u3369429@murdu.OZ.AU.UUCP (03/14/87)
I'm not impressed by the 'digest' format at all, but I take it that this is only a temporary feature. But some items have been published in digests #1-#3 which I would like to comment on. (BTW, why is this a moderated news group anyway?) >From: Arno Diehl <DIEHL%iravcl%germany.csnet@RELAY.CS.NET> >Subject: swing > >SWING crashed when you tried: > SET DEFAULT DUA0:[000000] ! to get all directories on that > RUN SWING > >The reason was LIB$FIND_FILE returning: DUA0:[000000]000000.DIR;1 >causing SWING to search for DUA0:[000000.000000]*.DIR; >This returned DUA0:[000000.000000]000000.DIR;1 and so on... > >A simple modification in procedure LOAD_NODES.FOR will handle this >situation. > [fix follows] Thank you very much. I will incorporate your code in my version. I got SWING quite stable by now, but I would like to continue working on it. I think I might re-publish it in about five weeks time in net.sources. I will probably distribute it then in shar-ed format, using D. Wecker's Ultrix/VMS/Amiga shar. The major advantage will be that every line starts with an ordinary character, rather than a string of blanks or tabs. ------------------------------ From: <BEN%TECHMAX.BITNET@wiscvm.wisc.edu> >Subject: MAJOR SWING problem >It seems that swing is a great program, except that users in the same root >directory can swing to each others directories and delete whole trees! This >is even more compunded by a problem of a user doing set defaulkt to syslib >or some other such area and deleting the whole structure. Isn't there some >owner prot/id routine in SWING? I see a problem here... SWING will only delete directory structures which you could delete from DCL anyway. It has the same privileges as the user running it. ------------------------------ >From: "Thomas_W._Taylor.WBST147"@Xerox.COM >Subject: CHECKSUM for SWING > >My copy of SWING came through fine (I think), however I had a lot of >trepidation. Please, please, please for the future: if such large >sources are sent, send along a checksum(s) so we can determine if >everything is as intended. The command is simple: > $CHECKSUM 'file' > $SHOW SYMBOL CHECKSUM$CHECKSUM >This will make a lot of us sleep a lot sounder. > Next time, I will distribute SWING in net.sources using D. Wecker's shar. The archive file will include a character count for every file. Maybe I will think about this CHECKSUM as well. ------------------------------ >From: welch%UMASS.BITNET@wiscvm.wisc.edu >Subject: Scanning logical name tables > >I have several projects in mind which could use a routine that returns >all the logical names in a given table. If anyone has such a routine, >or the data structure of the VMS V4.x logical name tables, I would be >glad to hear from you. > From DCL, see HELP SHOW LOGICAL . For high level languages, I don't know. But a procedure using SHOW LOGICAL can be built and it can re-direct the output to a file which the HLL program can read. ------------------------------ >From: <KND@DHDMPI5.BITNET> >Subject: Supress Error messages > >hi, >in my logout procedure I delete all .JOU files. But if there are none >I get an error message that there are no .JOU files. If you see the >text "error..." on the screen you look at it even it isn't a "real" >error. Is there a way to suppress the error message > >%DELETE-W-SEARCHFAIL, error searching for DRB2:[KND...]*.JOU;* > >Assigning SYS$ERROR or SYS$OUTPUT to NL: doesn't help neither a >SET MESSAGE /DELETE. Could anyone help? Of course SET MESSAGE/DELETE doesn't do what you want. Did you read HELP SET MESSAGE /DELETE ? To get to your problem: Did you try: Set Message/NoFacility/NoSeverity/NoIdentification/NoText ? The lexical function F$Environment("Message") can be used to retrieve the previous setting into a symbol and then you can restore it, e.g.: $ Old_Message=F$Environment("Message") $ Set Message/NoFacility/NoSeverity/NoIdentification/NoText $ Set Default SYS$LOGIN $ Delete [...]*.JOU;* $ Set Message 'Old_Message ------------------------------ >From: Richard Steinberger <STEINBERGER@SRI-KL.ARPA> >Subject: CONTROL-T and LOAD info > >In earlier versions of VMS (4.2 and before I think), typing CONTROL-T displayed >a "LOAD factor" which gave some idea of how busy the machine was. I don't recall that facility on any VMS. However, TOPS-20 had exactly that. ------------------------------ Michael Bednarek (u3369429@murdu.oz.au)