home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Atari Mega Archive 1
/
Atari Mega Archive - Volume 1.iso
/
lists
/
gem
/
l_0799
/
495
< prev
next >
Wrap
Internet Message Format
|
1994-08-27
|
4KB
From: Mark.Baker@mettav.royle.org (Mark Baker)
Date: 26 Jun 94 18:24:10
Message-Id: <UUCP.772743608@mettav>
Subject: Re: A Proposal For Everyone
To: gem-list@world.std.com (gem-list@world.std.com)
Precedence: bulk
Evan:
> OK, how about shift-tab for backtab, and ^tab for cycle windows? or
> is that bad too? TAB could stay as cycle object for object based apps
> and non-text apps. Shift-tab for reverse cycle (of objects in non-text
> apps) and shift ^ tab for reverse cycle windows?
That makes sense. Should ^tab cycle windows within the current app, or use
wind_get(WF_BOTTOM) as clicking on the title bar would? If the latter, how
would you implement shift ^tab on the Atari?
> Hmm .. I kinda like ^d because its easy (no shift keys) and finding a
> next isn't really the opposite of finding previous. If you look at the
> position of the keys, its really nice. ^f for find in the middle. The
> "next" key (position wise) is ^g for find next. And teh previous key is ^d
> for find previous. That makes ALOT of sense to me, and again its a NeXT
> standard. However, ATARI doesn't define find next OR find previous, so
> they aren't gonna back me up here :-)
This makes sense, but find previous is as near the opposite to find next as
you're going to get and shift-ctrl-g leaves ctrl-d free for something else,
perhaps deselect block?
> I'll make the changes the my end for the TAB differences. And remove ^e.
Don't remove ctrl-e, if it's what I think it is then it's very useful.
Ken:
> Sliders MUST be handled in a redraw-as-you-drag (or active-drawing)
> method in order to provide full functionality. Any sliders that are
> designed to display information (such as scaling percentages, etc) should
> be displayed either within the slider bar, or to the direct right or
> left of the slider track. Sliders that take time to draw information
> based on their position must be ghost-drag type.
Obviously active-draw sliders are not something that can be handled by
let-em-fly automatically and require the app. programmer to handle it.
Generally this all makes sense, using a solid button for active-draw and a
ghost button for ones that redraw afterwards.
> Hotkeys must be identified by either an underline, an inverted character,
Not particularly to do with the user interface, but a feature I would like to
see in let-em-fly would be a way to stop it assigning hotkeys to particular
objects, eg. items on a scrolling list (since these change as the list is
scrolled) Only assigning hotkeys to BUTTONs, not STRING, TEXT etc, would help.
> + SHIFT-CTRL : Insert text from clipboard at its current
> position, then move the counter up one. Press
> again to get the next previously entered string.
Do you mean the clipboard, or the history buffer, or will these be the same?
> - Delete : Delete character in front of cursor.
> + LSHIFT : Delete all text to left of cursor.
> + RSHIFT : Delete all text to right of cursor.
> + L&RSHIFT : Delete entire editable field. (Same as [ESC])
I don't like programs differentiating between the shift keys, since they are
marked indentically. How about shift-del for delete line right and shift-bs for
delete line left, as on Everest? Or use ctrl-shift-{del|bs} for these and keep
shift-del for delete line as normal.
> Instead of having the clipboard at a set place (like U:\CLIP or whatever)
> why not have it set in an environmental variable? I propose that the
> following four environmental variables be reserved for clipboard
> directory locations:
>
> CLIPBRD, SCRAPDIR -- These are standards to GEM
> TMP, TEMP -- These are standards to Unix/Linux
Surely you can use the AES scrp_xxx functions for this? The environment
variables should also be st, for use by TOS programs, but we're talking GEM
programs here.
Ofir:
>> - Insert : Toggle insert/overwrite mode (change cursor too.)
>> + SHIFT : Bring up a character table to select characters to
>> input into the editable field.
>
> I thought you were going to follow the current propsal Shift+CTRL+Z
I prefer shift-insert TBH, but I agree that the existing proposal should be
stuck to.