setedit-users — General list about setedit

Re: sLisp

[salvador <sal...@in...>]

"Thiago F.G. Albuquerque" wrote:

> On 24/09/01 09:50:18, salvador wrote:
>
> >> let
> >
> >I think what you wrote is currently:
> >(setv "str" (GetSelection))
>
> There is a difference: the variables created by 'let' are local. They only exist inside the let
> expression. The variables created by setv are all global, right? I don't know, setv solves my
> problem, but I think it would be nice to have local variables in the language too.

Sure, but not a high priority thing ;-)

> >> cond
> >
> >And I think that's
> >(if (not (strcmp "#i")) (InsertText "..."))
>
> There is a difference here too. "cond" is like if-elseif-else:
>
> (cond (test1) (action1)                 ; if (test1) (action1)
>         (test2) (action2)                       ; elseif (test2) (action2)
>         (test3) (action3)                       ; elseif (test3) (action3)
>       t       (else_action)             ; else (else_action)
> )

The same, currently the language lacks some indispensable structures. Adding stuff that can be
done with other statements (just typing more or needing to pay more attention) aren't high
priority. Of course patches are welcome ;-)

> The alternatives are mutualy exclusive. If one of the tests succeed, it executes the
> corresponding action and jumps out of the cond. It doesn't go trough any of the other tests.

I think that's the same to use: (if (test1) (action1) (if (test2) (action2) (...)))
Like in C: if (test1) {action1} else {if (test2) {action2} else {...}}

> >> ==
> >> (Lisp has several equality predicates (eq, equal, =,...). I don't know to which the one I am
> >> asking for corresponds, but it would be good to use the same notation as Lisp).
> >
> >I don't know, the only reason of the "Lisp" like stuff is that it is really easy to parse.
> >I preffer to make it closer to C and Perl.
>
> Yes, you're right. Lisp has some confusing stuff. What was I thinking of? :)
>
> >I agree, something like (call "function" [arguments ...]) could help.
> >The implicit way (function ....) could be also a solution, not sure.
>
> Since we are talking about this, I'd like to ask for two more things:
>
> 1) defun

What about it?

> 2) The ability to define macros with more than one func call without using 'eval'. The eval
> would be implicit.

Why? I agree a "call" is needed but is the same again: I don't want to add now things that are
redudant, just to save typing.

> Another idea is: replace defmacro with defun. The macro becomes a function with no args.

Maybe.

SET

--
Salvador Eduardo Tropea (SET). (Electronics Engineer)
Visit my home page: http://welcome.to/SetSoft or
http://www.geocities.com/SiliconValley/Vista/6552/
Alternative e-mail: se...@co... se...@ie...
Address: Curapaligue 2124, Caseros, 3 de Febrero
Buenos Aires, (1678), ARGENTINA Phone: +(5411) 4759 0013

Re: sLisp

[Thiago F.G. A. <tf...@za...>]

On 24/09/01 09:50:18, salvador wrote:

>> let
>
>I think what you wrote is currently:
>(setv "str" (GetSelection))

There is a difference: the variables created by 'let' are local. They only exist inside the let 
expression. The variables created by setv are all global, right? I don't know, setv solves my 
problem, but I think it would be nice to have local variables in the language too.

>> cond
>
>And I think that's
>(if (not (strcmp "#i")) (InsertText "..."))

There is a difference here too. "cond" is like if-elseif-else:

(cond (test1) (action1)			; if (test1) (action1)
	(test2) (action2)			; elseif (test2) (action2)
	(test3) (action3)			; elseif (test3) (action3)
      t       (else_action)		; else (else_action)
)

The alternatives are mutualy exclusive. If one of the tests succeed, it executes the 
corresponding action and jumps out of the cond. It doesn't go trough any of the other tests.

>> ==
>> (Lisp has several equality predicates (eq, equal, =,...). I don't know to which the one I am
>> asking for corresponds, but it would be good to use the same notation as Lisp).
>
>I don't know, the only reason of the "Lisp" like stuff is that it is really easy to parse.
>I preffer to make it closer to C and Perl.

Yes, you're right. Lisp has some confusing stuff. What was I thinking of? :)

>I agree, something like (call "function" [arguments ...]) could help.
>The implicit way (function ....) could be also a solution, not sure.

Since we are talking about this, I'd like to ask for two more things:

1) defun
2) The ability to define macros with more than one func call without using 'eval'. The eval 
would be implicit.

Another idea is: replace defmacro with defun. The macro becomes a function with no args.

Thiago

Re: sLisp (and maybe something else too?)

[salvador <sal...@in...>]

Grzegorz Adam Hankiewicz wrote:

> On Mon, Sep 24, 2001 at 09:50:18AM -0300, salvador wrote:
> > What really considered was using a powerful script language, like Perl,
> > but for this I must investigate how to do it.
> > I know Perl can be used in C through libperl, but I don't know if it
> > can handle "persistence" between calls.
>
> Hmmm... could you give an example of "persistence"?

Simple:
1) I call Perl with a program like it:

"$a=1;"

2) So Perl compiles it and then executes.
3) Control is back to C code.
4) Now I call Perl library again with:

"print $a;"

5) I should expect 1 as result.

In the past, I didn't really make any effort to solve it I saw the calls
acted as independent. The variables of the first call didn't exist in the
second.
I don't know if it was solved. In fact I didn't make the tests, a friend did
it.

> > Having Perl embedded in the editor could be really nice. It will be
> > added before version 1.0.0, of course it could be in 20 years ;-)
> > But it should be optional, so I need a basic macro language.
>
> And speaking about optional, I guess you will make us _all_ happy and let
> as optional a python parser too? I obviously volunteer to implement such
> an option when the time comes, and if nobody else steals the priviledge :)

:-)), no problem. I said Perl just because I use Perl. And I use Perl because
is much more popular than Python and hence you find much more information,
tools, etc. I don't have nothing against Python.

SET

--
Salvador Eduardo Tropea (SET). (Electronics Engineer)
Visit my home page: http://welcome.to/SetSoft or
http://www.geocities.com/SiliconValley/Vista/6552/
Alternative e-mail: se...@co... se...@ie...
Address: Curapaligue 2124, Caseros, 3 de Febrero
Buenos Aires, (1678), ARGENTINA Phone: +(5411) 4759 0013

Re: sLisp (and maybe something else too?)

[Grzegorz A. H. <gr...@te...>]

On Mon, Sep 24, 2001 at 09:50:18AM -0300, salvador wrote:
> What really considered was using a powerful script language, like Perl,
> but for this I must investigate how to do it.
> I know Perl can be used in C through libperl, but I don't know if it
> can handle "persistence" between calls.

Hmmm... could you give an example of "persistence"?

> Having Perl embedded in the editor could be really nice. It will be
> added before version 1.0.0, of course it could be in 20 years ;-)
> But it should be optional, so I need a basic macro language.

And speaking about optional, I guess you will make us _all_ happy and let
as optional a python parser too? I obviously volunteer to implement such
an option when the time comes, and if nobody else steals the priviledge :)

-- 
 Grzegorz Adam Hankiewicz   gr...@te...   http://gradha.infierno.org

Re: OpenFileUnderCursor

[salvador <sal...@in...>]

"Thiago F.G. Albuquerque" wrote:

> On 22/09/01 21:16:04, I wrote:
> Now, if I press ^Enter under "Token.h", it looks for "Token.h" in the current directory (the
> "main" source directory, the dir from which I started RHIDE). But it is not there, it is inside
> Lexer/.

Be sure to indicate *all* directories that contains headers in Option|Directories|Includes.
RHIDE uses this list.

SET

--
Salvador Eduardo Tropea (SET). (Electronics Engineer)
Visit my home page: http://welcome.to/SetSoft or
http://www.geocities.com/SiliconValley/Vista/6552/
Alternative e-mail: se...@co... se...@ie...
Address: Curapaligue 2124, Caseros, 3 de Febrero
Buenos Aires, (1678), ARGENTINA Phone: +(5411) 4759 0013

Re: OpenFileUnderCursor

[salvador <sal...@in...>]

"Thiago F.G. Albuquerque" wrote:

> The other day I wrote:
>
> > OpenFileUnderCursor doesn't work with <filename.h>, just with "filename.h"
>
> It turned out that this is not always the case. Sometimes it works just with <filename.h>,
> sometimes just with "filename.h" and sometimes with neither of them. I don't know the reason and I
> was not able to identify the circumstances under which each of these errors occurs. I guess this
> is not very helpful, isn't it? :-/

The problem is simple: For RHIDE the editor looks in %DJDIR%/include and the current directory if the
file isn't there fails. For Linux it just looks in the current directory (even worst).

> There is another thing. I don't know if this is already fixed in SETEDIT (in which case I
> apologize), but this is something I noticed in RHIDE. When you have something like this:
>
> #include "Lexer/Interface.h"
>
> OpenFileUnderCursor tries to open "Interface.h" in current directory. Can this be fixed?

RHIDE provides your own mechanism to load files under cursor. It uses the editor, but Robert provides
a function that is just in the middle of the process. So any complains about it goes to RHIDE list
;-)

About it. After discovering I wasn't even trying to look in /usr/include for Linux I added a new
configuration dialog to setup a list of directories where the editor will search. I had it in my todo
because I code setedit using setedit (not RHIDE) and I needed to have quick access to my headers
(spread in 5 directories).
That's in 0.4.49, lamentably I don't have TV in S.F. and newer versions of the editor needs a TV
that's newer. So I must first release a new TV, but I must close some bug reports first .... will be
available soon, but all of you must be patient.

SET

--
Salvador Eduardo Tropea (SET). (Electronics Engineer)
Visit my home page: http://welcome.to/SetSoft or
http://www.geocities.com/SiliconValley/Vista/6552/
Alternative e-mail: se...@co... se...@ie...
Address: Curapaligue 2124, Caseros, 3 de Febrero
Buenos Aires, (1678), ARGENTINA Phone: +(5411) 4759 0013

Re: sLisp

[salvador <sal...@in...>]

"Thiago F.G. Albuquerque" wrote:

> Today I would like to talk about some sLisp-related stuff.
>
> 1)In Emacs there is this "scratch buffer", where you can type Lisp expressions and then you
> press C-j and it evaluates it on the fly. It would be nice if SETEDIT had something like this.
> It would be great to debug macros.

Yes, I guess it could help. And isn't hard to implement. But I also think it will expose some
limitations ;-)

> 2) Sometime ago I asked SET for the possibility of assigning a (real) macro to a pseudo-macro
> trigger. I was thinking about it, and I wondered if it could be implemented as a sLisp macro,
> that would be assigned to Ctrl-Space. It would take the last two characters and depending on
> them, trigger other macros. Here's the code:
>
> (defmacro 'pmacro'
>  (eval
>   (SendCommands cmcSelectOn cmcCharLeft cmcCharLeft cmcSelectOff)
>   (let ((str (GetSelection)))
>           (cond ((== str "#i") (InsertText "#include "))
>                ; ((== str "")   ())  ; put your other macros here
>                ; ((== str "")   ())
>                 (t             (SendCommands cmcCharRight cmcCharRight))
>           )
>   )
>  )
> )
>
> But I need some new sLips commands:
>
> let

I think what you wrote is currently:
(setv "str" (GetSelection))

> cond

And I think that's
(if (not (strcmp "#i")) (InsertText "..."))

strcmp is just like the C function, so a match is 0, that's why I need the `not', `if' is just
like the C counterpart.

> ==
> (Lisp has several equality predicates (eq, equal, =,...). I don't know to which the one I am
> asking for corresponds, but it would be good to use the same notation as Lisp).

I don't know, the only reason of the "Lisp" like stuff is that it is really easy to parse.
I preffer to make it closer to C and Perl.

> Also, I would like a macro to be able to call another macro. Currently, code like this:
>
> (defmacro 'macro2'
>  ; something
> )
>
> (defmacro 'macro1'
>  (macro2)
> )
>
> doesn't work.

I agree, something like (call "function" [arguments ...]) could help.
The implicit way (function ....) could be also a solution, not sure.

> 3)
>
> SET,
>
>  I guess there is a number of open source embbedable Lisp interpreters out there. I've heard of
> one called "xlisp". Have you ever considered using one of them instead of sLisp in SETEDIT?

Not at all.

> The
> advantage is that all of Lisp's standard constructs (such as cond, let, etc...), types and
> functions are already implemented; you just have to add the editor-specific stuff. You don't
> have to reimplement everything from scratch.

And the disadvantage is that I must learn how it was coded and live with any thing I don't like
from it.
Adding new commands to my sLisp parser is really easy, just go and take a look how is implemented
strcmp.

> Please, don't take this as a criticism to sLisp. It is just a suggestion. What do you think?

What really considered was using a powerful script language, like Perl, but for this I must
investigate how to do it.
I know Perl can be used in C through libperl, but I don't know if it can handle "persistence"
between calls.
Having Perl embedded in the editor could be really nice. It will be added before version 1.0.0, of
course it could be in 20 years ;-)
But it should be optional, so I need a basic macro language.

SET

--
Salvador Eduardo Tropea (SET). (Electronics Engineer)
Visit my home page: http://welcome.to/SetSoft or
http://www.geocities.com/SiliconValley/Vista/6552/
Alternative e-mail: se...@co... se...@ie...
Address: Curapaligue 2124, Caseros, 3 de Febrero
Buenos Aires, (1678), ARGENTINA Phone: +(5411) 4759 0013

Re:OpenFileUnderCursor

[Thiago F.G. A. <tf...@za...>]

On 22/09/01 21:16:04, I wrote:

>#include "Lexer/Interface.h"
>
>OpenFileUnderCursor tries to open "Interface.h" in current directory. Can this be fixed?

Actually, this works. :) Sorry.

What does not work is this:

Now I am in "Lexer/Interface.h". It includes another file:

#include "Token.h"

Now, if I press ^Enter under "Token.h", it looks for "Token.h" in the current directory (the 
"main" source directory, the dir from which I started RHIDE). But it is not there, it is inside 
Lexer/.

The thing is: OpenFileUnderCursor should look for the file relative to the directory the current 
file is located, and not the working directory of RHIDE.

Thiago

OpenFileUnderCursor

[Thiago F.G. A. <tf...@za...>]

The other day I wrote:

> OpenFileUnderCursor doesn't work with <filename.h>, just with "filename.h"

It turned out that this is not always the case. Sometimes it works just with <filename.h>, 
sometimes just with "filename.h" and sometimes with neither of them. I don't know the reason and I 
was not able to identify the circumstances under which each of these errors occurs. I guess this 
is not very helpful, isn't it? :-/

There is another thing. I don't know if this is already fixed in SETEDIT (in which case I 
apologize), but this is something I noticed in RHIDE. When you have something like this:

#include "Lexer/Interface.h"

OpenFileUnderCursor tries to open "Interface.h" in current directory. Can this be fixed?

Thiago

sLisp

[Thiago F.G. A. <tf...@za...>]

Hi,

Today I would like to talk about some sLisp-related stuff.

1)In Emacs there is this "scratch buffer", where you can type Lisp expressions and then you 
press C-j and it evaluates it on the fly. It would be nice if SETEDIT had something like this. 
It would be great to debug macros.

2) Sometime ago I asked SET for the possibility of assigning a (real) macro to a pseudo-macro 
trigger. I was thinking about it, and I wondered if it could be implemented as a sLisp macro, 
that would be assigned to Ctrl-Space. It would take the last two characters and depending on 
them, trigger other macros. Here's the code:

(defmacro 'pmacro'
 (eval
  (SendCommands cmcSelectOn cmcCharLeft cmcCharLeft cmcSelectOff)
  (let ((str (GetSelection)))
          (cond ((== str "#i") (InsertText "#include "))
               ; ((== str "")   ())  ; put your other macros here
               ; ((== str "")   ())
                (t             (SendCommands cmcCharRight cmcCharRight))
          )
  )
 )
)

But I need some new sLips commands:

let
cond
==
(Lisp has several equality predicates (eq, equal, =,...). I don't know to which the one I am 
asking for corresponds, but it would be good to use the same notation as Lisp).

Also, I would like a macro to be able to call another macro. Currently, code like this:

(defmacro 'macro2'
 ; something
)

(defmacro 'macro1'
 (macro2)
)

doesn't work.

3)

SET,

 I guess there is a number of open source embbedable Lisp interpreters out there. I've heard of 
one called "xlisp". Have you ever considered using one of them instead of sLisp in SETEDIT? The 
advantage is that all of Lisp's standard constructs (such as cond, let, etc...), types and 
functions are already implemented; you just have to add the editor-specific stuff. You don't 
have to reimplement everything from scratch.

Please, don't take this as a criticism to sLisp. It is just a suggestion. What do you think?

Thiago

Small survey

[salvador <sal...@in...>]

Hi All!

  Next time you are downloading some part of the editor from Source Forge
please take a minute to answer a small survey I created. It have 5 simple
questions about the speed and memory consumption of the editor.

  If you want to go directly to the survey just use:

http://sourceforge.net/survey/survey.php?group_id=32835&survey_id=12225

Thanks, SET

--
Salvador Eduardo Tropea (SET). (Electronics Engineer)
Visit my home page: http://welcome.to/SetSoft or
http://www.geocities.com/SiliconValley/Vista/6552/
Alternative e-mail: se...@co... se...@ie...
Address: Curapaligue 2124, Caseros, 3 de Febrero
Buenos Aires, (1678), ARGENTINA Phone: +(5411) 4759 0013

TV and Big endian platforms!

[salvador <sal...@in...>]

Hi All:

  I don't know how much care about it but I managed to solve the tree endian
problems that were obvious in the TV demo example:

1) Mouse letting characters here and there.
2) Bogus shadows.
3) Non working keyboard (was worst to what I saw at first sight).

  The demo works quite well in a Linux/PPC.

SET

--
Salvador Eduardo Tropea (SET). (Electronics Engineer)
Visit my home page: http://welcome.to/SetSoft or
http://www.geocities.com/SiliconValley/Vista/6552/
Alternative e-mail: se...@co... se...@ie...
Address: Curapaligue 2124, Caseros, 3 de Febrero
Buenos Aires, (1678), ARGENTINA Phone: +(5411) 4759 0013

Re: SETEDIT: bugs

[salvador <sal...@in...>]

Hi Thiago:

> Thank you for your reply.
>
> >Why you don't want to have any of them? they are at the end of the list.
>
> When I close a window, I don't want to see it anymore. If I am still going to work on this
> buffer, I don't close it, I simply change to another window.

:-)), but you could open it again, no?

> >> When I run setedit as root, my options are overwriten (Don't make backups, Don't save desktop
> >> files, ...) They go back to the defaults.
> >
> >Sorry but I don't understand it, please clarify. The user `root' have their own settings.
>
> Ok. Here's exactly what I did. I was logged as 'thiago':
>
> su root
> e
> exit
>
> After this, the options of user 'thiago' go back to the defaults.

Please monitor these files:
1) /home/thiago/.setenvir.dat
2) /root/.setenvir.dat
3) /usr/share/setedit/.setenvir.dat (shouldn't exist)
Do it before going to root and after it.
Also: Are you using su and running setedit from /home/thiago? If yes just avoid it, it can create
files owned by root in your directory.

> I happened twice. Since then, I've been using another editor when I have to edit some file as
> root.

Do a backup of the files I mention (plus default desktops) and try it. I use setedit both, as root
and as a regular user and never saw a mix of settings.

> >> Font problem
> >> ============

[snip]

> Thanks for the URL. It contains very useful information.

I'm glad to heard it. I wrote it after configuring a system by nth time, I saw I was forced to
solve the same details again and again. So I put it in an easy to read way and available for
others.

SET

--
Salvador Eduardo Tropea (SET). (Electronics Engineer)
Visit my home page: http://welcome.to/SetSoft or
http://www.geocities.com/SiliconValley/Vista/6552/
Alternative e-mail: se...@co... se...@ie...
Address: Curapaligue 2124, Caseros, 3 de Febrero
Buenos Aires, (1678), ARGENTINA Phone: +(5411) 4759 0013

Re: SETEDIT: bugs

[Thiago F.G. A. <tf...@za...>]

Hi, SET!

Thank you for your reply.

>Why you don't want to have any of them? they are at the end of the list.

When I close a window, I don't want to see it anymore. If I am still going to work on this 
buffer, I don't close it, I simply change to another window.

>> When I run setedit as root, my options are overwriten (Don't make backups, Don't save desktop
>> files, ...) They go back to the defaults.
>
>Sorry but I don't understand it, please clarify. The user `root' have their own settings.

Ok. Here's exactly what I did. I was logged as 'thiago':

su root
e
exit

After this, the options of user 'thiago' go back to the defaults.

I happened twice. Since then, I've been using another editor when I have to edit some file as 
root.

>You could add them as wishes in the Source Forge page. It will help me to remmember

I will.

>> What happened to "Use no shadows" option?
>
>Did the standalone editor have it? I think that's an RHIDE setting.
>You could put it as a wish.

ok.

>> Font problem
>> ============
>>
>> I use lat1u-16, for Portuguese diacritics.
>
>I'll sugest you using lat1-16. They have the codes in the right place.
>The following URL:
>http://www.balug.org.ar/ConfTips/index.html
>Contains some tips about how to install the right fonts.
>Is in spanish but I think you'll understand it. It also explains how to avoid getting "broken
>frames" typical of lat1-16 fonts.
>
>> But then the graphic characters used by the editor
>> get all messed up (window borders, etc.). I guess I have to edit the font, but I don't know 
how
>> to do it. Does anybody know?
>
>Is much more easy:Tool&Ops|Options|Screen Options
>Look for the options with "(Linux)" because the Linux fonts adds some funny characters in the
>"undefined" range (to support VT100 charsets, in favor of PC one, which I think is quite
>questionable).

Thanks for the URL. It contains very useful information.

Thiago.

Cygwin port news

[salvador <sal...@in...>]

I did promess to inform about the news in this front:

Thanks to Ronald help the library is compiling and the demo.exe works. Some
details remains to be solved but they can be solved after a release. The
interesting thing is that no code was introduced, just adjustments about
what code to compile. Of course the configure mechanism needed a lot of fine
tunning to support the Cygwin stuff.
I also fixed things in the MinGW port and now is working with the last MinGW
stable release.

I hope to upload a new version with it in the next days but I want to solve
other details first.

SET

--
Salvador Eduardo Tropea (SET). (Electronics Engineer)
Visit my home page: http://welcome.to/SetSoft or
http://www.geocities.com/SiliconValley/Vista/6552/
Alternative e-mail: se...@co... se...@ie...
Address: Curapaligue 2124, Caseros, 3 de Febrero
Buenos Aires, (1678), ARGENTINA Phone: +(5411) 4759 0013

Re: [Rhide-users] Keyboard defaults

[Robert H. <rob...@gm...>]

At 14:32 17.09.01 -0300, salvador wrote:
>Hi All!
>
>   The question is: ¿Should we use it by default and enable the faster
>routine by user request?

I think, if switching to the bios functions will solve
the problems, we should use it by default.

******************************************************
* email:   Robert Hoehne <rob...@gm...>     *
* Post:    Am Berg 3, D-09573 Dittmannsdorf, Germany *
* WWW:     http://www.rhide.com                      *
******************************************************

Getting more from Source Forge

[salvador <sal...@in...>]

Hi All!

  I know these days I posted too much to the list and sorry if I'm over
encreasing the traffic. That's temporary until the move to Source Forge
stabilizes.
  I know the Source Forge pages are a little big heavy (around 35 Kb each
one, without much caching), but they offer some very interesting tools. One
is the "Tracker".
  Bug reports, Feature Requests, etc. can be tracked using it. I'm sure the
best is first ask in the list, but I also think it will be nice if after
determining something is a bug, a request, etc. you could add it to the
tracker.
  I have a rudimentary tracking system at home, but I think we will benefit
using Source Forge tracker.
  The main point is that using it things get public. So a user not
subscribed to the list can go and check if a bug was already reported, etc.

SET

--
Salvador Eduardo Tropea (SET). (Electronics Engineer)
Visit my home page: http://welcome.to/SetSoft or
http://www.geocities.com/SiliconValley/Vista/6552/
Alternative e-mail: set...@bi... se...@co...
                    se...@ie...
Address: Curapaligue 2124, Caseros, 3 de Febrero
Buenos Aires, (1678), ARGENTINA Phone: +(5411) 4759 0013

Re: SETEDIT: bugs

[salvador <sal...@in...>]

Hi Thiago:

> I am using SETEDIT 0.4.4.1 for Linux, binary distribuition.

No dot, is really .41, the 41th version of the 0.4 serie! (the build must be something like 500,
I use CVS only since 1998 ;-).

> Bugs
> ====
>
> Why I can't set "max. closed to remmember" = 0?

Is not a bug, is on purppose.

> The minimum allowed is 3. If those 3 were the special windows (Infview, Clipboard and the other
> that I can't remember), it would be ok -- that's what I want. But these "don't count". When I
> set max closed to 3, it keeps 3 *editor windows* always in the list, besides the special ones I
> mentioned. Is this the intended behaviour?

Yes. In 0.4.47 you can control much better what the editor remembers and what not. One of the
options is never for closed files, they live only for the session you are running and then are
erased. The minimun stills 3.
Why you don't want to have any of them? they are at the end of the list.

> ------------------------------------------------------------------------
>
> When I run setedit as root, my options are overwriten (Don't make backups, Don't save desktop
> files, ...) They go back to the defaults.

Sorry but I don't understand it, please clarify. The user `root' have their own settings.

> ------------------------------------------------------------------------
>
> Man page viewer
> ===============
>
> Can't copy from it.
> Doesn't close with ESC.

It still under alpha stage. In fact the one used in 0.4.41 version used to SIGSEGV when jumping to
another man page (double clicking in the name of a manpage, like xxxx(n) ). It was fixed in 0.4.47,
but the viewer is really "green".
You could add them as wishes in the Source Forge page. It will help me to remmember

> ------------------------------------------------------------------------
>
> OpenFileUnderCursor doesn't work with <filename.h>, just with "filename.h"

Are you sure? I'll investigate it.

> -------------------------------------------------------------------------
>
> Tabs
> ====
>
> Why "Tab Indents" is not a Global Option too?

Sorry, I have only 0.4.47 here and it only have "Tab smart indents", is that it?
In 0.4.47 it have a global option.

> IndentBlock doesn't work according to [ ]Tab Indents.

Nope. Are you sure it should? I don't see why.

> I mean, if Tab Indents is off, IndentBlock
> should insert tabs

Ah?!

> (or the corresponding amount of spaces, if Real Tabs if off), instead of
> jumping to the next hole in the line above.

Ok, I'll think about it. I think you are right, *but* it inserts a tab if "use real tabs" is
enabled.

> -------------------------------------------------------------------------
>
> What happened to "Use no shadows" option?

Did the standalone editor have it? I think that's an RHIDE setting.
You could put it as a wish.

> -------------------------------------------------------------------------
>
> Font problem
> ============
>
> I use lat1u-16, for Portuguese diacritics.

I'll sugest you using lat1-16. They have the codes in the right place.
The following URL:
http://www.balug.org.ar/ConfTips/index.html
Contains some tips about how to install the right fonts.
Is in spanish but I think you'll understand it. It also explains how to avoid getting "broken
frames" typical of lat1-16 fonts.

> But then the graphic characters used by the editor
> get all messed up (window borders, etc.). I guess I have to edit the font, but I don't know how
> to do it. Does anybody know?

Is much more easy:Tool&Ops|Options|Screen Options
Look for the options with "(Linux)" because the Linux fonts adds some funny characters in the
"undefined" range (to support VT100 charsets, in favor of PC one, which I think is quite
questionable).

SET

--
Salvador Eduardo Tropea (SET). (Electronics Engineer)
Visit my home page: http://welcome.to/SetSoft or
http://www.geocities.com/SiliconValley/Vista/6552/
Alternative e-mail: set...@bi... se...@co...
                    se...@ie...
Address: Curapaligue 2124, Caseros, 3 de Febrero
Buenos Aires, (1678), ARGENTINA Phone: +(5411) 4759 0013

SETEDIT: bugs

[Thiago F.G. A. <tf...@za...>]

Hi,

I am using SETEDIT 0.4.4.1 for Linux, binary distribuition.

Bugs
====

Why I can't set "max. closed to remmember" = 0?

The minimum allowed is 3. If those 3 were the special windows (Infview, Clipboard and the other 
that I can't remember), it would be ok -- that's what I want. But these "don't count". When I 
set max closed to 3, it keeps 3 *editor windows* always in the list, besides the special ones I 
mentioned. Is this the intended behaviour?

------------------------------------------------------------------------

When I run setedit as root, my options are overwriten (Don't make backups, Don't save desktop 
files, ...) They go back to the defaults.

------------------------------------------------------------------------

Man page viewer
===============

Can't copy from it.
Doesn't close with ESC.

------------------------------------------------------------------------

OpenFileUnderCursor doesn't work with <filename.h>, just with "filename.h"

-------------------------------------------------------------------------

Tabs
====

Why "Tab Indents" is not a Global Option too?

IndentBlock doesn't work according to [ ]Tab Indents. I mean, if Tab Indents is off, IndentBlock 
should insert tabs (or the corresponding amount of spaces, if Real Tabs if off), instead of 
jumping to the next hole in the line above.

-------------------------------------------------------------------------

What happened to "Use no shadows" option?

-------------------------------------------------------------------------

Font problem
============

I use lat1u-16, for Portuguese diacritics. But then the graphic characters used by the editor 
get all messed up (window borders, etc.). I guess I have to edit the font, but I don't know how 
to do it. Does anybody know?

-------------------------------------------------------------------------

Regards,
Thiago

Keyboard defaults

[salvador <sal...@in...>]

Hi All!

  In the last months I saw about ten users reporting keyboard problems wi=
th
RHIDE.
  The problem seems to be related with some BIOSes, I guess these BIOSes
make a particular use of the BIOS circular keyboard buffer that makes our=

code fail.
  The easiest solution is just use the BIOS routines. They are a little b=
it
slower but I don't think the impact could be easily persived by the end
user.
  The question is: =BFShould we use it by default and enable the faster
routine by user request?

SET

Note: Today another user asked it to me and I told him to download the la=
st
RHIDE and use the BIOS setting for keyboard and he reported the problem w=
ent
away.

--
Salvador Eduardo Tropea (SET). (Electronics Engineer)
Visit my home page: http://welcome.to/SetSoft or
http://www.geocities.com/SiliconValley/Vista/6552/
Alternative e-mail: set...@bi... se...@co...
                    se...@ie...
Address: Curapaligue 2124, Caseros, 3 de Febrero
Buenos Aires, (1678), ARGENTINA Phone: +(5411) 4759 0013

FreeBSD TV: partial success

[salvador <sal...@in...>]

Hi All!

  Following with the compiler farm stuff now I tried a FreeBSD machine
running x86.
  The configure process of TV currectly detects FreeBSD (I made it but is
the first time I test it!).
  My first surprise was that the make program is some "BSD make" that is
incompatible with GNU make at command line and commands level. But this
system had GNU make installed renamed to gmake. So the user must type gmake
and I must add this detection to correct a call in linuxso/makemak.pl which
blindly calls make.
  The second problem was a bug in the configuration process that I thinked
was fixed: the system doesn't have gpm, but the configuration process put
-lgpm in one place.
  And the third was a lack of -lstdc++ in the demo.mak, I must investigate
it.
  Finally I got a binary version of the demo and runned it.
  For some silly reason the screen is detected as monochome but the demo
works and Alt+key works, the mouse also works. Note that in all tests I'm
using ssh from an Eterm so TERM=xterm.
  No error during compilation and no special changes to the source code
needed.
  The system says: FreeBSD 4.3-RELEASE

SET

--
Salvador Eduardo Tropea (SET). (Electronics Engineer)
Visit my home page: http://welcome.to/SetSoft or
http://www.geocities.com/SiliconValley/Vista/6552/
Alternative e-mail: set...@bi... se...@co...
                    se...@ie...
Address: Curapaligue 2124, Caseros, 3 de Febrero
Buenos Aires, (1678), ARGENTINA Phone: +(5411) 4759 0013

Re: Key assignment

[salvador <sal...@in...>]

"Thiago F.G. Albuquerque" wrote:

> On 12/09/01 10:33:27, salvador wrote:
>
> >Yes, but isn't safe. When you do it using the TUI you *must* type the key you want. Under DOS
> you
> >won't get any surprise, the key you type is exactly what the editor gets and the name is quite
> >right. But under other OSs or when you have a non-us keyboard the name could be really
> different or
> >the key could be absent. A good example is Linux, some key combinations doesn't generate any
> code
> >that the editor can understand, that's configurable at OS level (in Debian
> >/etc/console-tools/default.kmap.gz).
> >If you edit a file by hand you can get very bad surprises.
>
> This reminds me of something I saw in the Zed editor. It uses a text file for the key bindings.
> To edit this file, the editor has a comand that switches the editor to "code mode". In this
> mode, for every key you type, it inserts que corresponding code string (escape sequence)
> generated by the key. To exit code mode, you press ESC twice (or something like this, I don't
> remeber).

ESC sequences aren't good, but the idea of inserting the name of the pressed key into the text
sound interesting.

[snip]

> Don't be angry at me, man. I am on your side. I am not complaining about the documentation, I
> think it is very good. I knew that the the keys assigned to menues were configured in
> menubind.smn, I was just making a suggestion. I'm sorry if I didn't sound that way.

Is ok.

SET

--
Salvador Eduardo Tropea (SET). (Electronics Engineer)
Visit my home page: http://welcome.to/SetSoft or
http://www.geocities.com/SiliconValley/Vista/6552/
Alternative e-mail: set...@bi... se...@co...
                    se...@ie...
Address: Curapaligue 2124, Caseros, 3 de Febrero
Buenos Aires, (1678), ARGENTINA Phone: +(5411) 4759 0013

Re: Key assignment

[Thiago F.G. A. <tf...@za...>]

On 12/09/01 10:33:27, salvador wrote:

>Yes, but isn't safe. When you do it using the TUI you *must* type the key you want. Under DOS 
you
>won't get any surprise, the key you type is exactly what the editor gets and the name is quite
>right. But under other OSs or when you have a non-us keyboard the name could be really 
different or
>the key could be absent. A good example is Linux, some key combinations doesn't generate any 
code
>that the editor can understand, that's configurable at OS level (in Debian
>/etc/console-tools/default.kmap.gz).
>If you edit a file by hand you can get very bad surprises.

This reminds me of something I saw in the Zed editor. It uses a text file for the key bindings. 
To edit this file, the editor has a comand that switches the editor to "code mode". In this 
mode, for every key you type, it inserts que corresponding code string (escape sequence) 
generated by the key. To exit code mode, you press ESC twice (or something like this, I don't 
remeber).

>> This can be pretty
>> annoying; I reassigned f9 to "make", but it didn't take effect because some menu assigned it 
to
>> SDG. I think all key assignments should be in the same place.
>
>Is impossible. The keys assigned to menues and status bar accelerators are configured in the
>menubind.smn file.
>It can't be in another way. If you don't modify the menu at the same time you modify the
>accelerator you will get really silly situations. Suppose you put Shift+F9 == SDG, but you 
don't
>modify the menu, then the menu will clearly say SDG==F9 and that's more than confusing.
>I know the documentation is quite hard to read because is mainly a lot of separated pieces and 
not
>one coherent description. But what a hell that's currently 6438 lines, 29997 words and 192990
>characters long! I really need help to put all together.
>I'm sure the keyboard configuration section should clearly state how to configure the menues 
and
>status bar accelerators by pointing to the section that describes it. But beleive me that 
writing
>such a huge documentation is hard specially when you must express complex stuff in a foreign
>language.

Don't be angry at me, man. I am on your side. I am not complaining about the documentation, I 
think it is very good. I knew that the the keys assigned to menues were configured in 
menubind.smn, I was just making a suggestion. I'm sorry if I didn't sound that way.

>> This solution also addresses another issue: the lack of user-specific macros. The macros file 
is
>> system-wide and you can edit it only if root lets you do so.
>
>That's incorrect. In Linux you should copy this file to your home directory and modify this 
copy.

Hum... I didn't know that. Thank you.

Thiago

Announce: Setedit 0.4.47 binaries for DOS

[salvador <sal...@in...>]

Hi All!

  Now binaries for djgpp are available (some files will need about 30
minutes to appear listed).
  This announcement is a follow up of the previous annuncement.
  The file edi0447i.zip contains the installer and the file edi0447b.zip is
a regular djgpp style distribution. Only one is needed.
  Remmember this is a beta release.

  http://setedit.sourceforge.net/

SET

P.S. I also uploaded the sources again fixing some small details that failed
for gcc 2.95.3 and binutils 2.11.2. They aren't in the CVS yet.

--
Salvador Eduardo Tropea (SET). (Electronics Engineer)
Visit my home page: http://welcome.to/SetSoft or
http://www.geocities.com/SiliconValley/Vista/6552/
Alternative e-mail: set...@bi... se...@co...
                    se...@ie...
Address: Curapaligue 2124, Caseros, 3 de Febrero
Buenos Aires, (1678), ARGENTINA Phone: +(5411) 4759 0013

Linux SPARC: TV partial success

[salvador <sal...@in...>]

Hi All!

  Well, the TV library cleanly compiled for SPARC. It shows the same
problems I saw in PPC so it means they are endian details.
  But we are quite close to get a fully functional TV for Linux/any CPU.
  This machine is an Ultra60 with 2 CPUs also available at Source Forge:

cpu             : TI UltraSparc II  (BlackBird)
fpu             : UltraSparc II integrated FPU
promlib         : Version 3 Revision 17
prom            : 3.17.0
type            : sun4u
ncpus probed    : 2
ncpus active    : 2
Cpu0Bogo        : 897.84
Cpu2Bogo        : 897.84
MMU Type        : Spitfire
State:
CPU0:           online
CPU2:           online

SET

--
Salvador Eduardo Tropea (SET). (Electronics Engineer)
Visit my home page: http://welcome.to/SetSoft or
http://www.geocities.com/SiliconValley/Vista/6552/
Alternative e-mail: set...@bi... se...@co...
                    se...@ie...
Address: Curapaligue 2124, Caseros, 3 de Febrero
Buenos Aires, (1678), ARGENTINA Phone: +(5411) 4759 0013