HP3000-L Archives

March 2002, Week 1

HP3000-L@RAVEN.UTC.EDU

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
Wirt Atmar <[log in to unmask]>
Reply To:
Date:
Tue, 5 Mar 2002 23:15:30 EST
Content-Type:
text/plain
Parts/Attachments:
text/plain (40 lines)
A new version (Version 0.96b, March 5, 2002) has been put up at:

    http://aics-research.com/qcterm

This version should correct the vi problem that Mark Bixby reported (which
was in fact a bug we inadvertantly put into the last version of QCTerm). It
also corrects a long-standing and very subtle problem we've had with FORMSPEC
for some time. The section of terminal code that FORMSPEC calls is
particularly sensitive to your interpretation of the terminal's
specifications. We've fiddled with this section of code before, but it now
appears that we had fixed one particular problem in this section using two
wrongs to make a right. Ideally, we have that corrected now and a whole pile
of rights should now add up to a right, no matter how you exercise this code.

This newest version should also correct a particularly vexing problem that
just a very few people have been reporting. Some people have been reporting
that they repeatedly lose their configuration file and QCTerm seemingly
randomly reverts back to its default file, the primary symptom being that
QCTerm then telnets into our machine rather than theirs.

It is our best guess that this problem was being caused by an old Windows API
that we used in an internal routine called FileExists. This API dates from
the 16-bit, 8.3 file naming convention, Windows 3.x era. We use our
FileExists routine to determine if a configuration file exists. If it didn't,
we build one anew, using our defaults, just as if this were the first time
that QCTerm were turned on in your PC. It appears that in some versions of
the Windows operating system, the API that we were using doesn't always tell
us the truth, especially in more modern versions of the O/S.

To correct the problem, we've simply begun using a very recent file system
API that gives us similar information, thus the configuration file should now
appear to always be there, if it really is. If you're one of the people who
have suffered with this problem, please let us know if you continue to
experience the problem.

Wirt Atmar

* To join/leave the list, search archives, change list settings, *
* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *

ATOM RSS1 RSS2