Atari Mega Archive 1/Atari Mega Archive - Volume 1.iso/lists/gem/l_0799/569

Internet Message Format | 1994-08-27 | 9.3 KB

Date: Tue, 21 Jun 94 15:29 CDT
From: [e-mail] (Evan K. Langlois)
To: [e-mail]
Subject: State of the Union Address
Precedence: bulk


Mostly to Tim Miller, re: A Proposal for Everyone

========================================================================
Now, I REALLY disagree.  I do not want to have to implement more than one
block-handling system just because this standard says so.
========================================================================

Uhh .. we went through this with dicontinuous blocks.  The APP-DEFS.SYS
file tells you which one the user prefers.  You decide the rest.

========================================================================
 I will use
these short-cuts insofar as they are applicable to my application, but I
don't want to be FORCED to implement more than one.
========================================================================

More than one?  More than one what?  One standard?   One short-cut key?
Just parse the file and set some function pointers and you never have
to worry about short-cuts anymore.  Simple.

As to being forced to implement a particular feature for which a short-cut
exists, we already told you that isn't necessary.   And the next proposal,
as in NeXT, which I adapted, it clearly states that you don't have to
implement it, but if you DO implement the function, you should use that
key (and allow the user to change it if APP_DEFS.SYS is present).  It also
says which keys can be used if your application wants to support some
other function for which no short-cuts exits.  Undefined is OK, last row
is OK, second row you should find something else first, and you can't
change the first set of keys (the user can, the programmer can't).

========================================================================
And the shortcuts file will also allow people to deviate so far from the
standard that the standard becomes a moot point.  We won't have a
standard... we'll have everyone's different opinion, which is no better
than the apps we have now.
========================================================================

Bullshit, what we have now is multiple apps that all use different keys.
The "shortcuts file" will allow the user to make every application act the
same exact way.  Now, of course, if you are using someone else computer,
you better bring you own short-cut file if you or the other person has
changed keys from the standard.  The idea of the standard is to give a
starting place, a recommendation, not to dictate how users should use
an app, and not to dictate how programmers must write an app.  But, if
a programmer wants the utmost flexibility, he can support every entry in
APP-DEFS.SYS.  Or, none if you like, although some users may complain
if you don't support the short-cuts at least.

Then again, many may not care about the APP_DEFS at all, since if no APP_DEFS
exist, use the standardized global settings.   Either way, there is a
VAST improvement.  What we have now, is everyone's different opinion,
namely yours, and since we can't satisfy everyone with one set of
keyboard short-cuts, the APP_DEFS.SYS file is the best offer available.

========================================================================
I don't like your idea of using a capital or lower-case letter for
shift.
========================================================================

It's not my idea, its a NeXT standard.  Again, the idea is a merging of
current list proposals, and a standard.

  1 - People are talking about changing the standard up-arrow and fuller
      symbols and ^ symbol.  ONe said the degree and filled degree, another
      said ' for control and " for shift-control, which would be OK except
      these symbols are two small (but it does save space in the menu).
  2 - The proposed ^? is supposed to imply SHIFT for keyboards that use
      SHIFT to access a question mark, and not for those that don't need
      shift.  This implies that the character itself can imply what
      modifier keys are required, based on you normally generate that key.
  3 - Its a NeXT standard.  In fact, you can go one step further and remove
      the ^ or fuller too, so the menu just shows a single letter and
      that will be sufficient.  I went half-way and removed SHIFT.  I
       do NOT feel that an up-arrow is obvious.

If enough people want to keep the up-arrow then fine, keep it.  The
major concepts are the key assignments, not the rest.  BTW, the menu
listings are half-next modified to be compatible with what is listed
in the ATARI Compedium (which gets the info from ATARI's developer style
guide).

========================================================================
from.  Again, I think redraw should be Ctrl-A.
========================================================================

Damn it.  I swapped ^a and ^A just for you because you already agreed that
you'd accept a swapping of select all and de-select all just because of
a weird block handling method.  You won't support the file definitions
for short-cuts or block methods, and you want to change existing standards
just for your benefit.  Not only do I think you are being unreasonable,
but are the best argument for APP_DEFS.SYS that I can make.  If you want
^a to be redraw, fine, then you change the global settings in YOUR file.

 1 -  ATARI COMPENDIUM says ^a is select all.
 2 -  The list decided ^A is de-select all.
 3 -  You complained and most people decided that these could be swapped.
 4 -  ^l is already standard for redrawing the screen on Unix systems.
      I use ^l on this Unix system to redraw my VI window, of course,
      I have to quote it since I also have ^l set in TOSWIN to redraw
      my local GEM screen.   So, locally, ^l redraws my screen, sending
      ^l redraws the terminal screen.

There are four reasons why I will not support redraw for ^a.  In fact,
I wanted to keep the standard, but everyone agreed (including you) that
swapping these would be OK.  We thought it would quiet you down a bit!
:-)  But now you want it for redraw.  Most people don't even want a redraw
which is why I put it in the bottom of the list.

========================================================================
I don't get it.  Ctrl-W has absolutely no excuse for meaning `Close', except
that some programs already use it.
========================================================================

If that isn't good enough, ATARI's developer docs say to use ^W.  Its
ATARIs standard.  It is also standard on the NeXT, and probably other
GUIs.  ^U is only a minor standard in Germany.  ALT-ESC is even worse.

========================================================================
Please remember that these keys are for ALL APPLICATIONS, not just those
with cursors.  For many applications, select all is a fairly safe, very
common opperation.  Take for example, a vector graphic drawing program:
very often, the user selects all and does a group modification such as
moving them, scaling them, etc.
========================================================================

This needed to be repeated in case someone missed it!!

========================================================================
Solving the Ctrl-A danger by making the key obscure is an ALL-APP change
since all apps should use the same key, but the solution of using a pop-up
in the almost-always-a-typo case of Ctrl-A + character key can be implemented
just for those apps for which Ctrl-A is dangerous.
========================================================================

Yes!  Again, perfect application of APP-DEFS.SYS and once again, making
keyboard short-cuts hard to use is counter-productive!!

It isn't much of a short-cut if its hard to use or remember!

========================================================================
What about Shift, Alt, and Alt-Shift?
========================================================================

ATARI Says that Alt is no good since some countries use ALT to form
special characters.  So we leave them.  Someone mentioned having
text-styles assigned to ALT-number_pad keys.  First, ALT-Numberpad
is for special characters, has been for awhile, especially on DOS
machines.  Second, text-styles are already standard in the Compendium
and making them different is counter-productive.  All this hassle so
that ^U, a stupid short-cut to begin with, can stay as Close Window
which is in opposition to ATARIs style guide is ludicrous.

Shift doesn't work because text apps need shift.  Obviously you can use
it in another type of APP, one where you are guaranteed not to
need a capital letter .. no graphics text, no non-modal dialogs that
may accept capitals, etc, etc.  A desktop doesn't usually use these,
so it could reserve normal and shifted keys for icon selection (type
first letter and jump to it, like Myriad, etc).

ALT-SHIFT - see above :-)

ALT-TAB and a of course, all the CONTROL-ALT fkeys, and many other CONTROL-ALT
keys that MiNT uses (like CONTRL-ALT-\) are reserved.

========================================================================
1. A method that allows applications to consult a global list of
user-defined key equivalents.

2. A standard (base file maybe) that comes as the default for the
above. Beyond that the user is free to confuse himself.
========================================================================

This another one of those .. in case someone didn't read this,
read it.  I agree here completely!!!