to get around this >by using WShell, but am curious about what the problem is with Csh. I vaguely >remember (now that I'm away from the machine) something about 'run execute'. Is >this the problem and does one of the newer versions of Csh address this? It may be a problem under some circumstances, but in this case, it's not your problem. Your problem has been very recently addressed on the net as per message ID: $89110303365772@masnet.uucp$ date: 2 Nov 89 14:36:00 GMT from vic.rocha@canremote.uucp (VIC ROCHA) subject: re: Missing k directive BH> From: farhi@athena.mit.edu (Bill Hoston) BH> Orga: Massachusetts Institute of Technology BH> thing I recall doing to the script was modifying it to run dmouse BH> and adding ls3.1 to my list of resident commands. I didn't test to VR> Bill, check for an undefined default .bra and .ket pair such as a `run VR> <nil: >nil: df0:dmouse', could even be over two lines. Another way to VR> avoid the problem is to define .bra as something, ie: `.bra {', that way VR> it's unlikely to confuse `Execute's finding of any `<'s or `>'s!! and as per message ID: $585@medusa.informatik.uni-erlangen.de$ date: 20 Nov 89 13:51:37 GMT from mlelstv@immd4.informatik.uni-erlangen.de (Michael van Elst ) subject: re: Missing k directive - claudio@forty2.UUCP (Claudio Nieder) writes: CN> In article <1105@forty2.UUCP> I wrote: CN> Thanks to everybody who sent me mail, telling me that I probably have CN> a "<" somewherer in my Shell-Startup. After cleaning up my prompt CN> which contained that character everything worked well again. MVE> If you use EXECUTE, then parameters are parsed by the command and MVE> substituted in the script. Actually a second script is generated with MVE> those substitutes. MVE> If it is called from NEWSHELL or as a startup-sequence, you just feed MVE> it into the CLIs standard input without ever looking at parameters or MVE> meta-commands (.bra, .dot, .key, etc....) MVE> I use in my startup-sequence: MVE> EXECUTE shell-execute MVE> and my shell-execute file comes up with .BRA { since otherwise any MVE> io-redirections would cause a 'missing K-directive' error. MVE> Michael van Elst As pointed out by Claudio Nieder, the Dillon/Drew and Dillon/Drew/Borreo-Cesare shells use ">" in the default prompt, so that's probably where your problem is originating. >What is the current version? The current distributed version is 3.03A by Carlo Borreo & Cesare Dieni. They may have another version well under way by now (or they may not). The revs may be interpreted as meaning 'any cshell of the Dillon/Drew extraction with a version level of 2.xx is for AmigaDOS v1.2 and any with version level of 3.xx is for AmigaDOS v1.3 with ARP library support.' If you don't provide a home for arp.library, then you won't be able to use the v1.3 lineage of cshells. >A slightly related (maybe more than I realize) question. I also tried using the >'protect' command to set the script bit, but it seemed like under Csh it also >had no effect and the 's' didn't show up when running 'list' as I expected. I >didn't check the bit under WShell, but the script ran without using 'execute'. The versions of list intended for use under v1.2 AmigaDOS only show the 'rwed' protection bits. What this means is that even if you are using the v1.3 Protect command or the 'protect' command (lower case) from a shell that knows about the 'chsp' protection bits, you won't see the result when you invoke v1.2 list. > >Any enlightenment on this would be welcome. Thanks. { As cited by Dave Haynie (in DiskSalv v1.0 circa de 1985): "I gained nothing at all from Supreme Enlightenment, and for that very reason it is called Supreme Enlightenment." -Gotama Buddha } >"Could we be the bellwether | Michael Figg DSAC-FSD >of major societal shifts?" | DLA Systems Automation Center - Columbus,Oh >mfigg@dsacg2.dsac.dla.mil Interpreting something related by James Burke in "The Day the Earth Changed": Plate tektonics are nothing more than weather in what we thought was solid rock. In this concept, a volcano is evidence for the development of an occluded front. Howard Hull hull@ncar.ucar.edu