home *** CD-ROM | disk | FTP | other *** search
- Subject: Reader's Digest
- Date: Fri, 17 Jun 1994 18:02:23 +0200 (MDT)
- From: Annius.Groenink@cwi.nl (Annius Groenink)
- X-Face: "E3Hm]k]&:,OEP<{D2ixJf>-9[qOGLebNa0&cQyFL-a~)kTM3&&I"gFw=fJ]K%1IduGjOE`
- ZGu]&~G]QNGa7i/L!+#Xng<|+}HKYHj~5?fTInUEUh0$I1gBI7jrA!&_|e/pR1[cX:^xgJTPsrjA_9
- m8Zli[|.-u{]+c1(6C7mL*m`/_J\>.{4!:g
- Mime-Version: 1.0
- Precedence: bulk
-
-
-
- Ofir:
-
- og> I think you are being silly. Most comments to your annoying messages
- og> should have hinted at that. I will organise a vote just to keep you happy.
- og> I have taken you arguement seriously, I did say that I can see your point
- og> but that CTRL+A is so well established that it will be a bad idea to change
- og> it.
-
- You have done similar harm to another 'well-established' set of
- shortcuts in the more recent proposals, viz.
-
- Home/Shift Home, Contrl-Up/Down There has been only one person
- who suggested to change these,
- and no positive reactions as far
- as I can remember.
-
- As to Control + F debate, isn't it standard for 'Format All Disks
- Immediately'?
-
- Also, I think we should remove the 'paragraph' stuff, or formulate it
- in more general terms ('medium size unit' or something similar). It
- does not make any general sense (whereas I think things like a character
- map DO (as they also help for dialog windows).
-
- Finally, can anyone give me an example of frequent use of 'append to
- clipboard'? Is it worth a general shortcut? It should at least be
- generalized to 'ADD to clipboard' which also makes it understandable in
- the case of a drawing package (maintaining the parallel between short
- cuts + shift and selection + shift, namely ADDING).
-
- ----------
-
- Vincent Lefevre:
-
- vl> > I disagree... Ctrl-< is the same thing as ctrl-, This gets icky. I
- vl> """"""""""""""""""""""""""""""""""
- vl> Not on a French keyboard. So I think that ctrl should always be used with
- vl> a letter, a digit or a special key whose scancode is the same on every
- vl> keyboard.
-
- I think 'Control + ASCII'-value should stand for: CONTROL + any
- scancode whose lookup value in the key table (taking account of the
- status in the shift) key is 'ASCII-value'.
-
- So Control + < is, on some keyboards, contains SHIFT, and on some,
- it doesn't. But by checking the Keytbl() tables, applications do not
- need to know this.
-
- Here's the code I use:
-
- ; int ctrl_ascii
- ( int k_code
- , int k_shift
- ) { int lower = (char)k_code
-
- /*
- // If the keycode is already ASCII or extended set,
- // return.
- */
- ; if (!lower || lower >= 32) return 0
- ; else
- { KEYTAB* tab = Keytbl((void*)-1, (void*)-1, (void*)-1)
- ; int upper = k_code >> 8
- ; return
- ( (k_shift&(K_LSHIFT|K_RSHIFT))
- ? (char*)tab->shift
- : (char*)tab->unshift
- )[upper]
- ; } }
-
- ; int alt_ascii
- ( int k_code
- , int k_shift
- ) { int lower = (char)k_code
-
- /*
- // If the keycode is already ASCII or extended set,
- // return.
- */
- ; if (lower) return 0
- ; else
- { KEYTAB* tab = Keytbl((void*)-1, (void*)-1, (void*)-1)
- ; int upper = k_code >> 8
- ; return ((char*)tab->unshift)[upper]
- ; } }
-
- It is guaranteed to return the ASCII code produced by pressing the same
- combination WITHOUT control or alternate.
-
- Is there someone who has a good argument that this is NOT the only
- correct way to approach shortcuts?
-
- BTW if you include this in your sources, express your thanks to Edith
- in your document files...
-
- -----
-
- Tim Miller waffles on (TIM, are you READING this???)
-
- tm> You can't possibly believe that shift-backspace is a GOOD thing, can
- tm> you? Why do you want it in there? You can solve (kludge) the Ctrl-A
- tm> problem by making dramatic changes in philosophy from how many programs
- tm> function, but there's nothing you can do to fix shift-backspace. If
- tm> you're typing in a string of capital letters, make a space, and hit
- tm> backspace, your line is GONE! This is NOT a good design. Change it.
-
- I explained TWICE how the same sort of philosophical fix applies to
- Shift + Backspace. Tim, I am getting very tired of your
- single-mindedness. A program should simply set the UNDO buffer at a
- shift + backspace. Typing on a line should be one operation as far as
- UNDO is concerned, not each letter should be an operation (as that is
- simply no use). I sent a full Edith version to Tim's magazine, so
- let's assume he'll try Edith within a week and shut up about this...
-
- BTW, Tempus does the same kind of thing. (But is less secure)
-
- tm> If you think that ignoring me is going to get me to stop nagging, you're
- tm> wrong. I am the king of Atari fanatics
-
- We are not ignoring you. I get the feeling you are ignoring some of US.
-
- -----
-
- Ken Hollis:
-
- kh> Actually, Michael, Abandon can be taken two different ways. It can either
- kh> re-load the abandoned file (if it indeed was existent), or it can completely
- kh> delete it from memory. CTRL-H is usually the standard to this type of
- kh> thing, although I have seen CTRL-UNDO and SHIFT-UNDO to abandon a doc.
- kh> I propose that "CTRL-H" be used as a permanent standard to Abandon. Any
- kh> seconds?
-
- I prefer CTRL-UNDO, as control H is already 'Replace' (which is not the
- same as search next and replace), and CTRL-UNDO or SHIFT-UNDO makes
- more sense. ``Habandon''? Sorry for repeating myself.
-
- -----
-
- I was just thinking: if we add shortcuts for window operations to WINX,
- (having the closer box open a menu at a single click, and close at
- double click, like in Motif) using Control + Alt (hence reserved),
- we'd get a lot of new shortcuts available for other things.
-
- -----
-
- og> "Abandon" does not clearly convey this message. "Revert" would work
- og> better. "Abandon" to me means "Throw away what's in memory and close the
- og> window".
- og>
- og> AVOID AMBIGUITY!
-
- Abandon is a bad translation from German first made by Pure Software.
- It should be 'Restore' or 'Reload'.
-
- -----
-
- Evan:
-
- el> Can someone please check and see if the keyboard state variable (state of
- el> cntrl-alt-shifts) that is returned from evnt_multi() is valid when a
- el> keyboard occurs. It looks to me like it may only be valid when a MOUSE
- el> event occurs, since evnt_mouse and evnt_button contain this information
- el> but evnt_keybd() does not.
-
- Don't worry, it is valid. But a problem is that by the time the event
- arrives, the alternate key for example may be released, while the
- registered key has been put in the keyboard buffer while Alt was still
- pressed. There is no good solution to this problem (a mediocre solution
- is hacking into the key repeat to avoid buffering. I do this in Edith).
-
- -----
-
- og> Because most users will not bother changing the keyboard.cnf (or whatever
- og> it's called). Why does CTRL+W mean Close Window? It's not that obvious.
- og> The majority of programs I get to review in the UK use CTRL+U to close a
- og> window. Atari woke up to late with their style guide. The German
- og> developers had the sense to come up with their own standard long before we
- og> did. Give them credit and try to cooperate.
-
- The German devlprs came up with Control U when Apple had already
- introduced 'W' for close window. I blame them for using 'U'. It
- doesn't even mean something in German!
-
- -----
-
- Neil:
-
- nj> 2. Maybe Delete shouldn't be used on it's own for deleting a block? When a
- nj> block is highlighted you should still be able to type - which means still
- nj> being able to use Delete as well as other keys, so maybe Ctrl-Delete to
- nj> delete a block? (Haven't got the last proposal in front of me so I can't
- nj> check to see if I'm suggesting something that is already in place!)
-
- I absolutely agree. (Must use the word 'agree' as Ofir performs a
- 'grep' on this for his proposals. Shift + Control + X is good for
- Delete as Delete is the more harmful version of Cut. (In cheapo
- applications, delete means 'irrestorably discard')
-
- -----
-
- Bo Leuf:
-
- bl> I am however rather unhappy with Ctrl-G as the selection here, since
- bl> Crtl-G = Goto (line, page, whatever)
- bl> is a very common and well-established command in many applications which also
- bl> use Find&Replace. I'd like to instead suggest:
-
- Contrl + L is de facto for 'Goto Line'. Is it in the standard? <agree>
-
- -----
-
- Christian Nieber (to whom I owe much for Edith's philosophy)
-
- cn> CTRL-I italic, CTRL-B bold, CTRL-U underline are widely used Mac and NeXT
- cn> standard. CTRL-U of course can't be used because it is more important for
- cn> closing windows.
-
- Or... a good reason to use Control W for close window! That is ALSO
- widely used on Mac and NeXT.
-
- cn> Printer configuration and page layout are often different things. Therefore
- cn> I suggest (again from the NeXT guidelines)
- cn>
- cn> CTRL L - Page layout
-
- Hm. Too application specific to distinguish between these. Shift
- Control P should simply be 'Printer Configuration and/or Page Layout'.
- Control L is also often 'Goto Line'.
-
- Could people post the Mac and NeXT guidelines if they have them? That
- could prove useful information for the list! Warwick, is there a
- guideline document for the Motif codes you suggested?
-
- <one hour later I see that someone has already posted the NeXT
- guidelines>
-
- cn> Problematic when applications use modeless dialogs. Does Undo mean "undo
- cn> the last thing I did to my document" or "Cancel the dialog"? In papyrus
- cn> it's the first, since it makes sense to use a global Undo while a dialog is
- cn> top window, i.e. after changing parameters for a block without closing the
- cn> dialog, then deciding you didn't want that. How about F10 for Cancel? Undo
- cn> could be supported for campatibility as long as modeless dialogs are used.
-
- This makes sense. But as we already included into the standard that we
- should underline buttons with Alternate-codes, isn't UNDO for Cancel
- simply superfluous? It would be Alt-C in English, Alt-A in German
- software.
-
- -----
-
- Damien Jones
-
- dj> First, when it comes to mouse operations, there are two
- dj> operations that are the hardest and/or the most tiring to do; these
- dj> are the double-click and the click-drag.
-
- Only for those who still use the Atari mouse. Should we still support
- the Atari mouse?
-
- -----
-
- Michel Forget:
-
- mf> In this case, HiSoft is non-standard. All the programs that I have seen
- mf> which support this feature do call it "Abandon". i think Windows programs
- mf> also have this option. As for needing a keyboard equivalent, I would
- mf> say yes. I use the feature at least three times a day...
- ^^^^^
-
- O well, then surely you need a shortcut for that! :-) NOT! In Pure C I
- hit ^H by accident more often that I hit it on purpose.
-
- mf> I have to say that I like having separate keys for these functions; it
- mf> allows you to selectively replace without being forced to respond to
- mf> ten million alert boxes.
-
- The best way for selective replacement is to separate 'search' and
- 'replace' instead of having 'search' and 'search and replace'. You can
- then search one by one, and replace only those that need to be
- replaced. Having 'Search' and 'Replace' is also much more orthogonal.
-
- How can you know when selecting ^T that the next item found will be one
- you want to replace? Only by searching for it first... This must have
- been the reason for Atari to specify ^F/^G/^H in the MultiTOS
- guidelines. Very prudent, IMHO.
-
- mf> [>> In the case of block operations: Have discontinuous blocks in mind.
- mf> [>
- mf> [>I am! Actually, I've got one in my window right now!
- mf>
- mf> In Edith Pro? That is a feature I have not found yet. I'm still having
- mf> fun with the ability to mark a block that is not bound to the left
- mf> column of the screen (using the right mouse button). That comes in
- mf> handy every time I use Edith lately.
-
- A square block created with the right mouse button IS a discontinuous
- block! But it can also be created by shift + drag/double/triple click.
-
- mf> Why is everyone so down on MultiTOS? Sure, Mag!X is better, but MultiTOS
- mf> is quite a step forward compared to the last operating system Atari
- mf> released...
-
- All I care about is 'geschwindigkeit'. MultiTOS is slower than TOS 1.0
- on my TT! Maybe I am too fast, I don't know.
-
- mf> [>Ah! The same as the '!!' trick in C... Smart, but a bit cumbersome...
- mf>
- mf> What is the "!!" trick in C? (I live for these little tricks...)
-
- The '!!' operator makes 1 out of everything that is not a 0. The trick
- with Select/xorred colour makes white out of everything that is not
- black. So it is similar to using not-not. (Of course, for the
- Mathematicians among us, it only works in 'classical' C)
-
- mf> [>Not true. The latest Atari guidelines specify ^H for Replace (not
- mf> [>search AND replace, but just replace. Control + Undo would be a better
- mf> [>choice for this (undo all).
- mf>
- mf> I agree; Control + UNDO would be better. But will word processor/text
- mf> editor authors support it? (Will you make this a feature in Edith,
- mf> for example?)
-
- Edith already has that shortcut.
-
- -----
-
- Hm. Gotta stretch my legs for a while now...
-
- OK, I'm back. Now for the second 3000 lines...
-
- -----
-
- Ofir:
-
- og> Fine. I don't see why you are here then. A standard does not need to make
- og> sense. There is no way where every keyboard shortcut is going to make
- og> sense to everyone.
-
- I disagree. I think the proposal should have an explanation for each
- shortcut, explaining why it makes sense. One way of making sense
- *could* however be its being an accepted standard for years.
-
- -----
-
- Dan Wilga:
-
- dw> I have never understood the decision to make Control-U the equivalent for
- dw> closing the topmost window. This keypress is probably one of the most
- dw> commonly used, and yet it is impossible to type with one hand (unless you
- dw> use your thumb, like I sometimes have to.)
-
- I support the people who say that Control + U is bad as it cannot be
- typed with one left hand. This is a very good *additional* reason for
- using ^W and ^Tab for closing/cycling instead of the German stuff.
-
- Rick Flashman:
-
- rf> Each country *IS* different. In Spanish "Cierra Ventana" also has no "U",
- rf> and they would probably prefer something like Control-C or Control-V.
-
- Ah. But the German people will object: Fenster Schlie*en does have the U
- all over it! :-)
-
- -----
-
- Stefan Hintz:
-
- sh> >ALT-SPACE: Redraw all open windows
- sh> >ALT-KPENTER: Redraw current window
- sh> Why should this be nessesary? A bugfree programm does not need this.
- sh> And if you really need it, you can have it by fulling twice the topwindow=
- sh> or a similar action.
-
- A bugfree program *does* need it if it is running under MultiTOS! (at
- least in my experience). I think the AES (WINX!) should provide and alt
- + control reserved code (R for example) for Redraw Screen. Or, of
- course, MultiTOS could be released to the People so we can fix it.
-
-
- Ken Hollis:
-
- kh> I agree. If a program NEEDS window contents to be redrawn at user command,
- kh> then there is something wrong with the program.
-
- Give me any decent program, and I'll run it concurrently with a fine,
- AES-conform application which will trash the contents of its windows! In
- fact, MultiTOS will be fine for this purpose.
-
- -----
-
- Ofir:
-
- og> >I propose to have NKCC as standard keyboard handler routines.
- og>
- og> Sure, if someone can convert it to HiSoft BASIC, GFA BASIC...
-
-
- I don't know what NKCC is, but Ofir, do you honestly know any decent
- applications which were written in BASIC? High level BASICs can be nice
- for quick hacks, but that's really all, I should think.
-
- -----
-
- Warwick:
-
- Let's make that APP -underscore- DEFS instead of APP-DEFS.SYS. Having a
- minus sign in a file is faily non-standard on Atari's. Some file
- selectors don't allow typing one, for example.
-
- -----
-
- Timothy Miller:
-
- > > When the whole text is marked via mark all/CTRL A, a flag is set that the
- > > block is non-overwriteable. Delete works (after all there is still Undo),
- > > but when the user tries to overwrite it, the block is deselected and the
- typing
- > > ends up at the beginning of the document.
- > >
- > I had already suggested this as an alternative, but it can get to be
- > really damn annoying.
-
- So, tell me, how can that be annoying? Have you ever purposely
- replaced a whole document with a letter? When I want to start a new
- document, I usually open a new window for it.
-
- -----
-
- Ofir:
-
- og> Christian Nieber:
- og> > The following is an excerpt from the NeXTStep User Interface Guidelines
- og> >...
- og> > The guidelines place keyboard alternatives into three groups--reserved,
- og> > required, and recommended.
- og>
- og> I suggest we adopt the same divisions. This will cut down on contention
- og> in this list, as we all agree that `Control-O' is `Open', but not all
- og> agree that `Control-F' should only be `Find'.
-
-
- I suggest we take it entirely as a basis for our proposal. I can only
- see 3 or 4 of its shortcuts that we don't want to see in our proposal.
- It may bring back those who have left Atari for NeXT as soon as we have
- a PowerPC based TOS machine... Anyone writing the Atari emulator for
- the PowerMac yet?
-
- -----
-
- 451 lines. Sorry, folks.
-
- regards,
-
- Annius.
-
-
- --
- Annius V. Groenink | E-mail: avg@cwi.nl | Private & ZFC:
- CWI, Kruislaan 413 | Room: M233 | P.O. Box 12079
- 1098 SJ Amsterdam | Ext: 4077 | NL 1100 AB Amsterdam
- Netherland | Phone: +31 20 592 4077 | Phone: +31 20 695 9901
-