home *** CD-ROM | disk | FTP | other *** search
- id m0ufGgx-0007uca; Sat, 13 Jul 96 20:13 MDT
- Sender: owner-executor
- Received: from ardi.com by ftp.ardi.com
- (Smail3.1.29.1 #3) id m0ufGgN-0007uVn; Sat, 13 Jul 96 20:13 MDT
- Path: sloth.swcp.com!usenet
- From: Clifford T. Matthews <ctm@ardi.com>
- Newsgroups: comp.emulators.mac.executor
- Subject: Re: E/DOS handling of extended keys
- Date: 13 Jul 1996 19:23:52 -0600
- Organization: ARDI
- Lines: 50
- Message-ID: <uf3f2vtyon.fsf@ftp.ardi.com>
- References: <4rvjvi$cec@dfw-ixnews9.ix.netcom.com>
- NNTP-Posting-Host: gw.ardi.com
- In-reply-to: David Ellsworth's message of Wed, 10 Jul 1996 06:48:25 GMT
- X-Newsreader: Gnus v5.0
- To: executor@ardi.com
- X-MailNews-Gateway: From newsgroup comp.emulators.mac.executor
- Sender: owner-executor@ardi.com
- Precedence: bulk
-
- >>>>> "David" == David Ellsworth <davidell@ix.netcom.com> writes:
- In article <4rvjvi$cec@dfw-ixnews9.ix.netcom.com> David Ellsworth <davidell@ix.netcom.com> writes:
-
-
- David> There is one bug that has always existed in E/DOS that has
- David> consistently annoyed me (a registered Executor
- David> user). (Well, actually there is another one -- if E/DOS is
- David> started in 50 row text mode, when it exits, you are left in
- David> 50 row mode, but the font is 8x16. But you already know
- David> that :).)
-
- Yes, that's a bug that we know about that will not be relevant when we
- have a native Win32 version.
-
- David> The Macintosh has two ways of dealing with the keyboard:
-
- David> 1) When a non-modifier key is pressed, released, or
- David> repeated, it generates an event, passing the keycode to the
- David> event handler. 2) It maintains a array containing an
- David> element for each key on the keyboard. This array includes
- David> modifiers (Shift, Ctrl, etc.). For each element, a nonzero
- David> value signals that the corresponding key is being held
- David> down.
-
- David> On the PC, extended key scancodes are prefixed by 0xE0. For
- David> example, the scancode for Left Alt is 0x38, while the
- David> scancode for Right Alt is 0xE0 0x38. In (1), E/DOS handles
- David> this fine. But in (2), it ignores occurences of 0xE0. If
- David> Left Ctrl is pressed and held, then Right Ctrl is pressed
- David> and released, E/DOS will forget that Left Ctrl is being
- David> held down. Even worse, Left and Right Alt are both
- David> translated to Command; there is no way to make the emulated
- David> Mac think you are holding down Option. This is significant
- David> in at least one program: Spin Doctor.
-
- David> To the people at ARDI: please consider fixing this in a
- David> future version of E/DOS. And thank you for an outstanding
- David> product!
-
- I just fixed it now, although I didn't test it with Spin Doctor (I
- tested it with Asymptote Demo). We were already doing a bunch of
- extended translations, but I guess I overlooked the fact that the
- right alternate uses them. This had been on my TODO list for a while,
- thanks to Gene DeClark.
-
- David> --- David Ellsworth davidells@aol.com
- David> davidell@ix.netcom.com
-
- --Cliff
- ctm@ardi.com
-
-