[comp.unix.microport] SysV/AT ksh emacs mode broken!

" Maynard) (07/09/88)

A followup to the ksh debate:
I've finally gotten some documentation for ksh (no, I can't pass it
along, as I'm not sure of its status), and have run into a problem:
If I do a 'set -o emacs', I can't edit the history file interactively.
It ignores the control characters to move around the line. If I 'set -o
vi', command editing works fine.
I'd really like emacs mode to work, since I've discovered that I can't
stand vi with emacs around. Help!


-- 
Jay Maynard, EMT-P, K5ZC...>splut!< | Never ascribe to malice that which can
uucp:       uunet!nuchat!           | adequately be explained by stupidity.
   hoptoad!academ!uhnix1!splut!jay  +----------------------------------------
{killer,bellcore}!tness1!           | Birthright Party '88: let's get spaced!

hedrick@aramis.rutgers.edu (Charles Hedrick) (07/09/88)

On the 286, there is not enough address space to include both emacs
and vi line editing modes.  (The makefile is set up for small model.
I haven't had the courage to try large model.)  The binary on the
uport bbs is compiled with vi.  That's why emacs mode didn't work for
you.  vi is bigger than emacs, so if you compile ksh with emacs
instead of vi, you also have room for the Berkeley-style job-control
stuff.  (If you use the patches I have developed, job control will
even work.)

richardh@killer.UUCP (Richard Hargrove) (07/09/88)

In article <581@splut.UUCP>, jay@splut.UUCP (Jay "you ignorant splut!" Maynard) writes:
> If I do a 'set -o emacs', I can't edit the history file interactively.
> It ignores the control characters to move around the line. If I 'set -o
> vi', command editing works fine.

It appears that emacs mode command line editing has not been implemented.
You shouldn't have to execute 'set -o emacs'. To quote my ksh docs:

	EDITOR
		If the value of this variable ends in emacs,
		gmacs, or vi and the VISUAL variable is not set,
		then the corresponding option (see Special Command
		set below) will be turned on.

Needless to say, this doesn't work either. The irony is that it DOES work
with the MKS Toolkit version of ksh which I use when I boot up MS-DOS.

Another ksh'ism that doesn't seem to work is using $(...) instead of `...`
for command substitution. Fortunately, `...` does work.

I'm looking forward to receiving the official ksh that is supposed to be
coming with the 2.3.1 upgrade. It's one of the primary reasons I purchased
the upgrade contract. I hope it implements all standard functionality.

richard hargrove
...!{ihnp4 | codas | cbosgd}!killer!richardh
--------------------------------------------

" Maynard) (07/10/88)

In article <Jul.9.04.30.18.1988.18068@aramis.rutgers.edu> hedrick@aramis.rutgers.edu (Charles Hedrick) writes:
>On the 286, there is not enough address space to include both emacs
>and vi line editing modes.  (The makefile is set up for small model.
>I haven't had the courage to try large model.)

Somehow, I'm not surprised.

>The binary on the
>uport bbs is compiled with vi.  That's why emacs mode didn't work for
>you.  vi is bigger than emacs, so if you compile ksh with emacs
>instead of vi, you also have room for the Berkeley-style job-control
>stuff.  (If you use the patches I have developed, job control will
>even work.)

Great...but what about those of us who don't have sources?

Maybe we can convince Microport to recompile it either in large model,
with both modes in place, or provide another version with emacs instead?

-- 
Jay Maynard, EMT-P, K5ZC...>splut!< | Never ascribe to malice that which can
uucp:       uunet!nuchat!           | adequately be explained by stupidity.
   hoptoad!academ!uhnix1!splut!jay  +----------------------------------------
{killer,bellcore}!tness1!           | Birthright Party '88: let's get spaced!

plocher@uport.UUCP (John Plocher) (07/11/88)

In article <581@splut.UUCP> jay@splut.UUCP (Jay "you ignorant splut!" Maynard) writes:
>
>A followup to the ksh debate:
>If I do a 'set -o emacs', I can't edit the history file interactively.

The 286 ksk was compiled without the emacs support.  ksh only works if
compiled small model and vi + emacs modes makes ksh too big.

If there is enough call for emacs mode I will do what I can to recompile
it ...

  -John Plocher

wtr@moss.ATT.COM (07/11/88)

In article <581@splut.UUCP> jay@splut.UUCP (Jay "you
ignorant splut!" Maynard) writes:

>I've finally gotten some documentation for ksh (no, I can't pass it 
>along, as I'm not sure of its status), and have run into a problem:
>If I do a 'set -o emacs', I can't edit the history file interactively.  
>It ignores the control characters to move around the line. If I 'set -o 
>vi', command editing works fine.  

>I'd really like emacs mode to work, since I've discovered that I can't 
>stand vi with emacs around. Help!

>Jay Maynard

jay,

this is being posted in advance of any uport technical
response.  so take this answer with a grain of salt ;-) 

the 'bug' in ksh is that the source will not compile under
the large flag.  this has to do with the assumption that the
size of an integer is the size of pointers, etc...

under small model, there is not enough room to fit both the
vi and emacs editting modules in.  so it's one or the other
at link time!  uport took the 'standard' unix editing mode,
vi.

this info is based on e-mail from david korn about compiling
ksh for a 286-based unix box.  i use my 286-box for
work-related  jobs and was trying to port ksh down to it
many moons ago.  (BTW: i gave up once uport posted theirs ) 

to anyone at uport: 

could you please post an alternate ksh with the emacs
command sequences enabled?  just call it ksh-e or something
like that.  it would be much appreciated by the user who
likes the emacs command line editting.  

hope this helps.
=====================================================================
Bill Rankin
Bell Labs, Whippany NJ
(201) 386-4154 (cornet 232)

email address:		...![ ihnp4 ulysses cbosgd allegra ]!moss!wtr
			...![ ihnp4 cbosgd akgua watmath  ]!clyde!wtr
=====================================================================

pdg@chinet.UUCP (Paul Guthrie) (07/12/88)

In article <367@uport.UUCP> @uport.UUCP (John Plocher) writes:
>The 286 ksk was compiled without the emacs support.  ksh only works if
>compiled small model and vi + emacs modes makes ksh too big.
>If there is enough call for emacs mode I will do what I can to recompile
>it ...
>  -John Plocher
What about the 386 version of ksh?  My release (3.0-L2.2) came also
with only vi compiled in.  There is really no reason for this on
a 386, but a subsequent call to uPort found there were no plans to
add emacs mode, even when I mentioned it might only be a compile.

Or am I out of date and this too is on the uPort BBS?

Paul Guthrie					I'm Pink
ihnp4!chinet!nsacray!paul			Therefor I'm Spam.
ihnp4!chinet!pdg

" Maynard) (07/12/88)

In article <367@uport.UUCP> @uport.UUCP (John Plocher) writes:
>The 286 ksk was compiled without the emacs support.  ksh only works if
>compiled small model and vi + emacs modes makes ksh too big.

Lemme guess...it was written in "all the world's a VAX" mode, and
assumes that pointers and ints can be freely interchanged.
I had hoped that AT&T wouldn't fall into THAT trap.

>If there is enough call for emacs mode I will do what I can to recompile
>it ...

Add me to whatever list there is...

-- 
Jay Maynard, EMT-P, K5ZC...>splut!< | Never ascribe to malice that which can
uucp:       uunet!nuchat!           | adequately be explained by stupidity.
   hoptoad!academ!uhnix1!splut!jay  +----------------------------------------
{killer,bellcore}!tness1!           | Birthright Party '88: let's get spaced!

fortin@zap.UUCP (Denis Fortin) (07/16/88)

In article <583@splut.UUCP> jay@splut.UUCP (Jay "you ignorant splut!" Maynard) writes:
  [about ksh]
> Maybe we can convince Microport to recompile it either in large model,
> with both modes in place, or provide another version with emacs instead?

Yes, I would also like to have the emacs version of ksh!  So uPort, 
please try to make `emacs ksh' available in some way...

PS. Of course, distributing two versions of ksh might not be feasible,
    but I would settle for receiving the vi version and the docs on floppy
    for the upgrade, and then downloading the emacs version from the BBS!
-- 
Denis Fortin, fortin@zap.uucp           
fortin%zap.uucp@larry.mcrcim.mcgill.edu	| CAE Electronics Ltd
uunet!utai!mcgill-vision!zap!fortin     | The opinions expressed above are mine

plocher@uport.UUCP (John Plocher) (07/23/88)

+---- Lots and Lots of you wrote:
| +---- and more and more of you wrote:
| | Maybe we can convince Microport to recompile ksh to support emas mode
| +----
| I would also like to have the emacs version of ksh
+----

ksh-v (vi) and ksh-e (emacs) are avaliable on the uport BBS in the
286-UPGR file area.  For the 386 there is a version in 386-UPGR called
simply ksh with both vi and emacs modes.

We decided not to compile in the job control changes at this time because
we didn't have time to test them out - I will try to do this testing as
soon as I can and get updated versions on the BBS.

BBS #s
    (408) 438-1680  (1200)
    (408) 438-8567  (Trailblazer)

   -John Plocher
    Customer Support Manager