Jon Green wrote:
>
> Thomas Hundt wrote:
> >  > On all but isearch C-w can be currently used to clear the message line
> >  > so it is not easy for us to simply change this as it may hinder more
> >  > people than it helps.
> >
> > I like the ^W feature in isearch (now that I know it's there!) but agree
> > that there ought to be a way to clear the search text (^A ^K doesn't
> > work) since I do this all the time (by cancelling the search and
> > starting over) when I mistype something at the beginning of the search text.
> >
> > -Th
> >
>
> I must admit I did not know that ^W existed.

Ditto - ^W to me on cmdline is delete a word
> I also think
> that ^A (possibly ^P) that returned to an editable
> search-forward type command line which is resumed to
> isearch-forward by ^S would actually be quite nice.

big emacs/Xemacs does something like this - I found it more confusing
than helpful
in being able to determine what searching mode you were in and what
keystrokes toggled you back and fwd and did searches and ...

-- hops

Before I forget heres a patch for a change I made to make the binary
self identifying

I was fiddling with changes in 3 different versions of jasspa me
(2001, Jan and Dec 2002) and got into a right state of confusion trying
to determine
why changes done werent working ...
turned out the path found different versions than the one I thought it
was supposed to.

Trying to get it to tell me (on command line) using the std gnu switch
(-V)
or anything else was something of an exercise in frustration so I fixed
  (the latest src) to emit its version in response to the -V switch
  like all well behaved software..

i.e
# give me version of software ( mej0 is Jan 2002 .. maybe ??)
713 localhost:/home/hops> mej0 -V
mej0 Error: Unknown option -V
Option -h gives further help
714 localhost:/home/hops> mej0 -h
usage     : me [options] [files]

where options can be:-
   @<file> : Setup using <file>[.emf], default is me.emf
   -b      : Load next file as a binary file
   -c      : Continuation mode (last edit, must have history setup)
   -d      : Debug mode (for macro files)
   -h      : For this help page
   -k[key] : Load next file as a crypted file optionally giving the <key>
   +<n> or
   -l <n>  : Go to line <n> in the next given file
   -m <msg>: Post message <msg> to MicroEmacs server
   -n      : For not X window, uses termcap instead
   -o      : One MicroEmacs, use ME server if available
   -p      : Pipe stdin into *stdin*, when saved output to stdout
   -r      : Read-only, all buffers will be in view mode
   -s <s>  : Search for string <s> in the next given file
   -u <n>  : Set user name to <n> (sets $MENAME)
   -v <v=s>: Set variable <v> to string <s>
   -x      : Don't catch signals

# still no @#$%^ version info

715 localhost:/home/hops> mej -V
MicroEmacs 02 - Date 02/12/16 - linux
  mej -h
usage     : me [options] [files]

where options can be:-
   @<file> : Setup using <file>[.emf], default is me.emf
   -b      : Load next file as a binary file
   -c      : Continuation mode (last edit, must have history setup)
   -d      : Debug mode (for macro files)
   -h      : For this help page
   -k[key] : Load next file as a crypted file optionally giving the <key>
   +<n> or
   -l <n>  : Go to line <n> in the next given file
   -m <msg>: Post message <msg> to MicroEmacs server
   -n      : For no X window, use console instead
   -o      : One MicroEmacs, use ME server if available
   -p      : Pipe stdin into *stdin*, when saved output to stdout
   -r      : Read-only, all buffers will be in view mode
   -s <s>  : Search for string <s> in the next given file
   -u <n>  : Set user name to <n> (sets $MENAME)
   -v <v=s>: Set variable <v> to string <s>
   -V      : Display Version info and exit
   -x      : Don't catch signals
   -y      : Load next file as a reduced binary file

# arrh much better
# though arguably the full usage (-help) should aso show its version
info
--------

Heres the patch :
------------------

*** main.c.orig Tue Oct 15 10:50:52 2002
--- main.c      Fri Oct 10 03:07:02 2003
***************
*** 97,102 ****
--- 97,103 ----
   "  -s <s>  : Search for string <s> in the next given file\n"
   "  -u <n>  : Set user name to <n> (sets $MENAME)\n"
   "  -v <v=s>: Set variable <v> to string <s>\n"
+ "  -V      : Display Version info and exit\n"
   #ifdef _UNIX
   "  -x      : Don't catch signals\n"
   #endif
***************
*** 1456,1461 ****
--- 1457,1475 ----
                       }
                       break ;
                   }
+
+             case 'V':
+                 snprintf((char *)evalResult, sizeof(evalResult),
+                          "%s %s - Date %s - %s\n",
+ #ifndef _NANOEMACS
+                          ME_MICROEMACS_FULLNAME,
+ #else
+                          ME_NANOEMACS_FULLNAME,
+ #endif

+                         meVERSION, meDATE, meSYSTEM_NAME) ;
+                 mePrintHelpMessage(evalResult) ;
+                 meExit(0) ;
+
   #ifdef _UNIX
               case 'x':
                   sigcatch = 0 ;
***************
*** 1862,1867 ****
--- 1876,1882 ----
           else if (cc == 'C')
               break;                      /* Let the sucker go !! */
       }
+     return 0;
   }
   #endif

---

Any likelihood of switching from a date based version to a familiar
triplet
major-minor-patch ???

-- hops

*** main.c.orig Tue Oct 15 10:50:52 2002
--- main.c Fri Oct 10 03:07:02 2003
***************
*** 97,102 ****
--- 97,103 ----
   "  -s <s>  : Search for string <s> in the next given file\n"
   "  -u <n>  : Set user name to <n> (sets $MENAME)\n"
   "  -v <v=s>: Set variable <v> to string <s>\n"
+ "  -V      : Display Version info and exit\n"
   #ifdef _UNIX
   "  -x      : Don't catch signals\n"
   #endif
***************
*** 1456,1461 ****
--- 1457,1475 ----
                       }
                       break ;
                   }
+
+             case 'V':
+                 snprintf((char *)evalResult, sizeof(evalResult),
+                          "%s %s - Date %s - %s\n",
+ #ifndef _NANOEMACS
+                          ME_MICROEMACS_FULLNAME,
+ #else
+                          ME_NANOEMACS_FULLNAME,
+ #endif
+                         meVERSION, meDATE, meSYSTEM_NAME) ;
+                 mePrintHelpMessage(evalResult) ;
+                 meExit(0) ;
+
   #ifdef _UNIX
               case 'x':
                   sigcatch = 0 ;
***************
*** 1862,1867 ****
--- 1876,1882 ----
           else if (cc == 'C')
               break;                      /* Let the sucker go !! */
       }
+     return 0;
   }
   #endif

Mike's got a great idea there.  I've had the same problem myself, and
hope it stays in the implementation.

It might be nice if there was a way to get the non-server version of ME
to execute an arbitrary command (option -m).  Because then one might do
something like 'me -m print $version' and get the same information.

(I never use the client-server mode.  Not sure what it's for.  Save
memory?  But memory's cheap!)

-Th


Mike Hopkirk wrote:

> I was fiddling with changes in 3 different versions of jasspa me
> (2001, Jan and Dec 2002) and got into a right state of confusion trying
> to determine
> why changes done werent working ...

> Trying to get it to tell me (on command line) using the std gnu switch
> (-V)
> or anything else was something of an exercise in frustration so I fixed

> -----Original Message-----
> From: Thomas Hundt [mailto:thundt@...]
> Sent: 10 December 2003 08:26
> To: jasspa@yahoogroups.com
> Subject: Re: [jasspa] Determining jasspa-me version
>
>
> Mike's got a great idea there.  I've had the same problem myself, and
> hope it stays in the implementation.

I agree, I think the -V option should be added.

> It might be nice if there was a way to get the non-server
> version of ME
> to execute an arbitrary command (option -m).  Because then
> one might do
> something like 'me -m print $version' and get the same information.

I think the '@' option could do this (i.e. me @printversion) but windows console
etc makes this a pain.

> (I never use the client-server mode.  Not sure what it's for.  Save
> memory?  But memory's cheap!)

I am an abuser of ME, I typically have one session of ME running for weeks and
can end up with 300+ buffers and 500+ Mb of memory (its cheap!). I use the
server because I want all the buffers in one me session so I can easily swap
between them (I hate using a mouse - too slow).

Steve

>
> -Th
>
>
> Mike Hopkirk wrote:
>
> > I was fiddling with changes in 3 different versions of jasspa me
> > (2001, Jan and Dec 2002) and got into a right state of
> confusion trying
> > to determine
> > why changes done werent working ...
>
> > Trying to get it to tell me (on command line) using the std
> gnu switch
> > (-V)
> > or anything else was something of an exercise in
> frustration so I fixed
>
>
>
> ------------------------ Yahoo! Groups Sponsor
> ---------------------~-->
> Buy Ink Cartridges or Refill Kits for your HP, Epson, Canon or Lexmark
> Printer at MyInks.com. Free s/h on orders $50 or more to the
> US & Canada.
> http://www.c1tracking.com/l.asp?cid=5511
> http://us.click.yahoo.com/mOAaAA/3exGAA/qnsNAA/89EolB/TM
> --------------------------------------------------------------
> -------~->
>
> ______________________________________________________________
> ____________
>
> This is an unmoderated list. JASSPA is not responsible for
> the content of
> any material posted to this list.
>
> To unsubscribe, send a mail message to
>
>     mailto:jasspa-unsubscribe@yahoogroups.com
>
> or visit http://groups.yahoo.com/group/jasspa and
> modify your account settings manually.
>
>
>
>
> Your use of Yahoo! Groups is subject to
> http://docs.yahoo.com/info/terms/
>
>
>

Thanks for this submission, I will add it to the next release.

Steve

> -----Original Message-----
> From: Mike Hopkirk [mailto:hops@...]
> Sent: 09 December 2003 18:35
> To: jasspa@yahoogroups.com
> Subject: [jasspa] Determining jasspa-me version
>
>
> Before I forget heres a patch for a change I made to make the binary
> self identifying
>
> I was fiddling with changes in 3 different versions of jasspa me
> (2001, Jan and Dec 2002) and got into a right state of
> confusion trying
> to determine
> why changes done werent working ...
> turned out the path found different versions than the one I thought it
> was supposed to.
>
> Trying to get it to tell me (on command line) using the std gnu switch
> (-V)
> or anything else was something of an exercise in frustration
> so I fixed
>  (the latest src) to emit its version in response to the -V switch
>  like all well behaved software..
>
> i.e
> # give me version of software ( mej0 is Jan 2002 .. maybe ??)
> 713 localhost:/home/hops> mej0 -V
> mej0 Error: Unknown option -V
> Option -h gives further help
> 714 localhost:/home/hops> mej0 -h
> usage     : me [options] [files]
>
> where options can be:-
>   @<file> : Setup using <file>[.emf], default is me.emf
>   -b      : Load next file as a binary file
>   -c      : Continuation mode (last edit, must have history setup)
>   -d      : Debug mode (for macro files)
>   -h      : For this help page
>   -k[key] : Load next file as a crypted file optionally
> giving the <key>
>   +<n> or
>   -l <n>  : Go to line <n> in the next given file
>   -m <msg>: Post message <msg> to MicroEmacs server
>   -n      : For not X window, uses termcap instead
>   -o      : One MicroEmacs, use ME server if available
>   -p      : Pipe stdin into *stdin*, when saved output to stdout
>   -r      : Read-only, all buffers will be in view mode
>   -s <s>  : Search for string <s> in the next given file
>   -u <n>  : Set user name to <n> (sets $MENAME)
>   -v <v=s>: Set variable <v> to string <s>
>   -x      : Don't catch signals
>
> # still no @#$%^ version info
>
> 715 localhost:/home/hops> mej -V
> MicroEmacs 02 - Date 02/12/16 - linux
>  mej -h
> usage     : me [options] [files]
>
> where options can be:-
>   @<file> : Setup using <file>[.emf], default is me.emf
>   -b      : Load next file as a binary file
>   -c      : Continuation mode (last edit, must have history setup)
>   -d      : Debug mode (for macro files)
>   -h      : For this help page
>   -k[key] : Load next file as a crypted file optionally
> giving the <key>
>   +<n> or
>   -l <n>  : Go to line <n> in the next given file
>   -m <msg>: Post message <msg> to MicroEmacs server
>   -n      : For no X window, use console instead
>   -o      : One MicroEmacs, use ME server if available
>   -p      : Pipe stdin into *stdin*, when saved output to stdout
>   -r      : Read-only, all buffers will be in view mode
>   -s <s>  : Search for string <s> in the next given file
>   -u <n>  : Set user name to <n> (sets $MENAME)
>   -v <v=s>: Set variable <v> to string <s>
>   -V      : Display Version info and exit
>   -x      : Don't catch signals
>   -y      : Load next file as a reduced binary file
>
> # arrh much better
> # though arguably the full usage (-help) should aso show its version
> info
> --------
>
> Heres the patch :
> ------------------
>
> *** main.c.orig Tue Oct 15 10:50:52 2002
> --- main.c      Fri Oct 10 03:07:02 2003
> ***************
> *** 97,102 ****
> --- 97,103 ----
>   "  -s <s>  : Search for string <s> in the next given file\n"
>   "  -u <n>  : Set user name to <n> (sets $MENAME)\n"
>   "  -v <v=s>: Set variable <v> to string <s>\n"
> + "  -V      : Display Version info and exit\n"
>   #ifdef _UNIX
>   "  -x      : Don't catch signals\n"
>   #endif
> ***************
> *** 1456,1461 ****
> --- 1457,1475 ----
>                       }
>                       break ;
>                   }
> +
> +             case 'V':
> +                 snprintf((char *)evalResult, sizeof(evalResult),
> +                          "%s %s - Date %s - %s\n",
> + #ifndef _NANOEMACS
> +                          ME_MICROEMACS_FULLNAME,
> + #else
> +                          ME_NANOEMACS_FULLNAME,
> + #endif
>
> +                         meVERSION, meDATE, meSYSTEM_NAME) ;
> +                 mePrintHelpMessage(evalResult) ;
> +                 meExit(0) ;
> +
>   #ifdef _UNIX
>               case 'x':
>                   sigcatch = 0 ;
> ***************
> *** 1862,1867 ****
> --- 1876,1882 ----
>           else if (cc == 'C')
>               break;                      /* Let the sucker go !! */
>       }
> +     return 0;
>   }
>   #endif
>
> ---
>
> Any likelihood of switching from a date based version to a familiar
> triplet
> major-minor-patch ???
>
> -- hops
> ------------------------ Yahoo! Groups Sponsor
> ---------------------~-->
> Buy Ink Cartridges or Refill Kits for your HP, Epson, Canon or Lexmark
> Printer at MyInks.com. Free s/h on orders $50 or more to the
> US & Canada.
> http://www.c1tracking.com/l.asp?cid=5511
> http://us.click.yahoo.com/mOAaAA/3exGAA/qnsNAA/89EolB/TM
> --------------------------------------------------------------
> -------~->
>
> ______________________________________________________________
> ____________
>
> This is an unmoderated list. JASSPA is not responsible for
> the content of
> any material posted to this list.
>
> To unsubscribe, send a mail message to
>
>     mailto:jasspa-unsubscribe@yahoogroups.com
>
> or visit http://groups.yahoo.com/group/jasspa and
> modify your account settings manually.
>
>
>
>
> Your use of Yahoo! Groups is subject to
http://docs.yahoo.com/info/terms/

All,

Isearch is a bit of an oddity and breaks many of the conventions in this area,
this is not just ME's isearch but Big emacs too and there are good reasons for
this.

I will not bore you with all the reasons I can think of, but I will try to
highlight the main one. Consider doing an isearch at the top of this email and
isearch for 'isX' ME would first find the 'is' on the first line and then jump
to the last line which has the first occurrence of "isX". Now consider executing
C-a C-k, does this mean you want to go back to the top and search for a new
string or start from the current location for a new string, I think probably the
latter. Given this now consider executing C-a ' instead, i.e. changing the
search string from "isX" to "'isX", given that the cursor is at the first isX
which does have a ' preceding it should ME move one character back and match at
the current location? Almost certainly so, in which case what does it mean to do
C-a . * instead (i.e. change to ".*isX") and how is the history (using the
backspace) meant to cope?

For those who lost the will to live in the middle of that paragraph the summary
is that if you start allowing standard message line editing such as C-a the
behaviour becomes ambiguous, very complex and everyone will starts losing the
will to live (especially me who'd probably end up trying to implement it!).

So I think the trick is to keep the isearch command simple and predictable and
learn how to get around its short comings in the most efficient way possible. So
to take the above as an example, if I was searching for isX and I wanted to
change this to 'isX I would do the following:

	 C-a - go to beginning of the line, this exits the Isearch keeping 'isX' in the
search history.

	 C-s - execute isearch

	 C-up - enter the standard message line editing with "isX" (this works only for
the first key), 'esc p' also works. This gives you full access to all the search
history and all standard message-line editing.

	 C-a - go to beginning of search string

	 ' - alter search string

	 C-s - search for it.

A very quick and easy way to overcome this common problem (and works in big
emacs as well),

Steve

> -----Original Message-----
> From: Mike Hopkirk [mailto:hops@...]
> Sent: 09 December 2003 18:16
> To: jasspa@yahoogroups.com
> Subject: Re: [jasspa] How to search forward for selection
>
>
> Jon Green wrote:
> >
> > Thomas Hundt wrote:
> > >  > On all but isearch C-w can be currently used to clear
> the message line
> > >  > so it is not easy for us to simply change this as it
> may hinder more
> > >  > people than it helps.
> > >
> > > I like the ^W feature in isearch (now that I know it's
> there!) but agree
> > > that there ought to be a way to clear the search text (^A
> ^K doesn't
> > > work) since I do this all the time (by cancelling the search and
> > > starting over) when I mistype something at the beginning
> of the search text.
> > >
> > > -Th
> > >
> >
> > I must admit I did not know that ^W existed.
>
> Ditto - ^W to me on cmdline is delete a word
> > I also think
> > that ^A (possibly ^P) that returned to an editable
> > search-forward type command line which is resumed to
> > isearch-forward by ^S would actually be quite nice.
>
> big emacs/Xemacs does something like this - I found it more confusing
> than helpful
> in being able to determine what searching mode you were in and what
> keystrokes toggled you back and fwd and did searches and ...
>
> -- hops
>
> ------------------------ Yahoo! Groups Sponsor
> ---------------------~-->
> Buy Ink Cartridges or Refill Kits for your HP, Epson, Canon or Lexmark
> Printer at MyInks.com. Free s/h on orders $50 or more to the
> US & Canada.
> http://www.c1tracking.com/l.asp?cid=5511
> http://us.click.yahoo.com/mOAaAA/3exGAA/qnsNAA/89EolB/TM
> --------------------------------------------------------------
> -------~->
>
> ______________________________________________________________
> ____________
>
> This is an unmoderated list. JASSPA is not responsible for
> the content of
> any material posted to this list.
>
> To unsubscribe, send a mail message to
>
>     mailto:jasspa-unsubscribe@yahoogroups.com
>
> or visit http://groups.yahoo.com/group/jasspa and
> modify your account settings manually.
>
>
>
>
> Your use of Yahoo! Groups is subject to
http://docs.yahoo.com/info/terms/

One thing I neglected to mention that I guess I should is that if you do 'C-s
return' (i.e. execute isearch and then press Enter) isearch drops into the
standard message line editing mode in which it remains until you type one of the
following:

	 C-g - abort
	 C-s - isearch forward for the current message line content
	 C-r - isearch backward for current ml content
	 return - perform a search-forward for current ml content, i.e. isearch becomes
search-forward

So at the cost of an extra return isearch can become the standard search-forward
(I bet no one knew that either - does anyone rtfm??). So what ever you can do in
search-forward you can easily do with isearch-forward!

So who's still using search-forward??

Steve

> -----Original Message-----
> From: Phillips, Steven
> Sent: 10 December 2003 13:27
> To: jasspa@yahoogroups.com
> Subject: RE: [jasspa] How to search forward for selection
>
>
> All,
>
> Isearch is a bit of an oddity and breaks many of the
> conventions in this area, this is not just ME's isearch but
> Big emacs too and there are good reasons for this.
>
> I will not bore you with all the reasons I can think of, but
> I will try to highlight the main one. Consider doing an
> isearch at the top of this email and isearch for 'isX' ME
> would first find the 'is' on the first line and then jump to
> the last line which has the first occurrence of "isX". Now
> consider executing C-a C-k, does this mean you want to go
> back to the top and search for a new string or start from the
> current location for a new string, I think probably the
> latter. Given this now consider executing C-a ' instead, i.e.
> changing the search string from "isX" to "'isX", given that
> the cursor is at the first isX which does have a ' preceding
> it should ME move one character back and match at the current
> location? Almost certainly so, in which case what does it
> mean to do C-a . * instead (i.e. change to ".*isX") and how
> is the history (using the backspace) meant to cope?
>
> For those who lost the will to live in the middle of that
> paragraph the summary is that if you start allowing standard
> message line editing such as C-a the behaviour becomes
> ambiguous, very complex and everyone will starts losing the
> will to live (especially me who'd probably end up trying to
> implement it!).
>
> So I think the trick is to keep the isearch command simple
> and predictable and learn how to get around its short comings
> in the most efficient way possible. So to take the above as
> an example, if I was searching for isX and I wanted to change
> this to 'isX I would do the following:
>
>  C-a - go to beginning of the line, this exits the
> Isearch keeping 'isX' in the search history.
>
>  C-s - execute isearch
>
>  C-up - enter the standard message line editing with
> "isX" (this works only for the first key), 'esc p' also
> works. This gives you full access to all the search history
> and all standard message-line editing.
>
>  C-a - go to beginning of search string
>
>  ' - alter search string
>
>  C-s - search for it.
>
> A very quick and easy way to overcome this common problem
> (and works in big emacs as well),
>
> Steve
>
> > -----Original Message-----
> > From: Mike Hopkirk [mailto:hops@...]
> > Sent: 09 December 2003 18:16
> > To: jasspa@yahoogroups.com
> > Subject: Re: [jasspa] How to search forward for selection
> >
> >
> > Jon Green wrote:
> > >
> > > Thomas Hundt wrote:
> > > >  > On all but isearch C-w can be currently used to clear
> > the message line
> > > >  > so it is not easy for us to simply change this as it
> > may hinder more
> > > >  > people than it helps.
> > > >
> > > > I like the ^W feature in isearch (now that I know it's
> > there!) but agree
> > > > that there ought to be a way to clear the search text (^A
> > ^K doesn't
> > > > work) since I do this all the time (by cancelling the search and
> > > > starting over) when I mistype something at the beginning
> > of the search text.
> > > >
> > > > -Th
> > > >
> > >
> > > I must admit I did not know that ^W existed.
> >
> > Ditto - ^W to me on cmdline is delete a word
> > > I also think
> > > that ^A (possibly ^P) that returned to an editable
> > > search-forward type command line which is resumed to
> > > isearch-forward by ^S would actually be quite nice.
> >
> > big emacs/Xemacs does something like this - I found it more
> confusing
> > than helpful
> > in being able to determine what searching mode you were in and what
> > keystrokes toggled you back and fwd and did searches and ...
> >
> > -- hops
> >
> > ------------------------ Yahoo! Groups Sponsor
> > ---------------------~-->
> > Buy Ink Cartridges or Refill Kits for your HP, Epson, Canon
> or Lexmark
> > Printer at MyInks.com. Free s/h on orders $50 or more to the
> > US & Canada.
> > http://www.c1tracking.com/l.asp?cid=5511
> > http://us.click.yahoo.com/mOAaAA/3exGAA/qnsNAA/89EolB/TM
> > --------------------------------------------------------------
> > -------~->
> >
> > ______________________________________________________________
> > ____________
> >
> > This is an unmoderated list. JASSPA is not responsible for
> > the content of
> > any material posted to this list.
> >
> > To unsubscribe, send a mail message to
> >
> >     mailto:jasspa-unsubscribe@yahoogroups.com
> >
> > or visit http://groups.yahoo.com/group/jasspa and
> > modify your account settings manually.
> >
> >
> >
> >
> > Your use of Yahoo! Groups is subject to
> http://docs.yahoo.com/info/terms/
>
>
>
> ------------------------ Yahoo! Groups Sponsor
> ---------------------~-->
> Buy Ink Cartridges or Refill Kits for your HP, Epson, Canon or Lexmark
> Printer at MyInks.com. Free s/h on orders $50 or more to the
> US & Canada.
> http://www.c1tracking.com/l.asp?cid=5511
> http://us.click.yahoo.com/mOAaAA/3exGAA/qnsNAA/89EolB/TM
> --------------------------------------------------------------
> -------~->
>
> ______________________________________________________________
> ____________
>
> This is an unmoderated list. JASSPA is not responsible for
> the content of
> any material posted to this list.
>
> To unsubscribe, send a mail message to
>
>     mailto:jasspa-unsubscribe@yahoogroups.com
>
> or visit http://groups.yahoo.com/group/jasspa and
> modify your account settings manually.
>
>
>
>
> Your use of Yahoo! Groups is subject to
http://docs.yahoo.com/info/terms/

>>(I never use the client-server mode.  Not sure what it's for.  Save
>>memory?  But memory's cheap!)
>
>
> I am an abuser of ME, I typically have one session of ME running for weeks and
can end up with 300+ buffers and 500+ Mb of memory (its cheap!). I use the
server because I want all the buffers in one me session so I can easily swap
between them (I hate using a mouse - too slow).
> Steve

That's great ;-)

I find the sessions so quick to start that it doesn't matter.  It is
quick to switch between them using Alt-Tab.  But I'm using it mainly in
non-console mode (and mainly in Win2k).

But when working on a project, I have many buffers going.  I've mapped
C-page-down to next-buffer and C-page-up to "-1 next-buffer" to make
switching between those very quick.

(BTW, I really appreciate being able to drag and drop files onto a
running ME.  Wish it would accept drag-and-drop of arbitrary files and
especially URLs, so I don't have to copy and paste them.  Not using ME
to browse, mind you, just to stick them into text documents.  I've
looked at it briefly and it looked like dropping files is a different
API section than dropping arbitrary.)

--
Thomas Hundt

Steve wrote:
> So at the cost of an extra return isearch can become the standard
> search-forward

Cool :-)

> (I bet no one knew that either - does anyone rtfm??).

Not if we think we know how it works already :-D

Also RTFM is an area where ME could use some improvement.  It's pretty
painful at the moment.  Specifically, the box that pops up when you do
General Help from the menu is cumbersome.  One can't isearch, for
example ;-)  And using the mouse wheel to scroll moves the darned thing
instead of scrolling.  And there's no scroll bar.  And scrolling wraps
from the end back to the beginning.  And the cursor bar seems to fight
between mouse and keyboard control (when you do PgUp and PgDn, for
example).  And the search order is case-sensitive.   And the window is
too small.  And doesn't go away when you select something.

Sorry for all the gripes.  I'm sure nobody wants to work on the
documentation!!  ;-)


--
Thomas Hundt

> Also RTFM is an area where ME could use some improvement. It's pretty
> painful at the moment.  Specifically, the box that pops up when you do
> General Help from the menu is cumbersome.  One can't isearch, for
> example ;-)  And using the mouse wheel to scroll moves the darned thing
> instead of scrolling.  And there's no scroll bar.  And scrolling wraps
> from the end back to the beginning.  And the cursor bar seems to fight
> between mouse and keyboard control (when you do PgUp and PgDn, for
> example).  And the search order is case-sensitive.   And the window is
> too small.  And doesn't go away when you select something.
>
> Sorry for all the gripes.  I'm sure nobody wants to work on the
> documentation!!  ;-)

So how come it has taken you several years before you bring these problems to
our attention? There I was merrily thinking that the world was rosy and all the
time the docs interface was pants!

Seriously though, please send us you gripes when you find them otherwise they
may never be fixed. Just because we implemented a feature does not mean that we
use it ourselves and therefore aware of its short comings.

Steve

The proposal made for handling keys in isearch sounds fine = Its a
thorny problem
given the dynamic behaviour and uncertainty about intent that editing
the search string
causes ..
FWIW I'd say editing the isearch string shpuld restart the search from
the start point rather than any current hit but thats just me...

"Phillips, Steven" wrote:
>
> One thing I neglected to mention that I guess I should is that if you do 'C-s
return' (i.e. execute isearch and then press Enter) isearch drops into the
standard message line editing mode in which it remains until you type one of the
following:
>
>         C-g - abort
>         C-s - isearch forward for the current message line content
>         C-r - isearch backward for current ml content
>         return - perform a search-forward for current ml content, i.e. isearch
becomes search-forward
>
> So at the cost of an extra return isearch can become the standard
search-forward

Sure - thats sometime useful when isearch isnt doing what you want, its
not
ideal if what you want is to just do a search...

>(I bet no one knew that either - does anyone rtfm??). So what ever you can do
in search-forward you can easily do with isearch-forward!

The cost - an extra cognitive dissonance keystroke to get into the right
mode is
still non-optimal
It'd be as convenient if there was a keystroke in search that boosted
the
search  to an isearch as well...
making both searches just search modes..

> So who's still using search-forward??

me - extra keystrokes to get where I want isnt goanna cut it....
-- hops

I asked this earlier but I guess it got lost in my other verbiage.

Are there any plans/thoughts of moving from a date based version
numbering system
i.e  Dec 2002
  to the (now) more common triplet
         major-minor-patch

  perhaps with production releases being even numbered ??

-- hops

The latest and greatest thing is 'Server 2003', MS moved away from year codes
with XP and then came running back to our way of thinking :-)

Our versioning is really the date code (see 'about' or your -V output, i.e. my
ME version is 03/10/01) which just happens to be a triplet number version which
is always increasing (apart from the millennium bug).

The main reason for the triplet number approach is for people who intend to
support and patch old versions, i.e. while the latest version is 15.0.0, the
supplier may still be supporting 14.1.23, 14.2.4  and 13.5.7 etc. We are not
aiming to do this for time reason (this is a hobby for us and as it takes us
best part of a week to generate a release and we simply cannot do this too often
due to other commitments). Instead we tend to fix major issues in the latest
version only; we may mail out either diff logs for users to fix their own or
sometimes a binary fix to a specific user to resolve their issue but this is
very rare.

However, given that 3rd parties are taking the base release of ME and
customizing it (you being one such third party I believe) we probably need a
standard way for you to put your own paw print on it so you know what version it
is and we know that it is not a standard release. Is this the reason you have
raised this issue?

Steve

> -----Original Message-----
> From: Mike Hopkirk [mailto:hops@...]
> Sent: 11 December 2003 19:11
> To: group
> Subject: [Fwd: [jasspa] version numbering]
>
>
>
> I asked this earlier but I guess it got lost in my other verbiage.
>
> Are there any plans/thoughts of moving from a date based version
> numbering system
> i.e  Dec 2002
>  to the (now) more common triplet
>         major-minor-patch
>
>  perhaps with production releases being even numbered ??
>
> -- hops
>
> ------------------------ Yahoo! Groups Sponsor
> ---------------------~-->
> Buy Ink Cartridges or Refill Kits for your HP, Epson, Canon or Lexmark
> Printer at MyInks.com. Free s/h on orders $50 or more to the
> US & Canada.
> http://www.c1tracking.com/l.asp?cid=5511
> http://us.click.yahoo.com/mOAaAA/3exGAA/qnsNAA/89EolB/TM
> --------------------------------------------------------------
> -------~->
>
> ______________________________________________________________
> ____________
>
> This is an unmoderated list. JASSPA is not responsible for
> the content of
> any material posted to this list.
>
> To unsubscribe, send a mail message to
>
>     mailto:jasspa-unsubscribe@yahoogroups.com
>
> or visit http://groups.yahoo.com/group/jasspa and
> modify your account settings manually.
>
>
>
>
> Your use of Yahoo! Groups is subject to
> http://docs.yahoo.com/info/terms/
>
>
>

Mike Hopkirk wrote:
> I asked this earlier but I guess it got lost in my other verbiage.
>
> Are there any plans/thoughts of moving from a date based version
> numbering system
> i.e  Dec 2002
>  to the (now) more common triplet
>         major-minor-patch
>
>  perhaps with production releases being even numbered ??
>
> -- hops

No, we only make 1 or 2 releases per year so it is
much easier to stick with the year numbering that
we currently use. It is also much easier to reference.
We do not make a release every 2 weeks where the
minor or revision numbers may increase, hence it
is not necessary for us to use this type of numbering.
There are still quite a few packages out there
that use date numbering, typically 8 chars i.e.
20031212 which may appear in a package name.

Internally we may generate say 50 different versions
between one release and the next, some of these
changes are trivial, others may be substantial
code re-work. I think this year we have laid down
2 new dated versions internally where the
functionality has substantially changed for testing.
(i.e the version I use at the moment is date
coded 03/10/01, this has been patched multiple
times - 2 times this week, but I cannot identify
that this has been patched from the version)

We do not have enough time to release everyone of
these changes as we would have to build across
all systems and re-test everything in addition
all of the documentation then has to be brought up
to date. It is much easier for us to sick with a
single major release, with maybe one minor. A major
version release will will typically take 1 month to
prepare by the time all of the changes have been
documented, packages built for different
platforms and verified.

Hence the date format works nicely for us !

Jon.

Proposed new version numbering scheme: 20031212.0.0

:-)

-Th


>>Are there any plans/thoughts of moving from a date based version
>>numbering system
>>i.e  Dec 2002
>> to the (now) more common triplet
>>        major-minor-patch
>>
>> perhaps with production releases being even numbered ??
>>-- hops
>
>
> No, we only make 1 or 2 releases per year so it is
> much easier to stick with the year numbering that
> we currently use. It is also much easier to reference.
> Jon.

Thomas Hundt wrote:
> Proposed new version numbering scheme: 20031212.0.0
>
> :-)
>
> -Th
>

Which is pretty much the same as the package version
numbers on the JASSPA site as generated in 2002/3.
I assume that .0.0 means midnight !

Jon.

"Phillips, Steven" wrote:


> However, given that 3rd parties are taking the base release of ME and
customizing it (you being one such third party I believe) we probably need a
standard way for you to put your own paw print on it so you know what version it
is and we know that it is not a standard release. Is this the reason you have
raised this issue?

Not really - any numbering scheme is modifiable by augmentation for
customization
mods  - just add another letter/number

Wondered about major-minor-patch(or not) numbering cos it was the scheme
Dan Lawrences original uemacs used ( duet anyway - 3.9, 4.0, ..)
  - I'd got used to it with that and other things
and i tend to think it gives a bit more more info
  - i.e whether the new version is a major change rev ( new features or
incompat change)
or just some minor revs/additions - also immediately apparent where a
rev fits in the
previous sets gone before.

Date based just indicates that theres a new rev and that its after any
previous ones

....
wrt Jons notes about multiple internal revs but only 1 or 2 releases per
year and only
working on latest rev ( without patches)  I understand about the release
  situation - I'm more asking about marking the releases more
granularly..

I tend to think that even internally you'd be as well off using a
major-minor-patch/tweak triplet rather than just a date ... that way
theres some indication of how widespread/big the change is rather than
having to know it a priori from
the date release

fro your example
"Internally we may generate say 50 different versions
between one release and the next, some of these
changes are trivial, others may be substantial
code re-work. I think this year we have laid down
2 new dated versions internally where the
functionality has substantially changed for testing."

I'd imagine you could tag these as the trivial changes being tagged with
incremental patch
version numbers ( indicating 'triviality') with the major changes -
substantial rework-
marked as minor changes...
The actual release may be major rev or a minor rev jump reflecting
internal
minor rev steps..

Not that it matters much to me anyway - Your project and work - mark it
how you like
but I was just wondering if anyone had considered it at all
given that many openSrc projects use that version scheme..

> > -----Original Message-----
> > From: Mike Hopkirk [mailto:hops@...]
> > Sent: 11 December 2003 19:11
> > To: group
> > Subject: [Fwd: [jasspa] version numbering]
> >
> >
> >
> > I asked this earlier but I guess it got lost in my other verbiage.
> >
> > Are there any plans/thoughts of moving from a date based version
> > numbering system
> > i.e  Dec 2002
> >  to the (now) more common triplet
> >         major-minor-patch
> >
> >  perhaps with production releases being even numbered ??
> >
> > -- hops

-- hops

I would like to:

1)  Open the file browser using F10
2)  Select a bunch of .h and .cpp files
3)  Apply Format | Buffer to each
4)  Apply Format | Clean Buffer to each
5)  Save each file.

How do I do this?

Preferably I would like to invoke the 3 commands from the command
line.  One after another.

Then I can, using the find -exec command, recursively do this to an
entire directory tree.

Salman

Salman,

Look at dos2unix.emf, this defines a start-up macro which will be called once
ME's initialisation is complete. It should not be difficult to change this to do
a clean etc.

Steve

> -----Original Message-----
> From: salmankhilji [mailto:salmankhilji@...]
> Sent: 15 December 2003 22:50
> To: jasspa@yahoogroups.com
> Subject: [jasspa] Format Buffer a bunch of files?
>
>
> I would like to:
>
> 1)  Open the file browser using F10
> 2)  Select a bunch of .h and .cpp files
> 3)  Apply Format | Buffer to each
> 4)  Apply Format | Clean Buffer to each
> 5)  Save each file.
>
> How do I do this?
>
> Preferably I would like to invoke the 3 commands from the command
> line.  One after another.
>
> Then I can, using the find -exec command, recursively do this to an
> entire directory tree.
>
> Salman
>
>
> ------------------------ Yahoo! Groups Sponsor
> ---------------------~-->
> Buy Ink Cartridges or Refill Kits for your HP, Epson, Canon or Lexmark
> Printer at MyInks.com. Free s/h on orders $50 or more to the
> US & Canada.
> http://www.c1tracking.com/l.asp?cid=5511
> http://us.click.yahoo.com/mOAaAA/3exGAA/qnsNAA/89EolB/TM
> --------------------------------------------------------------
> -------~->
>
> ______________________________________________________________
> ____________
>
> This is an unmoderated list. JASSPA is not responsible for
> the content of
> any material posted to this list.
>
> To unsubscribe, send a mail message to
>
>     mailto:jasspa-unsubscribe@yahoogroups.com
>
> or visit http://groups.yahoo.com/group/jasspa and
> modify your account settings manually.
>
>
>
>
> Your use of Yahoo! Groups is subject to
> http://docs.yahoo.com/info/terms/
>
>
>

Is there a way of increasing the limit on the number of keystrokes
saved by cntl-x (, cntl-x ) ?

You need to increase the size of meKBDMACRO_SIZE_MAX defined in
estruct.h and recompile. I have already increased this to 1024 for the
next release!

Steve

On 16 Dec 2003 at 15:08, jhines229@... wrote:

> Is there a way of increasing the limit on the number of keystrokes
> saved by cntl-x (, cntl-x ) ?
>
>
> __________________________________________________________________________
>
> This is an unmoderated list. JASSPA is not responsible for the content of
> any material posted to this list.
>
> To unsubscribe, send a mail message to
>
>     mailto:jasspa-unsubscribe@yahoogroups.com
>
> or visit http://groups.yahoo.com/group/jasspa and
> modify your account settings manually.
>
>
>
>
>
> Yahoo! Groups Links
>
> To visit your group on the web, go to:
>  http://groups.yahoo.com/group/jasspa/
>
> To unsubscribe from this group, send an email to:
>  jasspa-unsubscribe@yahoogroups.com
>
> Your use of Yahoo! Groups is subject to:
>  http://docs.yahoo.com/info/terms/
>
>

How can I get the "t" flag (to update Last Modified) enabled in my mode?

Also, I am following the tutorial on creating a new mode and it looks
quite a bit different than the modes that already exist. Is their a
reason for this? For instance, in the C mode I see such lines as:

set-variable .fhook-c.setup ® "/history" "fhook/c" "bdfghnopx"
set-variable .fhook-c.setup-mask "abdefghikmnoptux>"

I never see anything like:

!if &sequal .hilight.lua "ERROR"
     ; Allocate a highlighting scheme number
     set-variable .hilight.lua &pinc .hilight.next 1
!endif

which is in my mode (Lua by the way) (http://www.lua.org) and I use
quite extensively:

hilight .hilight.lua ... ... ...

Is their docs on how the other modes are created? What type of vars I
can use on a define-macro ? (assuming that indeed is what .fook-c is.

Thanks!

----
jeremy <jc@...>

Q: What do you call a boomerang that doesn't come back?
A: A stick.

Is their an existing Wiki for MicroEmacs?

----
jeremy <jc@...>

The box said Windoze 95 or greater, so I installed Linux!

Where is the unmark-buffer command? I have me '02 and it's in the docs,
but my editor says the command does not exist.

Thanks!

----
jeremy <jc@...>

Q: What do you call a boomerang that doesn't come back?
A: A stick.

Unmark-buffer is only defined in the MicroEMACS compatibility mode, see docs on
the command and macro file meme3_8.emf for a definition. The Jasspa ME way to do
this out of the box is to remove the 'edit' mode, given that the mode will be
set, toggling the mode has the same effect you simply do:

	 C-x m edit

Steve

> -----Original Message-----
> From: Jeremy Cowgar [mailto:jc@...]
> Sent: 19 December 2003 04:59
> To: jasspa@yahoogroups.com
> Subject: [jasspa] unmark-buffer ... where is it?
>
>
> Where is the unmark-buffer command? I have me '02 and it's in
> the docs,
> but my editor says the command does not exist.
>
> Thanks!
>
> ----
> jeremy <jc@...>
>
> Q: What do you call a boomerang that doesn't come back?
> A: A stick.
>
> ______________________________________________________________
> ____________
>
> This is an unmoderated list. JASSPA is not responsible for
> the content of
> any material posted to this list.
>
> To unsubscribe, send a mail message to
>
>     mailto:jasspa-unsubscribe@yahoogroups.com
>
> or visit http://groups.yahoo.com/group/jasspa and
> modify your account settings manually.
>
>
>
>
> Yahoo! Groups Links
>
> To visit your group on the web, go to:
>  http://groups.yahoo.com/group/jasspa/
>
> To unsubscribe from this group, send an email to:
>  jasspa-unsubscribe@yahoogroups.com
>
> Your use of Yahoo! Groups is subject to:
>  http://docs.yahoo.com/info/terms/
>
>
>

4 years ago there were only a dozen  or so file hooks, each hook was quite large
and had to do a lot of work to setup things like a Tools menu, help page etc.
There are now 64 file hooks and in an effort to keep maintenance down and
reliability up they now use the buffinit.emf and buffstp.emf macros to do the
majority of the work. This has vastly cut down the amount of code in each
hk???.emf file, increased the consistency and functionality of each file type.
The one downer is that the documentation has been left behind! Apologies for
that.

The best way to create a new file hook is to find an existing file hook that has
the same features (use find-buffer new.awk to create a new buffer of a given
file type and then use Help->Buffer Setup to get a list of supported features
(template, hilighting, indentation rules etc). Once you have found a suitable
hook copy it and make appropriate changes. The last thing to do, but probably
the most important, is to submit it to support@... (please attach the
hook files and some examples of the file type if you can). This means that we
can maintain the file hook for you (as ME and buffinit.emf matures the file
hooks change rapidly and updating the file hook can be tricky).

If you have further problems with creating your hklua.emf I suggest you mail
support with your hook and examples of what you're trying to do,

Steve

> -----Original Message-----
> From: Jeremy Cowgar [mailto:jc@...]
> Sent: 19 December 2003 00:59
> To: jasspa@yahoogroups.com
> Subject: [jasspa] Mode creation questions
>
>
> How can I get the "t" flag (to update Last Modified) enabled
> in my mode?
>
> Also, I am following the tutorial on creating a new mode and it looks
> quite a bit different than the modes that already exist. Is their a
> reason for this? For instance, in the C mode I see such lines as:
>
> set-variable .fhook-c.setup ® "/history" "fhook/c" "bdfghnopx"
> set-variable .fhook-c.setup-mask "abdefghikmnoptux>"
>
> I never see anything like:
>
> !if &sequal .hilight.lua "ERROR"
>     ; Allocate a highlighting scheme number
>     set-variable .hilight.lua &pinc .hilight.next 1
> !endif
>
> which is in my mode (Lua by the way) (http://www.lua.org) and I use
> quite extensively:
>
> hilight .hilight.lua ... ... ...
>
> Is their docs on how the other modes are created? What type of vars I
> can use on a define-macro ? (assuming that indeed is what .fook-c is.
>
> Thanks!
>
> ----
> jeremy <jc@...>
>
> Q: What do you call a boomerang that doesn't come back?
> A: A stick.
>
> ______________________________________________________________
> ____________
>
> This is an unmoderated list. JASSPA is not responsible for
> the content of
> any material posted to this list.
>
> To unsubscribe, send a mail message to
>
>     mailto:jasspa-unsubscribe@yahoogroups.com
>
> or visit http://groups.yahoo.com/group/jasspa and
> modify your account settings manually.
>
>
>
>
> Yahoo! Groups Links
>
> To visit your group on the web, go to:
>  http://groups.yahoo.com/group/jasspa/
>
> To unsubscribe from this group, send an email to:
>  jasspa-unsubscribe@yahoogroups.com
>
> Your use of Yahoo! Groups is subject to:
>  http://docs.yahoo.com/info/terms/
>
>
>

On Fri, 2003-12-19 at 06:06, Phillips, Steven wrote:

Can you tell me what the "bdfghnopx" and "abdefghikmnoptux" means? or
where to find this info?

set-variable .fhook-lua.setup ® "/history" "fhook/lua" "bdfghnopx"
set-variable .fhook-lua.setup-mask "abdefghikmnoptux>"

Thanks,

----
jeremy <jc@...>

jeremy@myhome $ ./think
Segmentation Fault

Greetings.

I am curious if you can refresh the tools in the toolbar via a macro?
Specifically, I want to refresh list items tool.

Thank you

----
jeremy <jc@...>

jeremy@myhome $ ./think
Segmentation Fault

Jeremy Cowgar wrote:
> On Fri, 2003-12-19 at 06:06, Phillips, Steven wrote:
>
> Can you tell me what the "bdfghnopx" and "abdefghikmnoptux" means? or
> where to find this info?
>
> set-variable .fhook-lua.setup ® "/history" "fhook/lua" "bdfghnopx"
> set-variable .fhook-lua.setup-mask "abdefghikmnoptux>"
>
> Thanks,

There is a bit of a hole in the documentation here
which has not been released. The '04 documentation
will address this issue - just takes lots of time to
write !

Bet place to look is in the header of the macro file
"buffstp.emf" which provides a bried description of
each letter.

Apologies for not finding enough time to fully
address this issue.

Jon.

Jeremy Cowgar wrote:
> Is their an existing Wiki for MicroEmacs?
>

No. Agreed that this might be a nice way to
run a FAQ and post new material. We do not
posess a public server with Wiki facilities,
and to be honset I'm not prepared to open
up a server with a public interface to support
it.

There are probably some sites that we can
use that may host this facility, but we have not
looked - if you have some ideas then please
post them.

Jon.