WGLP09@SLACVM.SLAC.STANFORD.EDU (02/20/91)
So, there's this bug in VLT 5.034 that snuck in when I added
tooltype support from the Workbench. If you run VLT from the CLI, ignore
this message, otherwise read on.
The bug is kind-of nasty, especially under 1.3. Some versions of
2.0 mask the bug entirely and that's how it was able to sneak by. What it
is is a number of reads from location zero, and maybe some other things.
It will most definitely show up if you run enforcer, but there is at least
one condition in which it can cause a spurious Read/Write Error requester
to pop up: spurious because there really isn't an error anywhere. Just
cancelling the requester will make things run along just fine.
Yet, at this time I cannot recommend running VLT from the Workbench.
I will upload a fixed version to BIX in the next week or so, and I'm sure
that version will show up on a BBS near you shortly after.
Meanwhile, if you know of any other bugs that might be fixed at
the same time, let me know.
Willy.WGLP09@SLACVM.SLAC.STANFORD.EDU (02/27/91)
Yes, the fifo handler really isn't completely stable (sorry Matt 8).
I'm sure Matt will get around to fixing it sooner or later, though.
Meanwhile, it seemed to be useful enough to leave the feature in for
this release of VLT.
Will VLT ever be residentable... Well, not so long ago I would have
said (and in fact did say) that it would be impossible. However, I now have
an alpha version ported to SAS/C that more or less runs okay. If I read the
docs correctly, I should be able to link with cres.o, and have VLT
automagically be residentable... I still need to read the docs on that,
though.... 8)
Willy.
----------
Willy Langeveld - Bitnet: WGLP09 @ SLACVM - BIX: langeveld