home *** CD-ROM | disk | FTP | other *** search
- From: Mark.Baker@mettav.royle.org (Mark Baker)
- Date: 29 Jun 94 21:57:16
- Message-Id: <UUCP.773001152@mettav>
- Subject: Re: Dialog Box Proposal Part 1
- To: gem-list@world.std.com (gem-list@world.std.com)
- Precedence: bulk
-
- Craig:
- > But some applications DO allow select/cut/paste from inside an edittable
- > text field using the mouse (eg. the Myriad MiNT only desktop).
- > I think that the mouse cursor should be changed to reflect any object
- > it can effect (as it does under Xwindows/Motif).
-
- Motif does it for you, trying to do it yourself would need some fairly
- complicated code, particularly since the aes only lets you watch two
- rectangles.
-
- > The clipboard CPX's I've tried all crash on a falcon......
-
- The SDS one does; the Atari one seems to work on my falcon, although it doesn't
- have the features of the SDS one.
-
- Ofir:
- >> How about having always the 2 possibilities? e.g. active-drawing with
- >> the left button, and ghost-drag with the right button (or shift-left
- >> button).
- >
- > Due to the way GEM works, I think we should leave the right mouse button
- > alone since it lets you click on background windows.
-
- This is the convention the desktop uses; it isn't done automatically and it
- isn't easier to program than any other use of the right button. OTOH, the fact
- that the desktop and a few other programs use this convention mean we should
- stick to it.
-
- Christian:
- > Since we want to set standards for other things as well, we should better
- > put the big-cursor paradigm in the standard now, since it's well
- > established on the ATARI and in all other major GUI's. Those who still
- > want to implement the other sort of blocks (damn, isn't there a good name
- > for it? ) can make up their own shortcuts.
-
- The big-cursor paradigm is easier for using with the mouse, but not so good for
- keyboard users IMO. The standard should allow for apps to use either method and
- have suitable shortcuts for both. I don't mean all programs should support
- both, although the turbo-C++ editor for windows does very nicely.
-
- > Tempus), but everyone has a different set of shortcuts for the block, and
- > none of them understands ^X/^C/^V.
-
- I use LED all the time which does use ^X/^C/^V.
-
- > I feel misunderstood. More precisely, apps that only support undo and
- > redo can use Undo alternatively for both. Those with multi level-Undo can
- > use Undo for step-wise undo and Control or Shift - Undo for Step-wise
- > redo.
-
- You're right, I did misunderstand you - what you say here makes sense and I
- think the original proposal was a bit confusing.
-
- [Hide]
- >> What's wrong with just iconifying?
- > Doesn't work for MagiC (yet), and the menu bar will stay there.
-
- The menu bar's hardly obtrusive, since it will be replaced by another apps as
- soon as you try and use it.
-
- > Apple's Human Interface Guidelines suggest "Revert to Saved"
-
- Yes, that's fairly unambiguous. Actually Apple's guidelines are worth a read,
- although they are too comprehensive IMO as there is no way you could remember
- all the recommendations.
-
- Rainer
- >> Conclusion: including a list of internationally available key
- >> combinations in the shortcut standard is necessary IMO.
- >
- > Not to forget the key combinations of alternative (PC-like) keyboards.
-
- Which combinations aren't available on PC-like keyboards?
-
- > Deselect sounds better to me. With 'hide' I think of something like
- > hiding behind a tree. (I hope you understand what I'm trying to say)
-
- Yes, you kind of expect the text to disappear from the screen. Deselect is
- better, deselect all in object based apps and deselect block in text based
- ones.
-
- Timothy:
- >> I might have a go. Mine would use standard resource files, with the
- >> extended types used for special objects. I would use the same extended
- >> types as interface despite not having it :)
- >>
- > If you use standard resource files, how are you going to put in all the
- > extentions? Of course, you'll have to have a totally custom written
- > dialog drawing routine.
-
- Just use lots of progdefs. Use the extended types, and when the resource file
- is loaded you call a routine which checks this byte then makes the objects
- progdefs, saves the old ob_spec value and fills in the ob_spec fields depending
- what type it is. I'll only have to supply custom routines for the objects I
- replace, which will be providing CUA-style toggles first, with other new
- features such as speedo fonts in editable fields coming later. I'll try to make
- it so it can be used by a programmer with as little extra effort as possible.
-
- Ken:
- > Warwick, I think my method is MUCH easier to do. You don't have to
- > check for the rectangles, all you have to do is check to see if the
- > mouse enters an editable object or not...
-
- But this means responding to timer events and it slows the system down.
- Warwick's method has less effect on the speed.
-
- > Ofir, I thought the method I had written down was clear cut and easy to
- > understand. LEFT Shift and Delete: Clear to the LEFT of the cursor.
- > RIGHT shift and Delete: Clear to the RIGHT of the cursor. BOTH shifts
- > and Delete: Same as ESC. What's so hard about that? :)
-
- Nothing hard about it, it's just totally different from anything else. Follow
- the existing standard. Also, IMO there is only one situation in which it is
- acceptable for the shift keys to be different, and that's turning the player in
- games.
-
-
-
-