Re: Autoskip bit in logon panels (Was APAR OA16111, BlueZone,

2006-09-07 Thread Shmuel Metz (Seymour J.)
In
[EMAIL PROTECTED],
on 09/06/2006
   at 12:13 AM, Alan Altmark [EMAIL PROTECTED] said:

VM allows the sysprog to set the line editing defaults.  The most 
reasonable are: CHARDEL OFF LINEND # ESCAPE  LINEDEL OFF.

Is there any reaon not to use Field Mark for LINEND?
 
-- 
 Shmuel (Seymour J.) Metz, SysProg and JOAT
 ISO position; see http://patriot.net/~shmuel/resume/brief.html 
We don't care. We don't have to care, we're Congress.
(S877: The Shut up and Eat Your spam act of 2003)

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html


Re: Autoskip bit in logon panels (Was APAR OA16111, BlueZone,

2006-09-07 Thread Shmuel Metz (Seymour J.)
In [EMAIL PROTECTED], on 09/06/2006
   at 01:22 AM, Paul Gilmartin [EMAIL PROTECTED] said:

Why?  How?

Command stacking. I assume that was what Alan had in mind as well.

People today are quite accustomed to graphic workstations where no
displayable character is usurped for editing functions.

ISPF usurps a character, and I find it useful. For that matter, so
do various shells.

I customarily just SET LINEDIT OFF.

De gustibus non disputandem est.
 
-- 
 Shmuel (Seymour J.) Metz, SysProg and JOAT
 ISO position; see http://patriot.net/~shmuel/resume/brief.html 
We don't care. We don't have to care, we're Congress.
(S877: The Shut up and Eat Your spam act of 2003)

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html


Re: Autoskip bit in logon panels (Was APAR OA16111, BlueZone,

2006-09-07 Thread Paul Gilmartin
In a recent note, Shmuel Metz (Seymour J.) said:

 Date: Thu, 7 Sep 2006 11:18:35 -0300
 
 Is there any reaon not to use Field Mark for LINEND?
 
I'd love to.  Depending on terminal emulator, keyboard, and map.

I tried to do the same for ISPF, but it wouldn't let me.  Apparently
ISPF has a set of characters it considers acceptable candidates
for LINEND, not including Field Mark.  IIRC, I edited my PROFILE
data set and set LINEND to Field Mark regardless.  It may or
may not work, but at least I recovered use of semicolon without
sacrificing another of the valuable characters in the list.

And I found an old CMS PROFILE that I used for CMS with a serial
terminal, perhaps in the age of 300 baud modems.  In it, I had
set all the line editing characters to CTRL-modified keys.  CMS
is far more tolerant than ISPF -- if the programmer wants to
set his line delete character to SPACE, it's his gun and his foot.

-- gil
-- 
StorageTek
INFORMATION made POWERFUL

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html


Re: Autoskip bit in logon panels (Was APAR OA16111, BlueZone,

2006-09-07 Thread Alan Altmark
On Thursday, 09/07/2006 at 11:18 ZW3, Shmuel Metz (Seymour J.) 
[EMAIL PROTECTED] wrote:
 VM allows the sysprog to set the line editing defaults.  The most
 reasonable are: CHARDEL OFF LINEND # ESCAPE  LINEDEL OFF.
 
 Is there any reaon not to use Field Mark for LINEND?

Because my fingers refuse to push the Field Mark key after 25 years 
pushing # ?  :-)

That is an excellent idea (works, too!).

Alan Altmark
z/VM Development
IBM Endicott

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html


Re: Autoskip bit in logon panels (Was APAR OA16111, BlueZone,

2006-09-06 Thread Paul Gilmartin
In a recent note, Alan Altmark said:

 Date: Wed, 6 Sep 2006 00:13:24 -0400
 
- Line editing ought to be disabled by default on graphic
  terminals.
 
  For for full-screen applications, yes: not for line-mode input.
 
 By coincidence, the userid and password fields on the VM logo are being
 changed to stop line editing since there is no ambiguity in what is

At first, it sounds like a good idea.

 essentially a full-screen app.  But if you press enter and go back into
 line mode, then the default VM line editing characters (those specified or
 defaulted in SYSTEM CONFIG) are in effect, even before you log on.
 
But then you had to spoil it.

  The feature is useful only on keyboard-printer terminals.
 
  The feature is useful for line-mode input on a 3270. BTDTGTS.
 
Why?  How?

 VM allows the sysprog to set the line editing defaults.  The most
 reasonable are: CHARDEL OFF LINEND # ESCAPE  LINEDEL OFF.  Character and
 
'#' is not reasonable because it is a national character valid in
filenames in both VM and MVS.  It took an APAR against XEDIT to
get LOAD FOO#BAR ft fm to work because there's no way to
disable LINEND before doing a LOAD.  IIRC, the APAR caused XEDIT
on entry to use the setting of CP LINEND.

 line delete aren't necessary any more and the default of CHARDEL @ just
 
They were never necessary; they're no longer even useful.

 creates havoc with internet addressing.  Users are free to set their own
 values to override the defaults.
 
The most reasonable values are all OFF in order to minimize the
initial learning effort and astonishment factor for new users.  As
they gain sophistication, they are free to set their own values ...

People today are quite accustomed to graphic workstations where
no displayable character is usurped for editing functions.  VM
is simply an anachronism in this respect.

Likewise, ISPF's default of ';' as a command separator is an
irritant; I disabled it so long ago that I forget how.  I want
to be able to search for a target that contains ';', and IIRC,
';' was not treated as an ordinary character even in a quoted
string.

I customarily just SET LINEDIT OFF.  Every character on the keyboard
is too precious to squander on editing functions of relatively
little value on a graphic terminal.  if I could use CTRL characters
for the editing functions, I might feel differently.

-- gil
-- 
StorageTek
INFORMATION made POWERFUL

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html


Re: Autoskip bit in logon panels (Was APAR OA16111, BlueZone,

2006-09-06 Thread Alan Altmark
On Wednesday, 09/06/2006 at 01:22 CST, Paul Gilmartin 
[EMAIL PROTECTED] wrote:
  essentially a full-screen app.  But if you press enter and go back 
into
  line mode, then the default VM line editing characters (those 
specified or
  defaulted in SYSTEM CONFIG) are in effect, even before you log on.
  
 But then you had to spoil it.

:-)  You can't please everyone.  Today is not your day.  Tomorrow is not 
looking good, either!

 
   The feature is useful only on keyboard-printer terminals.
  
   The feature is useful for line-mode input on a 3270. BTDTGTS.
 
 Why?  How?
 
  VM allows the sysprog to set the line editing defaults.  The most
  reasonable are: CHARDEL OFF LINEND # ESCAPE  LINEDEL OFF.  Character 
and
 
 '#' is not reasonable because it is a national character valid in
 filenames in both VM and MVS.  It took an APAR against XEDIT to
 get LOAD FOO#BAR ft fm to work because there's no way to
 disable LINEND before doing a LOAD.  IIRC, the APAR caused XEDIT
 on entry to use the setting of CP LINEND.

My definition of reasonable obviously differs from yours.  :-)  My view 
is from the compatibility perspective.  It is ALSO reasonable, as you note 
later, to turn LINEND OFF and ESCAPE OFF if you do not have a user 
population that would be upset to find a change in behavior.  For new VM 
installations, OFF is fine.  For me, I have to have LINEND turned on. 
XAUTOLOG user SYNCH#SET SECUSER user * is just too valuable to me, and 
my exec to handle it for me isn't on every system I use.  (sigh)

  line delete aren't necessary any more and the default of CHARDEL @ 
just
 
 They were never necessary; they're no longer even useful.

Eh? They were useful in the days of typewriter terminals and teletypes, 
when a physical backspace would have just obscured whatever you 
subsequently typed. There was no white-out on a 3215!  But I agree that 
once the CRT came into being and we all learned to use it, CHARDEL and 
LINEDEL went the way of the dodo  Maybe the next time we think about 
re-versioning z/VM we'll consider changing those two to OFF.

But I think LINEND OFF and ESCAPE OFF would cause a riot.  If you can 
convince the folks on the IBMVM mailing list  ;-)

Alan Altmark
z/VM Development
IBM Endicott

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html


Re: Autoskip bit in logon panels (Was APAR OA16111, BlueZone,

2006-09-06 Thread Gerhard Postpischil

Alan Altmark wrote:
once the CRT came into being and we all learned to use it, CHARDEL and 
LINEDEL went the way of the dodo  Maybe the next time we think about 
re-versioning z/VM we'll consider changing those two to OFF.


But they didn't vanish quietly. The console task maintained printable 
CHARDEL values long past the time when both the line mode and CRT 
consoles came with the backspace capability. We handled those by zapping 
the DCM backspace field to x'16' (backspace).


Gerhard Postpischil
Bradford, VT

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html


Re: Autoskip bit in logon panels (Was APAR OA16111, BlueZone,

2006-09-05 Thread Shmuel Metz (Seymour J.)
In [EMAIL PROTECTED], on 09/04/2006
   at 08:53 AM, Paul Gilmartin [EMAIL PROTECTED] said:

  Perhaps autoskip should operate in the reverse direction,

I'm not sure that is the best solution, but the problem is certainly a
nuisance.

o Of late, companies are requiring ever more complex passwords, with
  mixed-case, numerics, and special characters.  VM line editing
  usurps several special characters; to enter them literally, they
  must be prefixed with ''.  But the password field in VM full
  screen logon is only 8 characters wide, allowing no extra space
  for quoting characters.

Does VM line editing apply? I thought that was strictly for command
and line-mode input.

  - Line editing ought to be disabled by default on graphic
terminals.

For for full-screen applications, yes: not for line-mode input.

The feature is useful only on keyboard-printer terminals.

The feature is useful for line-mode input on a 3270. BTDTGTS.

-- 
 Shmuel (Seymour J.) Metz, SysProg and JOAT
 ISO position; see http://patriot.net/~shmuel/resume/brief.html 
We don't care. We don't have to care, we're Congress.
(S877: The Shut up and Eat Your spam act of 2003)

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html


Re: Autoskip bit in logon panels (Was APAR OA16111, BlueZone,

2006-09-05 Thread Alan Altmark
On Tuesday, 09/05/2006 at 10:11 ZW3, Shmuel Metz (Seymour J.) 
[EMAIL PROTECTED] wrote:

 o Of late, companies are requiring ever more complex passwords, with
   mixed-case, numerics, and special characters.  VM line editing
   usurps several special characters; to enter them literally, they
   must be prefixed with ''.  But the password field in VM full
   screen logon is only 8 characters wide, allowing no extra space
   for quoting characters.
 
 Does VM line editing apply? I thought that was strictly for command
 and line-mode input.

   - Line editing ought to be disabled by default on graphic
 terminals.
 
 For for full-screen applications, yes: not for line-mode input.

By coincidence, the userid and password fields on the VM logo are being 
changed to stop line editing since there is no ambiguity in what is 
essentially a full-screen app.  But if you press enter and go back into 
line mode, then the default VM line editing characters (those specified or 
defaulted in SYSTEM CONFIG) are in effect, even before you log on.
 
 The feature is useful only on keyboard-printer terminals.
 
 The feature is useful for line-mode input on a 3270. BTDTGTS.

VM allows the sysprog to set the line editing defaults.  The most 
reasonable are: CHARDEL OFF LINEND # ESCAPE  LINEDEL OFF.  Character and 
line delete aren't necessary any more and the default of CHARDEL @ just 
creates havoc with internet addressing.  Users are free to set their own 
values to override the defaults.

Alan Altmark
z/VM Development
IBM Endicott

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html


Autoskip bit in logon panels (Was APAR OA16111, BlueZone, and Vista (Was: 27x132?))

2006-09-04 Thread Chris Mason
Edward

There is (still I guess) an aspect of the VM logon panel, copied in the
NetView logon panel[1], which, to my mind indicated a lack of understanding
of human factors in the implementation of the 3270 data stream and its
effect on the operator experience.

When deciding whether or not to use the autoskip bit in the attribute
character which defined the end of an input field, my opinion is that it
should depend on whether the data to be entered necessarily fills the entire
input field or may not.

If the input data is guaranteed to fill the field, then the use of the
autoskip bit is appropriate. Ideally such a field will be part of a
sequence of such fields so that reaching for the tab key will not be
necessary within the whole sequence.

If the input data is *not* guaranteed to fill the field, then the use of the
autoskip bit is *not* appropriate. The operator will be assured that use
of the tab key will always be required after having keyed the current
character data. If such a field is part of a sequence of fields, it may even
be better ergonomics to require the use of the tab key for every one of the
fields - even if one or two meet the guaranteed to fill criterion.

The VM and NetView logon panels use the autoskip bit universally as if the
programmer thought neat (or cool) when reading up on the 3270 data
stream capabilities for the first time and never missed an opportunity to
use it - it's when you are irritated by such tricks that such evil thoughts
come to mind.

The irritations arose actually with NetView. In general I always used the
tab key after entering a userid and then my password. If the userid I
happened to be using filled the whole field - I guess it would happen to
have been eight characters long - I used the tab key as usual - I'd have
been concentrating on the keyboard as now since I'm a fairly fast - and
clumsy - two finger man - and the password would have appeared for the world
and his dog to see in the PROFILE field.

I believe something similar happened with the VM logon panel.

Now, of course, some will say, But surely you have only one userid and so
how can this ever be a problem; you will either need to use the tab key or
you will never need to use the tab key. Such people lack imagination. If
you are a systems programmer in charge of - I'll stick to NetView although I
believe the same would happen in VM - an automation system, you may well
need to log on as one of the automation operators in order to sort out a
problem and the automation userids could have a variety of lengths with some
being eight characters long.

Ideally there should have been an IBM corporate directive to impose the
principle I outlined above - especially where failure to apply the principle
could expose passwords. As it is it's as if there was a de facto opposite
corporate directive. And I wonder how many vendor products have copied VM
and NetView assuming that IBM always leads the way to best practice.

Chris Mason

[1] I can't recall whether or not the mistake was also present in the
TSO/E logon panel. It's easy to check.

- Original Message - 
From: Edward Jaffe [EMAIL PROTECTED]
Newsgroups: bit.listserv.ibm-main
To: IBM-MAIN@BAMA.UA.EDU
Sent: Sunday, 03 September, 2006 5:11 PM
Subject: Re: APAR OA16111, BlueZone, and Vista (Was: 27x132?)


...

 The VM guys did a *much* better job on their logon panel, which *was*
 official VM from the start. They obviously understood 3270 a whole lot
 better.

 -- 
 Edward E Jaffe
 Phoenix Software International, Inc
 5200 W Century Blvd, Suite 800
 Los Angeles, CA 90045
 310-338-0400 x318
 [EMAIL PROTECTED]
 http://www.phoenixsoftware.com/

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html


Re: Autoskip bit in logon panels (Was APAR OA16111, BlueZone,

2006-09-04 Thread Paul Gilmartin
In a recent note, Chris Mason said:

 Date: Mon, 4 Sep 2006 11:50:21 +0200
 
 If the input data is guaranteed to fill the field, then the use of the
 autoskip bit is appropriate. Ideally such a field will be part of a
 sequence of such fields so that reaching for the tab key will not be
 necessary within the whole sequence.
 
 If the input data is *not* guaranteed to fill the field, then the use of the
 autoskip bit is *not* appropriate. The operator will be assured that use
 of the tab key will always be required after having keyed the current
 character data. If such a field is part of a sequence of fields, it may even
 be better ergonomics to require the use of the tab key for every one of the
 fields - even if one or two meet the guaranteed to fill criterion.
 
A couple things:

o If I recognize by kinesthesis that I've mistyped a character, my
  reflexive recovery is to press the backspace key and overtype the
  erroneous character with the correct one.  If the mistyped character
  was the last in an autoskip field, the same reflex operates, but
  with bad effect.  I backspace into a nonmodifiable field; the
  keyboard locks at the next keystroke; and the audible alert starts
  sounding.  I'd be better off without autoskip.

  Perhaps autoskip should operate in the reverse direction, also,
  so the backspace would place me at the character I wish to overtype.

  We've all been spoiled by the behavior of text entry fields in
  browsers which scroll the field when more characters than their
  displayed widths are typed.  But emulating 3270 behavior feels
  ever more like providing a buggy whip for the space shuttle.

o Of late, companies are requiring ever more complex passwords, with
  mixed-case, numerics, and special characters.  VM line editing
  usurps several special characters; to enter them literally, they
  must be prefixed with ''.  But the password field in VM full
  screen logon is only 8 characters wide, allowing no extra space
  for quoting characters.

  - Entry fields ought to be wider to accommodate quoting characters.

  - Line editing ought to be disabled by default on graphic terminals.
The feature is useful only on keyboard-printer terminals.  I
turn it off early in my PROFILE EXEC, but, of course, this is
ineffective at the logon screen.

-- gil
-- 
StorageTek
INFORMATION made POWERFUL

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html


Re: Autoskip bit in logon panels (Was APAR OA16111, BlueZone, and Vista (Was: 27x132?))

2006-09-04 Thread Edward Jaffe

Chris Mason wrote:

If the input data is guaranteed to fill the field, then the use of the
autoskip bit is appropriate. Ideally such a field will be part of a
sequence of such fields so that reaching for the tab key will not be
necessary within the whole sequence.
  


I agree with the above.

My comment about the superiority of the VM logon panel to that supplied 
by TSO was never intended as a comment on the usability of the panel, 
but rather on the superior support contained therein for 3270 displays 
larger than 24x80.


--
Edward E Jaffe
Phoenix Software International, Inc
5200 W Century Blvd, Suite 800
Los Angeles, CA 90045
310-338-0400 x318
[EMAIL PROTECTED]
http://www.phoenixsoftware.com/

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html