S211KENO@HTIKHT5.BITNET (Kees Noyens KUB-DRC) (12/03/86)
There are some new but undocumented "features" in VMS V4.4. The word "feature" is more and more misused by DEC, I'm beginning to hate that word. Here is a list of some bugs (I prefer this word) in V4.4 : o $ analyze/disk/repair sys$sysdevice: This will cause JOB_CONTROL to log an error on the console, complaining that the accounting-file can't be written. Therefore, a new accounting-file is created. Very confusing for operators and not convenient for system-managers who wish to switch their account logs at regular times. o start/queue and set queue commands remove block-limits from print queues. They just disappear, all other items (like defaults, mounted forms, etc.) are retained. o $ set process/priv=all $ mount/write <tape with different UIC> The tape is mounted, but SOFTWARE write locked. So privileges like SYSPRV,BYPASS,VOLPRO,LOG_IO,PHY_IO don't help. The only solution is: $ set uic to a system-uic $ mount/write <same tape> and you are able to write the tape. This is a very strange one and I'm not sure how this worked with V4.3 o the set file/acl/like command has changed its behaviour. If you have granted a user CONTROL access with an ACE on file F2, with set file/acl=(id=<user-uic>,access=READ+WRITE+EXECUTE+DELETE+CONTROL) F2 that user will not be able to do $ set file/acl/like=F1 F2 , which did work with V4.3. Now it ends up with a privilege violation. What happens is that first the entire ACL is removed from F2, and then the ACL can not be copied from F1 to F2 because CONTROL access is not available anymore. Very, very clumsy. A workaround is to use PROTECTED ACE's. o the /SHEET_FEED qualifier (DEFINE/FORM) doesn't work since V4.4. the printjob doesn't finish but remains in the printing state; the printqueue is left in a paused state. o be careful with the &-character, it will terminate your process if it is not followed by an existing symbol. (for example, just try $ & ) o Documentation error: Command line recall capability is NOT extended for VAXTPU or EDT However, this is stated in the V4.4 release notes page 2-1. I know VMS is changing rapidly. I accept minor inconveniences. I would even accept all the existing bugs if V4.4 had brought our users any REAL improvements like IF THEN ELSE ENDIF or WHILE DO OD in DCL command procedures instead of the new subroutine commands (which are not bug-free). ------------------------------------------------------- Kees Noyens S211KENO@HTIKHT5.BITNET Tilburg University The Netherlands