1. Fix an "undocumented feature" in XLOCK in which an attempt to set a wildcard lock on X,0 when you already had a lock on X,Y would appear to succeed even though it had not. This is just a variation of the general problem of XLOCK returning success but actually doing nothing when the requested lock conflicts with a lock already set by the caller. In the case of a normal non-wildcard lock, this may be harmless, but in the case of the wildcard lock, it is probably quite dangerous since it would lead the caller to think that the he now had a lock on all X,* when he did not. So even though the AMOS version of XLOCK has the same problem, we decided that there was no excuse not to return an error (MODE = your own job number) when this happens. Note that this was not a problem with the SBR=MXLOCK flag set, since it's purpose was exactly that--to warn you of an attempt to lock something you already had locked.
2. XLOCK.SBR Enhancement: "Lock Reservations". A new mode, 4, has been added which acts similarly to 0 (request a lock but don't wait) except that if it fails to get the requested lock, it leaves a "reservation" for it in the lock queue. The intent of this is to facilitate a request for a wildcard lock X,0 in situations where heavy multi-user locking activity for X,(1-n) may mean that the the request for X,0 is not granted for a very long time. The "reservation" lock is ignored by jobs that already have at least one lock set on the LOCK1 value X, but it prevents any new jobs from setting any locks on X,*. Thus any locking
24 July 2000: Version 4.6(739)
1. Implement VAR.SBR, XVAR.SBR, and EDIT.SBR for Pat LaFollette. EDIT requires two entries in MIAME.INI:
ALIAS=EDIT:INFLD SBR=INFLDEDIT
(Without the ALIAS, EDIT.SBR will not be recognized at all. Without the SBR flag, it will work but the exit codes will not be returned as it will assume the parameter layout is that of INVUE, which is very similar to EDIT but unfortunately disagrees on the matter of the return code parameter.)
2. (LINUX) Add support for the "non-curses" arrow keys in the Gnome Terminal. Note that with this support added, the Gnome Terminal appears to work much better than the KDE Terminal, since the latter does have translatable function key sequences, nor do the graphic characters work.
3. (WINDOWS) Fix a bug with the selection of the border color which was introduced along with the support for 16 colors.
4. (UNIX) A new miame.ini switch (OPTIONS=RAWTABS) causes TABs to be output to the screen "raw" (i.e. instead of being converted to spaces.) This feature is rumored to be instrumental in some new and totally cool (but as yet undisclosed) function in ZTERM 2001.
