🚀 go-pugleaf

>>>>> In article <2ok1qb$nsq@lastactionhero.rs.itd.umich.edu>, Jim.Rees@umich.edu writes:
Jim> I won't be moving to lucid, because they made the unfortunate choice of
Jim> switching from C to a different source language.  I've tried emacs 19 but
Jim> haven't switched to it yet, partly because I haven't been able to build it
Jim> for all the machines I use, but partly because of this list of gripes.

  I have successfully switched from Epoch 4.2 to FSF Emacs 19. My answers
below are for Emacs 19.19, maybe some things have broken in newer versions...
I don't understand your problem with Lucid, though. I don't use it because
it can't work on an ordinary terminal when I'm away from my office
and I'm sick of maintaining two or three versions of my elisp.

Jim> cursor warps (and focus shifts) to Completions window, have to
Jim> manually move it back to the minibuf each time.

  True. The search-completions package is probably what you want 
(though I can't get the new versions to work for me anymore).

Jim> get a new Completions frame each time (investigate frame-pop package?)

  You'll get tired of this :-) 

Jim> Completions frame says "Click mouse-2 on a completion to select it," but
Jim> that's a lie.  Mouse 2 pastes like anywhere else.

  Untrue. Mouse 2 selects a completion.

Jim> how to suppress initial *scratch* frame?

  Try this:

	    (setq initial-frame-alist
		  '((height . 24) (width . 80) (minibuffer . nil)
		    (left . 495) (top . 156)))
	    (setq minibuffer-frame-alist
		  '((height . 2) (name . "minibuf") (width . 80)
		    (top . 116) (left . 495) (auto-raise . t)))
	    (setq default-frame-alist
		  '((menu-bar-lines . 1) (minibuffer . nil)))

Jim> minibuf autoraise doesn't work (wm bug?)

  Works for me with olvwm 3.3.

Jim> how to control minibuf geometry depending on display size?

  I haven't got around to this - I just move the windows when I
use a different sized display. You could use x-display-pixel-width
and x-display-pixel-height.

Jim> cursor is opaque sometimes

  That means that frame has the focus... :-) What do you really mean ?

Jim> better frame placement would be nice.

  Better, how ? Under olvwm frames are placed along the leading diagonal
like any other new X window - just what I want. If you set any X resources,
maybe they'll always pop up in the same place. How about setting
things in default-frame-alist ?

Jim> dragging mouse 1 shouldn't move point

  You mean you want secondary selections, rather than primaries ? They're
on M-mouse-1, M-mouse-2 etc. I just wish I could work out how to get 
OpenWindows style cut, copy and paste and drag'n'drop working. I do use

	(transient-mark-mode 1)
	(delete-selection-mode 1)

Jim> pasting from X is really strange!  funny delays.

  Untrue.

Jim> running gnuclient yields "cannot switch buffers in minibuf" if there are no
Jim> frames visible.  Or, if there are, it switches buffers in one of the visible
Jim> frames, leaving you with no view of whatever was in that frame before.

  Hmmm. I do get `Cannot switch buffers in a minibuf window', but for
other things (a periodic time-logger <Plod>, called from a wakeup process).

Jim> gnuclient doesn't return until emacs exits.
 
  Untrue. C-x # makes it return as usual.

  Hope this helps you make the switch. I'm glad I did.

--
|Kevin.Davidson@edinburgh.ac.uk +-+ Centre for Cognitive Science/HCRC,        |
|tkld@cogsci.ed.ac.uk           | | University of Edinburgh,                  |
|+44 31 650 4447         .oOo.  | | 2 Buccleuch Place, EH8 9LW.        .oOo.  |
`-------------------------------' `-------------------------------------------'

>>>>> "Jim" == Jim Rees <Jim.Rees@umich.edu> writes:

    Jim> I won't be moving to lucid, because they made the unfortunate
    Jim> choice of switching from C to a different source language.

Where did you ever get that idea?  Lucid Emacs (which is as close as
there comes to an "official" successor to Epoch) is written in the
same mix of languages that FSF 18, FSF 19, and Epoch are:  C and Emacs
Lisp.  It has never been any other way.


--
Chuck Thompson
Research Programmer
Department of Computer Science
University of Illinois @ Urbana-Champaign

In article <TKLD.94Apr15093643@scott.cogsci.ed.ac.uk>, tkld@cogsci.ed.ac.uk (Kevin Davidson) writes:

  Jim> Completions frame says "Click mouse-2 on a completion to select it," but
  Jim> that's a lie.  Mouse 2 pastes like anywhere else.
  
    Untrue. Mouse 2 selects a completion.

I'll be generous here and assume that what you mean is it works for you.  So
the interesting question is why it works for you but not for me.  I haven't
really investigated this yet.

  Jim> how to suppress initial *scratch* frame?
  
    Try this:

I did, and I still get a *scratch* frame when I start emacs.  In case it
isn't clear, what I want is the equivalent of inhibit-initial-screen-mapping.

  Jim> minibuf autoraise doesn't work (wm bug?)
  
    Works for me with olvwm 3.3.

I have verified that this is a problem with my window manager.

  Jim> cursor is opaque sometimes
  
    That means that frame has the focus... :-) What do you really mean ?

I'd like to get normal X behavior.  If the frame has the focus, I should
have a reverse video block cursor.  If it doesn't, I should have a box
cursor.  I should never have an opaque black cursor, and that's what I get
sometimes.  It's not consistent, and I don't know how to reproduce it.

  Jim> better frame placement would be nice.
  
    Better, how ?

Regular frames should never obscure the minibuf frame.  This isn't a killer,
just a wishlist item.

  Jim> dragging mouse 1 shouldn't move point
  
    You mean you want secondary selections, rather than primaries ?

I want the same thing epoch gives me, whatever that's called.  Dragging
mouse 1 should leave point unchanged, set and move the mark, and hilight the
region and copy it to the X cut buffer (or whatever they're calling it this
week).  I know I can do this, I just haven't figured out how yet.

  Jim> pasting from X is really strange!  funny delays.
  
    Untrue.

Again, I'll assume that what you mean is that it works for you.

  Jim> gnuclient doesn't return until emacs exits.
   
    Untrue. C-x # makes it return as usual.

I want it to return when I delete the frame.

In article <CTHOMP.94Apr15103523@hal.cs.uiuc.edu>, cthomp@hal.cs.uiuc.edu (Chuck Thompson) writes:

      Jim> I won't be moving to lucid, because they made the unfortunate
      Jim> choice of switching from C to a different source language.
  
  Where did you ever get that idea?  Lucid Emacs (which is as close as
  there comes to an "official" successor to Epoch) is written in the
  same mix of languages that FSF 18, FSF 19, and Epoch are:  C and Emacs
  Lisp.  It has never been any other way.

I'm afraid you're mistaken.  The lemacs README file, available at
ftp://lucid.com/pub/lemacs/README, says, "Note that building Lucid Emacs
requires an ANSI C compiler, like gcc, or Lucid's own (and the galaxy's
best) Lucid C/C++."  I've verified this by trying to build it with a K&R C
compiler.  Gnu emacs 19 and all its ancestors are written in K&R C.

The reason I say this is unfortunate is that ansi C is much less portable
than K&R C (which I just call "C" to avoid confusion).

>>>>> "Jim" == Jim Rees <Jim.Rees@umich.edu> writes:

    Jim> I'm afraid you're mistaken.  The lemacs README file,
    Jim> available at ftp://lucid.com/pub/lemacs/README, says, "Note
    Jim> that building Lucid Emacs requires an ANSI C compiler, like
    Jim> gcc, or Lucid's own (and the galaxy's best) Lucid C/C++."
    Jim> I've verified this by trying to build it with a K&R C
    Jim> compiler.  Gnu emacs 19 and all its ancestors are written in
    Jim> K&R C.

    Jim> The reason I say this is unfortunate is that ansi C is much
    Jim> less portable than K&R C (which I just call "C" to avoid
    Jim> confusion).

I think calling ANSI C a different source language than K&R C is a
poor choice of words.  As for ANSI C being less portable than K&R,
opinions differ.



				-Chuck

In article <CTHOMP.94Apr15174217@hal.cs.uiuc.edu>, cthomp@hal.cs.uiuc.edu (Chuck Thompson) writes:

  I think calling ANSI C a different source language than K&R C is a
  poor choice of words.

They're certainly not the same, and to me that makes them different.  How
would you have put it?

  As for ANSI C being less portable than K&R,
  opinions differ.

K&R C is accepted by every ansi C compiler I know of, but I don't know of
any K&R C compilers that will accept ansi C.  So a program written in K&R C
will run on more machines than one written in ansi C.  That's what I meant
by "portable."

>>>>> In article <2omfaj$4d@lastactionhero.rs.itd.umich.edu>,
Jim.Rees@umich.edu writes: Jim.Rees> In article
<TKLD.94Apr15093643@scott.cogsci.ed.ac.uk>, tkld@cogsci.ed.ac.uk (Kevin Davidson) writes: 
Jim> Completions frame says "Click mouse-2 on a completion to select it," but 
Jim> that's a lie.  Mouse 2 pasteslike anywhere else.

tkld> Untrue. Mouse 2 selects a completion.

Jim.Rees> I'll be generous here and assume that what you mean is it
Jim.Rees> works for you.  So the interesting question is why it works
Jim.Rees> for you but not for me.  I haven't really investigated this
Jim.Rees> yet.

  Ouch! Yes, I mean for me.
  Perhaps you have some Epoch resources or elisp still lying about.
Check (or remove) your X resources and start up emacs with the -q flag.

Jim.Rees> I did, and I still get a *scratch* frame when I start emacs.
Jim.Rees> In case it isn't clear, what I want is the equivalent of
Jim.Rees> inhibit-initial-screen-mapping.

  If you mean you want rid of the initial frame that opens while your
.emacs is being read, I want rid of it as well. The default-frame-alist
et. al. only seem to take effect after .emacs has been read. I suppose
this is to guarantee that there's a minibuffer if things go wrong.
From frame.el, it seems that the initial frame gets set up
as a before-init-hook, so there's no way to bypass it from .emacs.

Jim.Rees> I'd like to get normal X behavior.  If the frame has the
Jim.Rees> focus, I should have a reverse video block cursor.  If it
Jim.Rees> doesn't, I should have a box cursor.  I should never have an
Jim.Rees> opaque black cursor, and that's what I get sometimes.  It's
Jim.Rees> not consistent, and I don't know how to reproduce it.

  Works fine for me. Sounds like your window manager getting the focus 
wrong, or perhaps two dark colours both being shown as black on a mono
display.

Jim> better frame placement would be nice.
tkld> Better, how ?
Jim.Rees> Regular frames should never obscure the minibuf frame.  This
Jim.Rees> isn't a killer, just a wishlist item.

  When the frame is created, you mean ? Or at any time ? Isn't this 
a window manager issue ?

Jim.Rees> I want the same thing epoch gives me, whatever that's
Jim.Rees> called.  Dragging mouse 1 should leave point unchanged, set
Jim.Rees> and move the mark, and hilight the region and copy it to the
Jim.Rees> X cut buffer (or whatever they're calling it this week).  I
Jim.Rees> know I can do this, I just haven't figured out how yet.

  This is a secondary selection. M-mouse-1 is what you want.

Jim> gnuclient doesn't return until emacs exits.
tkld> C-x # makes it return as usual.
Jim.Rees> I want it to return when I delete the frame.

 Ah. I don't think there's a delete-frame-hook that could run server-edit.
I found that deleting frames (from the window manager) killed emacs
anyway and C-x 5 0 is an extra key stroke :-) Perhaps it should be the
other way round - the last call to server-edit could call delete-frame
if we're using a window system...

--
|Kevin.Davidson@edinburgh.ac.uk +-+ Centre for Cognitive Science/HCRC,        |
|tkld@cogsci.ed.ac.uk           | | University of Edinburgh,                  |
|+44 31 650 4447         .oOo.  | | 2 Buccleuch Place, EH8 9LW.        .oOo.  |
`-------------------------------' `-------------------------------------------'