[comp.mail.elm] Elm command misfeature

jgd@csd4.milw.wisc.edu (John G Dobnick,EMS E380,4142295727,) (02/14/89)

Somewhere along the line (Elm 2.0 ?), the "Z" command was added to Elm.
When one types this command, it echos "Command: ZZ" and Elm then procedes
to exit and SAVE ALL MESSAGES in the "save file".

I have two objections to this.  One is that I ALWAYS want to be asked
if I want messages saved, since I prefer to leave unanswered things in the
system mailbox.  This command does not allow me that choice.

The second, and more serious, objection is that Elm lies to me!  Yes, LIES!!!
I type a *single* "Z", and Elm says I typed "ZZ".  (And, needless to
say, I don't like this.)

I would like to know the reason for the "Z" command.  (I note that the "vi"
editor uses "ZZ" to exit, but this requires *TWO* keystrokes.  If the
intent was that Elm emulate this exit command, its implementation is
faulty.)

Also, please note that it is very easy to *accidentally* hit a "Z" if
one's fingers miss the "shift" key (on at least some keyboards, "shift"
and "Z" are adjacent).  This is poor human engineering.  (This keeps "getting"
me, and I get tired of re-editing my system mailbox to restore these 
"lost" messages.)

I would like to see one of the following changes made.

a) Correct the implementation of this option to require *two* keystrokes.
   A quick examination of the code reveals no obvious difficulty here.

b) Make this command configurable.  A compile-time option is fine, but
   a run-time option, specifiable in the .elmrc file, may be better.

c) Remove the "Z" command.


All my opinion, but this "Z" thing *really* irritates me.

All that said, I *like* elm!  It's my favorite, and only, mailer.
--
-- 
John G Dobnick
Computing Services Division @ University of Wisconsin - Milwaukee
INTERNET: jgd@csd4.milw.wisc.edu

rob@pbhyf.PacBell.COM (Rob Bernardo) (02/14/89)

In article <963@csd4.milw.wisc.edu> jgd@csd4.milw.wisc.edu (John G Dobnick,EMS E380,4142295727,) writes:
+Somewhere along the line (Elm 2.0 ?), the "Z" command was added to Elm.
+When one types this command, it echos "Command: ZZ" and Elm then procedes
+to exit and SAVE ALL MESSAGES in the "save file".

'Z' keeps all unread messages in the current file, saves all
read messages in the save-file, and deletes all messages marked deleted.

+I have two objections to this.  One is that I ALWAYS want to be asked
+if I want messages saved, since I prefer to leave unanswered things in the
+system mailbox.  This command does not allow me that choice.

The 'q' command is still available to do this. 'Z' is an addition, not
a substitute for 'q'. There is an option in the elmrc file "ask", which
if "ON" while cause ELM to ask these questions when the 'q' command is
entered, otherwise ELM will presume the default.

+The second, and more serious, objection is that Elm lies to me!  Yes, LIES!!!
+I type a *single* "Z", and Elm says I typed "ZZ".  (And, needless to
+say, I don't like this.)

I'll presume there was a missing smiley or two in this paragraph as ELM
also lies about just every other command as well. You enter a mere 'r'
and ELM says you typed "Reply to message" >gasp<. I didn't type all that
in! >sheesh<

+I would like to know the reason for the "Z" command.

It is a quick way to exit ELM leaving/putting mail messages in certain
places that many users find natural.

+  (I note that the "vi"
+editor uses "ZZ" to exit, but this requires *TWO* keystrokes.  If the
+intent was that Elm emulate this exit command, its implementation is
+faulty.)

Vi is in no way a user-interface standard for ELM, although our experience
with vi and other programs may give us ideas for ELM from time to time.

+Also, please note that it is very easy to *accidentally* hit a "Z" if
+one's fingers miss the "shift" key (on at least some keyboards, "shift"
+and "Z" are adjacent).

Easy for you, difficult for others. But, in fact, your complaint has been
noted. Now if this clutziness were found to be more common, we ELMgineers
will  be glad to *improve* (notice I didn't use a loaded word like
you did: "correct") ELM.
-- 
Rob Bernardo, Pacific Bell UNIX/C Reusable Code Library
Email:     ...![backbone]!pacbell!pbhyf!rob   OR  rob@pbhyf.PacBell.COM
Office:    (415) 823-2417  Room 4E750A, San Ramon Valley Administrative Center
Residence: (415) 827-4301  R Bar JB, Concord, California

jgd@csd4.milw.wisc.edu (John G Dobnick,EMS E380,4142295727,) (02/15/89)

From article <4693@pbhyf.PacBell.COM>, by rob@pbhyf.PacBell.COM (Rob Bernardo):
> In article <963@csd4.milw.wisc.edu> jgd@csd4.milw.wisc.edu (John G Dobnick) writes:
> 
[Comments and complaints about the "Z" command, which I claim is a
mis-feature.  (Or at least that it's *implementation* could be improved.)
These comments were interpreted as being somewhat "intemperate" by Rob,
and he was right.  I was not in the best mood at the time, and I let my
annoyance get away from me.  Sorry about that -- I'll try to be less
inflamatory.]
> 
> +  (I note that the "vi"
> +editor uses "ZZ" to exit, but this requires *TWO* keystrokes.  If the
> +intent was that Elm emulate this exit command, its implementation is
> +faulty.)
> 
> Vi is in no way a user-interface standard for ELM, although our experience
> with vi and other programs may give us ideas for ELM from time to time.
> 

Let me just say that this is "non-intuitive behavior" in my book.  "ZZ"
sure *looks* like a vi command. 

May I suggest that such an "exit command", that *changes* the state of
the mail file in an irrevocable manner, depend on more than a *single*
keystroke.  Slipped fingers (quite easy given the adjacency of shift and Z
keys on a lot of keyboards), modem noise, and possibly other things, can
lead to annoying loss of messages.  Elm is very good about requesting
confirmation for other actions that lead to the irrevocable deletion of
messages.  This command should also do that.

> +Also, please note that it is very easy to *accidentally* hit a "Z" if
> +one's fingers miss the "shift" key (on at least some keyboards, "shift"
> +and "Z" are adjacent).
> 
> Easy for you, difficult for others. But, in fact, your complaint has been
> noted. Now if this clutziness were found to be more common, we ELMgineers
> will  be glad to *improve* (notice I didn't use a loaded word like
> you did: "correct") ELM.

Klutzy I may be, but is there *anyone* who doesn't make typos?

See above comments.  I also note that the Z command is apparently the only
exit command that does *not* require or request confirmation.  ("x" is not
a problem because it doesn't *change* anything.  And I *do* use the ask=on
option.)  I still claim that a *single character* command that can 
potentially "lose" messages without confirmation is not "user friendly".
This strikes me as contrary to the philosophy, as I understand it, of Elm.

I repeat my request for a change:  Require *two* "Z"s.  Or make this command
"configurable".  Or delete it entirely.

Personally, I think this command causes more problems that it cures.
One can configure the "q" command to perform (almost all of) the "Z"
functions after all, can't one?  So why have yet another command to do this?

A thought.  Since "q" can do *almost* everything that "Z" does, perhaps
a new elmrc option, say "alwayssave", can control the remaining function,
whether already read messages are saved in the "save" file.  This would 
allow the set of
	alwaysdelete = ON
	alwaysleave = ON
	alwayssave = ON
	ask = OFF
options to cause the "q" command to behave exactly as the "Z" currently does.
[And then I can turn ask=ON and have my security blanket back!  :-) ]

[Note to the Elm developers.  Please copy me on discussion of this item,
assuming there *is* any further discusstion of it.  I really would like
to know what the general feeling is on the subject.  Thanks.]


--
-- 
John G Dobnick
Computing Services Division @ University of Wisconsin - Milwaukee
INTERNET: jgd@csd4.milw.wisc.edu

syd@dsinc.UUCP (Syd Weinstein) (02/15/89)

In article <963@csd4.milw.wisc.edu> jgd@csd4.milw.wisc.edu (John G Dobnick,EMS E380,4142295727,) writes:
>Somewhere along the line (Elm 2.0 ?), the "Z" command was added to Elm.
>When one types this command, it echos "Command: ZZ" and Elm then procedes
>to exit and SAVE ALL MESSAGES in the "save file".
>
>I have two objections to this.  One is that I ALWAYS want to be asked
>if I want messages saved, since I prefer to leave unanswered things in the
>system mailbox.  This command does not allow me that choice.
As the author of that command, I'd like to comment.

Thats the purpose of the command, a quick exit from mail, like the
older mailx exit that saves all read mail into the mbox.  If you
don't want that, use q. 

>The second, and more serious, objection is that Elm lies to me!  Yes, LIES!!!
>I type a *single* "Z", and Elm says I typed "ZZ".  (And, needless to
>say, I don't like this.)
Elm doesn't lie, it outputs the command name.  The name of the commnand
is ZZ, if you type m, elm echos mail, did you type mail?  Elm isnt lying
but acknowledging the command.

>I would like to know the reason for the "Z" command.  (I note that the "vi"
>editor uses "ZZ" to exit, but this requires *TWO* keystrokes.  If the
>intent was that Elm emulate this exit command, its implementation is
>faulty.)
The name is a steal from vi, the reason is to have an exit like mailx
that didnt ask the questions all the time.  Note, no messages are lost,
just moved to the default mail save folder.

>Also, please note that it is very easy to *accidentally* hit a "Z" if
>one's fingers miss the "shift" key (on at least some keyboards, "shift"
>and "Z" are adjacent).  This is poor human engineering.  (This keeps "getting"
>me, and I get tired of re-editing my system mailbox to restore these 
>"lost" messages.)
Agreed, Z is next to shift, but I've havent heard of others complaining
about this.

Secondly, this feature is not in Elm2.1, so most of the net
won't even know what you are talking about.  It's a new feature in
Elm 2.2, so you have a development copy of the system.

>I would like to see one of the following changes made.
>
>a) Correct the implementation of this option to require *two* keystrokes.
>   A quick examination of the code reveals no obvious difficulty here.
No command is two keystrokes, this violates the commands all being
a single keystroke (submenus don't count, the command changes mode)

>b) Make this command configurable.  A compile-time option is fine, but
>   a run-time option, specifiable in the .elmrc file, may be better.
If you don't like the command, turn it off in your version.

>c) Remove the "Z" command.
If you don't like the command, turn it off in your version.

-- 
=====================================================================
Sydney S. Weinstein, CDP, CCP                   Elm Coordinator
Datacomp Systems, Inc.				Voice: (215) 947-9900
{allegra,bpa,vu-vlsi}!dsinc!syd	                FAX:   (215) 938-0235

geoff@desint.UUCP (Geoff Kuenning) (02/15/89)

In article <4693@pbhyf.PacBell.COM> rob@pbhyf.PacBell.COM (Rob Bernardo) writes:

> Easy for you, difficult for others. But, in fact, your complaint has been
> noted. Now if this clutziness were found to be more common, we ELMgineers
> will  be glad to *improve* (notice I didn't use a loaded word like
> you did: "correct") ELM.

Loaded words?  You don't consider "klutzy" loaded (as well as misspelled)?

In any case, there are many people who have a problem hitting keys
adjacent to the one they intended.  One only has to attend a classical
piano recital to realize that this is a common human failing, even among
persons who have trained themselves to the virtuoso level.

I agree with John Dobnick:  elm should not take unilateral and irreversible
action without asking for confirmation, regardless of the particular
command syntax.  Furthermore, I think it would be nice if the "keep mail
in a different file" people would give a bit more consideration to the
equally valid "keep all mail in the incoming mailbox" view.  In other words,
no mail-reading agent should "helpfully" move already-seen mail into a
different file without giving the user a way to suppress this behavior.

There.  Now were any of those words loaded?  :-)
-- 
	Geoff Kuenning   geoff@ITcorp.com   uunet!desint!geoff