Re: U2 glitches

[Robert Holmgren holmgren@xxxxxxxx]

** Reply to notes from Judith and Harry Wed, 14 Jun 2000 21:39:07 -0400

> Is commanding ABort enough to save and store this change?

Good God, Judith. I think that question deserves some kind of award,
worst of the year or something. How -- in plain language -- can aborting
anything save or store it? Don't you have a manual? RTFM!!!
Specifically, the sections on the commands ABORT, SAVE, and STORE.
It's an imposition to ask us to replace the manual, surely you realize
that. The answer is a resounding NO. SAving or SToring saves and/or
stores. ABorting "gets rid of" any changes you might have made (since the
last SAve), and removes the file from the window. If you know NOTHING else
about XyWrite, in order to be able to use it at even the most minimal
level, you need to know what these three commands do.

≫> I also tried reassigning the function NL to another key -- the .inf
≫> file correctly identified the key assignment and didn't crash.

≫ Possibly the NL assignment on the Numeric Keypad with Windows 98 and
≫ trying to identify it through ≪VA_NL≫ is a lethal combination. Maybe
≫ it isn't a good idea to assign a func to the Numeric-Keypad-Enter key.

> Especially for those of us who don't even have such a key.

True -- if it's true. On my laptop, 104 is the special "Fn" key plus the
"{ [" key. Use "scan" to see if you have a 104. Most modern
laptops are able to reproduce a fullsize keyboard via combos with
"Fn". Unless you have an old clunker, I'll bet your machine does too.

Here's another possibility. Don't the various Virii have a tab in
Properties that displays checkboxes which enable or lock out certain
combinations of keys in the application? Is [Ctrl-]Enter(28) or
[Ctrl-]Grey-Enter(104) one of those boxes? Worth looking. Maybe the
bloody thing is disabled in Properties, but XyWrite -- which assumes,
being a DOS app from the old days, that it has total/exclusive control
over the keyboard -- is ignoring a msg from the OS and trying to access it
anyway, and that's causing a Protection Violation.

However you shake it, the evidence is beginning to point at
a bug in the Virus98 keyboard handler. A standard call to ask DOS what
is assigned to a particular key is causing DOS to bomb. I'm sure that
a polite request to Microsoft would induce them to fix it right away.
Bill said so on TV last night, in one of the new touchy-feely "we've
been writing *GREAT* software for 25 years" ads. This, mind you, is
the same organization that stuck us with DOS for eight more years after
the 286 came out.

> But, is there a keyboard file around anywhere that explains in English
> what each key does with its various shift combinations?

Dunno. All I remember is a template card that came with the distribution
(in the box) that gave common assignments. You can dope it out, though,
using IDKEY, then looking up the meaning of function mnemonics in
the funcs table (the same table that crashes when you try to look up
anything, Catch-104).

> Will anyone object to my assigning U2  to unshifted F10?

Nope!

> I just realized that my uncombined factory Alts have been
> giving me all along the same results as unshifted
> F10--access to drop-down help boxes.

You only just noticed that they've been dropping down? And F10 does the
same thing? Then why are you putting the Helpkey on unshifted F10, if it
conflicts with a built-in W98 operation?

I do hope this spate of unfinished business is drawing to an end. I feel
somewhat responsible for what happens in INF, because I designed it and
also wrote the various U2 frames that trigger it -- so I linger. If it
crashes for Harry when he just commands "help" then he's done
something to corrupt U2 itself. It works perfectly on my one W98 machine
-- alas, I won't have access to that hardware again for a month or more,
so can't do any of this debugging myself.

Ciao




---------
Robert Holmgren
holmgren@xxxxxxxx
---------