knowledge-database (beta)

Current group: comp.sys.apollo

sharing the DM experience

sharing the DM experience  
John S. Yates, Jr.
 Re: sharing the DM experience  
Craig A. Roloff
 Re: sharing the DM experience  
Stan Dunajski
 Re: sharing the DM experience  
Jim Rees
 Re: sharing the DM experience  
John S. Yates, Jr.
 Re: sharing the DM experience  
john at yates-sheets.org
From:John S. Yates, Jr.
Subject:sharing the DM experience
Date:1 Dec 2004 11:57:37 -0800
Jim Rees's posting about a VNC server for Domain got me to thinking...

I am interested in motivating addition of some DM feature into Linux
and window managers like WMI (http://wmi.modprobe.de/). It would be
really helpful if there were a way to let folk see a DM
input/transcript pair live. Does anyone have a way to do this over
the internet (even if it requires prior arrangement)?

/john (erstwhile Apollo compiler developer)
From:Craig A. Roloff
Subject:Re: sharing the DM experience
Date:Thu, 2 Dec 2004 23:24:13 GMT
If you haven't experienced ceterm, give it a try. It's real close to DM.

http://168.158.26.29/ce/ce/ce.html

John S. Yates, Jr. wrote:
> Jim Rees's posting about a VNC server for Domain got me to thinking...
>
> I am interested in motivating addition of some DM feature into Linux
> and window managers like WMI (http://wmi.modprobe.de/). It would be
> really helpful if there were a way to let folk see a DM
> input/transcript pair live. Does anyone have a way to do this over
> the internet (even if it requires prior arrangement)?
>
> /john (erstwhile Apollo compiler developer)
From:Stan Dunajski
Subject:Re: sharing the DM experience
Date:Mon, 06 Dec 2004 23:32:03 -0500
For information on Ce/CeTerm :

http://www.umich.edu/~archive/apollo/ce_readme.txt

For more information and executables :

http://168.158.26.29/ce/ce/ce.html

Stan
From:Jim Rees
Subject:Re: sharing the DM experience
Date:Thu, 02 Dec 2004 11:21:24 EST
In article <31fb0b12.0412011157.66624f66@posting.google.com>, john@yates-sheets.org (John S. Yates, Jr.) writes:

I am interested in motivating addition of some DM feature into Linux
and window managers like WMI (http://wmi.modprobe.de/).

One of the DM features I miss is the separate input pad. There are some
shells that allow editing the input, but usually only a single line and with
a special builtin editor that's not the same as your normal editor. Emacs
has a shell mode that comes close but it doesn't properly interleave the
input and output, and is missing such obvious features as the "again" key.

At one time I started writing a pad-like X terminal program but didn't get
very far. The main problem is that the terminal can't find out when the
attached process has consumed the input.

Note this is a feature of the terminal program, not the window manager.

It would be
really helpful if there were a way to let folk see a DM
input/transcript pair live. Does anyone have a way to do this over
the internet (even if it requires prior arrangement)?

I guess my vnc server would do it if it worked. I've got keyboard input
working now but not screen updates. Anyone know if there is an ec or some
other way to find out when the screen has been modified? Right now I just
check every few seconds but that is not satisfactory.
From:John S. Yates, Jr.
Subject:Re: sharing the DM experience
Date:3 Dec 2004 08:34:18 -0800
rees@umich.edu (Jim Rees) wrote in message news:<2004Dec2.112124@dumaguete.citi.umich.edu>...
> In article <31fb0b12.0412011157.66624f66@posting.google.com>, john@yates-sheets.org (John S. Yates, Jr.) writes:
>
> I am interested in motivating addition of some DM feature into Linux
> and window managers like WMI (http://wmi.modprobe.de/).
>
> One of the DM features I miss is the separate input pad. There are some
> shells that allow editing the input, but usually only a single line and with
> a special builtin editor that's not the same as your normal editor. Emacs
> has a shell mode that comes close but it doesn't properly interleave the
> input and output, and is missing such obvious features as the "again" key.

This is the DM feature I most want to present. I believe that it is
hard from just a written description for those who have not
experienced the input/transcript pad pair to adequately appreciate the
utility.

>
> At one time I started writing a pad-like X terminal program but didn't get
> very far. The main problem is that the terminal can't find out when the
> attached process has consumed the input.
>
> Note this is a feature of the terminal program, not the window manager.

Exactly. That is what I want to address. I understand it likely
involves some kernel hacking.

>
> It would be
> really helpful if there were a way to let folk see a DM
> input/transcript pair live. Does anyone have a way to do this over
> the internet (even if it requires prior arrangement)?
>
> I guess my vnc server would do it if it worked. I've got keyboard input
> working now but not screen updates. Anyone know if there is an ec or some
> other way to find out when the screen has been modified? Right now I just
> check every few seconds but that is not satisfactory.

I cannot answer the ec question. Please let me know if you do get it
working. For the sake of a simple demo would cranking the delay down
to a fraction of a section work?

/john
From:john at yates-sheets.org
Subject:Re: sharing the DM experience
Date:28 Dec 2004 10:14:16 -0800
Jim Rees wrote:
> In article <31fb0b12.0412011157.66624f66@posting.google.com>,
> john@yates-sheets.org (John S. Yates, Jr.) writes:
>
[..SNIP..]
>
> At one time I started writing a pad-like X terminal program
> but didn't get very far. The main problem is that the terminal
> can't find out when the attached process has consumed the input.
>
> Note this is a feature of the terminal program, not the window
> manager.
>
[..SNIP..]

Part of what revived my interest in this project was the inotify
functionality being added to the Linux kernel:

http://kerneltrap.org/node/view/3847

Following a quick skim of the code I felt reasonably confident
that I could create the necessary notification when a process
blocks waiting for input on a given stream. My model was to have
the input pad supply one line of input for each notification of a
desire to read from the monitored stream.

The other event prompting my renewed interest is that Drew Adams
has done a lot of work to wean Emacs of its old window bias. The
result is strongly reminicent of the DM

http://www.emacswiki.org/cgi-bin/wiki/OneOnOneEmacs

So what is the prognosis for that VNC capability? It does not
have to be efficient. I just want to demonstrate true input pad
behavior and universal editting.

/john
   

Copyright © 2006 knowledge-database   -   All rights reserved