gnuplot-beta — Developers and beta testers of new gnuplot code

Re: help exit

[<br...@ph...>]

Petr Mikulik wrote:
>> Good idea.  But let me point out a detail: at least for gih help,
>> some of the node already *have* such a display: it's shown after you 
>> come out of the pager, and it asks you about a choice of sub-node.

> I don't see any...

Then you didn't look in the right place.  Here's the end of "help fit"
in a .gih-based version, as an example:

Subtopics available for fit:
     adjustable_parameters               beginners_guide   control
     error             error_estimate    errors            guide
     multi-branch      parameters        starting_values   tips

Subtopic for fit:  <input cursor here>

If you type "tips" here, it'll put you to the same node otherwise 
reached by "help fit tips".

In Windows, you get a collection of clickable links to the sub-nodes 
instead.

Re: Replot in epslatex terminal

[Ethan M. <merritt@u.washington.edu>]

On Friday 24 June 2005 10:02 am, Ga=EBl Varoquaux wrote:
> On Fri, Jun 24, 2005 at 09:56:43AM -0700, Ethan Merritt wrote:
> > "Counter-intuitive" obviously has different meanings for each person.
>=20
>     Sure, then lets make this behaviour obvious by stressing it in the
> help of the postscript terminal.=20

I am reminded of a request that went by on either the newsgroup or
the mailing list a while ago.  Re-casting it in the light of the=20
current discussion, it would be to create a new variant of 'set output'
that specifies a function to generate a filename, rather than the
filename itself.

	FILENAME(plotno) =3D sprintf("./plot_%02d.eps",plotno)
	set output sequential FILENAME

Each plot command would increment plotno and trigger the creation of a
new output file.

No, I don't have a full implementation in my head. I'm not sure=20
what difficulties might crop up.  But it would provide a
terminal-independent solution to the current problem.
 =20


=2D-=20
Ethan A Merritt       merritt@u.washington.edu
Biomolecular Structure Center
Mailstop 357742
University of Washington, Seattle, WA 98195

RE: Setjmp(), longjmp() in gnuplot

[Nigel N. <nN...@au...>]

Hi Timoth=E9e,

For my gnuplot-as-library driven from wxWidgets, I replaced=0D
bail_to_command_line() with an "extern C" routine that throws=0D
an exception back the C++ caller.  Then I wrap calls to the=0D
gnuplot subsystem with try/catch, e.g.,

  try {

    lib_do_line("set term wx");
    lib_do_line("set grid");
    lib_do_line("plot sin(x)");
   =0D
    // execute an arbitrary command string=0D
    // generated during a simulation:
    lib_do_line( cmd.c_str() );=0D

  } catch(...)  {
    // recover
  }

Nigel


----- Original Message -----
> I keep working on the wxwidgets terminal, and I am now=0D
> facing a difficult problem.=0D
>=0D
> When there are errors on a command line, gnuplot uses its=0D
> functions int_error(token, string) to inform the user and=0D
> stop parsing the command line. This function prints an=0D
> error message and then calls bail_to_command_line() which=0D
> is simply a wrapper for the standard longjmp(env) function.=0D
> This is similar to "goto", and puts the program in its=0D
> initial state by restoring the registers set in main().=0D
> --- [snip] ---


___________________________________________________________________________=
__________=0D
This message is intended for the addressee named and may contain=
 confidential and=0D
privileged information. If you are not the intended recipient please note=
 that=0D
any form of distribution, copying or use of this communication or the=
 information=0D
in it is strictly prohibited and may be unlawful. If you receive this=
 message in error,=0D
please delete it and notify the sender. =0D
Keep up to date with what's happening in Australian sport. Visit=
 www.ausport.gov.au=0D
___________________________________________________________________________=
__________

Re: Setjmp(), longjmp() in gnuplot

[Ethan M. <merritt@u.washington.edu>]

Caveat: I don't really understand how your new terminal type is supposed
to work.  So maybe I have the wrong end of the stick here...

On Friday 24 June 2005 11:45 am, Timoth=C3=A9e Lecomte wrote:
>=20
> As far as I understand, this would be perfect if I had not to use a=20
> separate thread for my terminal. Indeed, I want to send a command from=20
> the terminal (through a menu for example) to gnuplot. I use events=20
> defined in mouse.c, and more precisely a "command" event (which doesn't=20
> seem to be used very often, or maybe in OS/2 only). The do_event(event)=20
> is executed in my gui thread. When an error appears on the command, we=20
> obtain a longjump which is invalid for *this* gui thread... and it ends=20
> with a segmentation fault.

I would think the proper fix is to re-initialize the longjump at the
start of each thread creation.  Then if you get an error it will jump
to a per-thread error handler and exit the thread cleanly.

> Another solution would be to build the terminal as a separate process,=20
> like x11 or os/2 ones. But I think it's not an optimal design.
> So I would be pleased if somebody can give me a tip in order to deal=20
> with this issue ;-) .

I might understand better if you would briefly describe the flow of
control you are trying to achieve.  What does the new terminal type do
that existing terminals do not handle already?  Is the user expected to run
gnuplot from a command line, but when 'set term wxwidgets' is selected a=20
separate control window appears?   Or does the user run a wrapping program
that executes gnuplot underneath with the terminal pre-set to communicate w=
ith
the wrapping layer? =20

If it's the latter, then I'm not sure you even  need a new terminal type.
You may be able to use 'set term x11' and direct the output to an embedded
panel of the wrapping program (see patchset #1027032).  I'm not saying this
is the best thing to do, but I point it out as a possibility.
=20


=2D-=20
Ethan A Merritt       merritt@u.washington.edu
Biomolecular Structure Center
Mailstop 357742
University of Washington, Seattle, WA 98195

Re: help exit

[Petr M. <mi...@ph...>]

>> Speaking of "help", attached is a short patch for displaying the topic at 
>> the very beginning of help, e.g.,
>> 
>> gnuplot> help gnuplot
>> TOPIC: gnuplot-defined variables

I really like this. I would prefer this title:
   GNUPLOT HELP TOPIC: x11 x11_fonts\n

Typing sth like "help x11 x11" will display where your abbreviation points 
to.

> Good idea.  But let me point out a detail: at least for gih help,
> some of the node already *have* such a display: it's shown after you come out 
> of the pager, and it asks you about a choice of sub-node.

I don't see any...

---
PM

Re: Replot in epslatex terminal

[V. <gae...@no...>]

On Fri, Jun 24, 2005 at 09:56:43AM -0700, Ethan Merritt wrote:
> "Counter-intuitive" obviously has different meanings for each person.

    Sure, then lets make this behaviour obvious by stressing it in the
help of the postscript terminal. As for the epslatex terminal, I really
think it should be turned off : it took me a while to understand what
was happening when I discovered this and I am sure many users would just
have given up.

    Ga=EBl

Re: Replot in epslatex terminal

[Ethan M. <merritt@u.washington.edu>]

On Friday 24 June 2005 08:44 am, Hans-Bernhard Br=F6ker wrote:
> Ethan Merritt wrote:
>=20
> > As a result, you can have as many showpages in an EPSF file as you
> > like, because they are all redefined to null.  I don't see any specific
> > mention of %%Page in connection with EPSF files.
>=20
> It wouldn't be --- %%Page is not part of the PostScript language proper,=
=20
> but rather of the DSC extensions. =20
> That's why it's formatted like a PostScript comment.

Thanks for that bit of jargon, since it lets me find the relevant
sections of the PostScript spec.

The example EPSF files in the PostScript Language Reference Manual
do use %%Page operators.  It is explicitly allowed, as part of the
Document Structure Convention, for EPSF and EPSI files.

On Friday 24 June 2005 12:29 am, Ga=EBl Varoquaux wrote:
>=20
>     Well I think that for the epslatex terminal it is a bug : what's
> printed by the replot command never gets shown.

It may well be undesirable for the epslatex terminal. I defer to those
who actual use this terminal type. =20

> And for the eps terminal=20
> it is a very counter-intuitive feature that will trick many users. I
> think this beahve should at least be controled by a terminal option and
> turned off by default.

But for plain *.eps output generated by 'set term post eps', so far as I
understand Adobe's reference documentation, it is allowed.

Existing viewers (ghostscript, gv) handle the presence of multiple %%Page
elements by displaying one screen per %%Page in an eps file.=20
My current PostScript printer (Xerox Phaser 8400) handles them by printing
them all, but without a page-feed in between.  So the individual=20
%%BoundingBox lines allow multiple plots per sheet of paper, just as in
a full *.ps file.

"Counter-intuitive" obviously has different meanings for each person.
To me it is intuitive that a *.eps file is the same as a *.ps file
except that it allows being embedded in another document. I realize there
is also provision for having an embedded preview image bitmap in the
*.eps case, but I rarely encounter these.  I imagine this was an
accommodation to the historically limited support for displaying the actual
PostScript contents under MS Windows.=20

Anyhow, I would argue against changing the current (and long-standing)
default behavior of 'set term post eps'.  For epslatex, I have no
opinion one way or the other.

=2D-=20
Ethan A Merritt       merritt@u.washington.edu
Biomolecular Structure Center
Mailstop 357742
University of Washington, Seattle, WA 98195

Setjmp(), longjmp() in gnuplot

[<tim...@en...>]

Hello !

I keep working on the wxwidgets terminal, and I am now facing a=20
difficult problem.

When there are errors on a command line, gnuplot uses its functions=20
int_error(token, string) to inform the user and stop parsing the command=20
line. This function prints an error message and then calls=20
bail_to_command_line() which is simply a wrapper for the standard=20
longjmp(env) function. This is similar to "goto", and puts the program=20
in its initial state by restoring the registers set in main().

As far as I understand, this would be perfect if I had not to use a=20
separate thread for my terminal. Indeed, I want to send a command from=20
the terminal (through a menu for example) to gnuplot. I use events=20
defined in mouse.c, and more precisely a "command" event (which doesn't=20
seem to be used very often, or maybe in OS/2 only). The do_event(event)=20
is executed in my gui thread. When an error appears on the command, we=20
obtain a longjump which is invalid for *this* gui thread... and it ends=20
with a segmentation fault.

I can't simply delete the longjump, as it is necessary to stop command=20
line parsing. If I do it, gnuplot either seems to enter in a infinite=20
loop or terminante with an other segfault.

Of course, there's a solution : avoid any error in commands sent from my=20
terminal. But it's like reinventing the wheel. (These errors can be=20
various : for example, if I write a dialog to "set xlabel", and the user=20
enters forbidden characters; or I want to send a "replot" while it's the=20
"test" command which has opened the terminal, etc.)

Another solution would be to build the terminal as a separate process,=20
like x11 or os/2 ones. But I think it's not an optimal design.

So I would be pleased if somebody can give me a tip in order to deal=20
with this issue ;-) .

On my side, I will look more thoroughly to existing code to see how=20
things are done (by windows gui, os/2 pm, etc. )

Greetings,
Timoth=E9e

Re: Replot in epslatex terminal

[<br...@ph...>]

Ethan Merritt wrote:
> On Thursday 23 June 2005 02:18 pm, Hans-Bernhard Bröker wrote:
>>Gaël Varoquaux wrote:
>>>    It doesn't seem to. I just tried, giving gnuplot two plot commands,
>>>and when I open the file with gv I can select from the menu "page" the
>>>entry "next", and switch from my first plot to my second.

> So far as I know, it has always behaved that way. Certainly 4.0 did
> (I just confirmed this). I consider it a feature, not a bug. 

It may be a (dubious) feature in "postscript eps", but in epslatex, it 
cannot be anything else but a bug to behave like that.

> you can edit the individual plots in a tool like Illustrator, 

I'm quite sure Illustrator can't figure out epslatex output.

Re: Replot in epslatex terminal

[V. <gae...@no...>]

On Thu, Jun 23, 2005 at 05:22:08PM -0700, Ethan Merritt wrote:
> I consider it a feature, not a bug.

    Well I think that for the epslatex terminal it is a bug : what's
printed by the replot command never gets shown. And for the eps terminal
it is a very counter-intuitive feature that will trick many users. I
think this beahve should at least be controled by a terminal option and
turned off by default.

--
    Ga=EBl

Re: help exit

[Ethan M. <merritt@u.washington.edu>]

On Thursday 23 June 2005 07:44 am, Petr Mikulik wrote:
> I propose this text:
> 
> 
>   The commands `exit` and `quit`, as well as the END-OF-FILE character
>   (usually  Ctrl-D) terminate input from the current input stream: terminal
>   session,  pipe, and file input (pipe).
> 
>   If input streams are nested (inherited `load` scripts), then reading will
>   continue in the parent stream. When the top level stream is closed, the
>   program itself will exit.
> 
>   See "help batch/interactive" for more details.

Looks OK to me.

 
>   Each of these commands will clear the output device (as does the `clear`
>   command) before exiting. 
> BUT: Is the last point right??? Definitely not for -persist option!

I see nowhere in the code that calls clear_command() except for an
explicit `clear` from the command line.  So I think this is not right.


-- 
Ethan A Merritt       merritt@u.washington.edu
Biomolecular Structure Center
Mailstop 357742
University of Washington, Seattle, WA 98195

Re: Replot in epslatex terminal

[Ethan M. <merritt@u.washington.edu>]

On Thursday 23 June 2005 02:18 pm, Hans-Bernhard Br=F6ker wrote:
> Ga=EBl Varoquaux wrote:
>=20
> >     It doesn't seem to. I just tried, giving gnuplot two plot commands,
> > and when I open the file with gv I can select from the menu "page" the
> > entry "next", and switch from my first plot to my second.
>=20
> That's a bug then, which needs to be fixed.  Off-hand I'd guess it to=20
> have been introduced when the pslatex and epslatex stuff was merged.

So far as I know, it has always behaved that way. Certainly 4.0 did
(I just confirmed this). I consider it a feature, not a bug.=20
You are free to save each plot to a new file if you like. But you also
have the option of saving a sequence of plots to a single file. Later
you can edit the individual plots in a tool like Illustrator, and save
the ones you want to individual *.eps files for inclusion in a document. =20

=2D-=20
Ethan A Merritt       merritt@u.washington.edu
Biomolecular Structure Center
Mailstop 357742
University of Washington, Seattle, WA 98195

Re: Replot in epslatex terminal

[Ethan M. <merritt@u.washington.edu>]

On Thursday 23 June 2005 02:31 pm, Hans-Bernhard Br=F6ker wrote:
>=20
> I'm a little surprised that "postscript eps" mode doesn't behave=20
> properly either --- it manages to avoid actually outputting %%Page DSCs,=
=20
> but still allows multiple /showpage instances in the same file.  I=20
> suspect that's a direct violation of the EPSF file definition.

The PostScript standard says:

  The showpage operator is permitted in EPS files because it is present
  in so many PostScript language files [...] The application importing
  the EPS file is responsible for redefining showpage [...] according
  to the following code segment:  /showpage {} def

As a result, you can have as many showpages in an EPSF file as you
like, because they are all redefined to null.  I don't see any specific
mention of %%Page in connection with EPSF files.



=2D-=20
Ethan A Merritt       merritt@u.washington.edu
Biomolecular Structure Center
Mailstop 357742
University of Washington, Seattle, WA 98195

Re: Replot in epslatex terminal

[Ethan M. <merritt@u.washington.edu>]

On Thursday 23 June 2005 01:52 pm, Daniel J Sebald wrote:
> Hans-Bernhard Br=F6ker wrote:
> >=20
> > It cannot act like the X11 terminal.  It has to act like (I hope) png=20
> > does: ignore all further plot commands after the first one until the=20
> > output is closed and (another one) re-opened.

The png terminal does not act that way, nor should it.
My normal mode of generating and viewing png output is
	set term png
	set output '| display png;-'

Then I can view the output from successive plot commands interactively
by hitting <space> in the display window.  If I want to save a particular
one to disk, then I hit "save" instead.

> How about keep overwriting the old one?  That way, if a person makes a mi=
stake,=20
> s/he can retype the command without having go through the "set output; se=
t term=20
> ...; set output ..." sequence again.

Maybe. I would like to use eps and gv through a pipe=20
	set output '| gv -'
similarly to the way I use png+display, but unfortunately gv is not
happy with that.  So a compromise would indeed be to continually=20
overwrite the same output file, and view successive images in gv
by hitting the re-read button.  But I don't immediately see how I
would keep the two ends of the process in sync as easily as the
png+display case.



=2D-=20
Ethan A Merritt       merritt@u.washington.edu
Biomolecular Structure Center
Mailstop 357742
University of Washington, Seattle, WA 98195

Re: Key layout patch

[Daniel J S. <dan...@ie...>]

Petr Mikulik wrote:
>> I think the key-cleanup
> 
> 
> Yes, that's interesting. Please update it for current cvs.
> There, I propose to change
>     ... outside | below | <position>}
> to
>     outside | below | at <position>}
> 
> as giving raw numbers is always confusing.

OK.  It's changed, consistent with "labels" placement:

  Syntax:
        set key {on|off} {default}
                {{inside | outside} | {at <position>} | {above | below}}
                {left | right | center} {top | bottom | center}
[...]

The output of 'key.dem' is at:

http://www.acer-access.com/~ds...@ac.../gnuplot/key.pdf

I'll be leaving town shortly, so I've put an up-to-date patch on S.F.  However, 
there is one detail left.  It has to do with a discussion between Hans and I 
about manually set margins.  I couldn't recall the exact issue, but I have all 
the old emails.  (The problem is that ybot is a global variable set all over the 
place.  This fix was right in the middle of some code that gets tossed by the 
new patch.)

Dan


#if 0
     /* Dan Sebald 21jun2005:  Hans had added this hunk of code after
      * a new key patch was created.  It was placed somewhere that made
      * no sense in the new patch.  Sorry if this has created a bug,
      * but I can't exactly recall what the issue was.  Perhaps if
      * that bug crops up again we can find a proper home for this code.
      */
		/* HBB 20040725: leave manually set bmargin alone */
		if (bmargin < 0) {
		    ybot += key_entry_height * key_rows
			+ (int) (t->v_char * (ktitl_lines + 1));
		    ybot += (int) (key->height_fix * t->v_char);
		}
#endif

Re: Replot in epslatex terminal

[Daniel J S. <dan...@ie...>]

Hans-Bernhard Br=F6ker wrote:
> I'm a little surprised that "postscript eps" mode doesn't behave=20
> properly either --- it manages to avoid actually outputting %%Page DSCs=
,=20
> but still allows multiple /showpage instances in the same file.  I=20
> suspect that's a direct violation of the EPSF file definition.

Probably is.  But most drivers probably handle such a case in a fairly gr=
aceful=20
way because it isn't too difficult to deal with.

Dan

Re: help exit

[Daniel J S. <dan...@ie...>]

Daniel J Sebald wrote:
> But seriously, programming alternate behavior wouldn't be that 
> difficult.  If gnuplot's pager is active then don't print the first (or 
> last) line "TOPIC: <help>" but create a header if possible.  Not 
> suggesting that, just arguing it would be easy.  And possibly, as the 
> pager seems like a configurable thing already, perhaps an environment 
> variable could be set to turn off the printing of "TOPIC: <help>".  
> You'd think that advanced pagers would have a header convention, but 
> maybe not.

Octave has two modes of help and both appear to simply print information at the 
beginning of the help; no header...  not worth the effort.

Dan

Re: Replot in epslatex terminal

[<br...@ph...>]

Daniel J Sebald wrote:

> How about keep overwriting the old one?  That way, if a person makes a 
> mistake, s/he can retype the command without having go through the "set 
> output; set term ...; set output ..." sequence again.

I would tend to value internal consistency with other drivers more 
important than that --- the PNG/GD driver family established "typical" 
gnuplot behaviour, which any new drivers faced with the same problem 
should follow.

I'm a little surprised that "postscript eps" mode doesn't behave 
properly either --- it manages to avoid actually outputting %%Page DSCs, 
but still allows multiple /showpage instances in the same file.  I 
suspect that's a direct violation of the EPSF file definition.  I.e. the 
problem is bigger then it originally appeared.

Re: help exit

[Daniel J S. <dan...@ie...>]

Hans-Bernhard Br=F6ker wrote:
> Daniel J Sebald wrote:
>=20
>> Eh, displaying multiple times isn't professional looking.  In some=20
>> way, we can think of the topic as a "header".  As you point out, we=20
>> really don't want to print it as part of the help text.  Rather, what=20
>> would be nice, if possible, is if the pager had it's own kind of=20
>> header, e.g., at the top of the terminal is the header in bold while=20
>> the text scrolls underneath it.  Is that possible?
>=20
>=20
> No --- if only because we have no control over what the $PAGER chosen b=
y=20
> the user might be.  IIRC we have our own fall-back pager implementation=
=20
> in which such things could be done, but I would have to strictly oppose=
=20
> any change that made gnuplot ignore $PAGER, or only work "properly" if=20
> $PAGER wasn't set.

Yes, I suppose.  Then an orderly convention would be the only alternative=
.  It=20
might mean going through the gnuplot.doc and clearing out first lines of =
help=20
that say, e.g., "gnuplot-variable":... that is, if there are any.

But seriously, programming alternate behavior wouldn't be that difficult.=
  If=20
gnuplot's pager is active then don't print the first (or last) line "TOPI=
C:=20
<help>" but create a header if possible.  Not suggesting that, just argui=
ng it=20
would be easy.  And possibly, as the pager seems like a configurable thin=
g=20
already, perhaps an environment variable could be set to turn off the pri=
nting=20
of "TOPIC: <help>".  You'd think that advanced pagers would have a header=
=20
convention, but maybe not.

Dan

Re: help exit

[<br...@ph...>]

Daniel J Sebald wrote:

> Eh, displaying multiple times isn't professional looking.  In some way, 
> we can think of the topic as a "header".  As you point out, we really 
> don't want to print it as part of the help text.  Rather, what would be 
> nice, if possible, is if the pager had it's own kind of header, e.g., at 
> the top of the terminal is the header in bold while the text scrolls 
> underneath it.  Is that possible?

No --- if only because we have no control over what the $PAGER chosen by 
the user might be.  IIRC we have our own fall-back pager implementation 
in which such things could be done, but I would have to strictly oppose 
any change that made gnuplot ignore $PAGER, or only work "properly" if 
$PAGER wasn't set.

Re: Replot in epslatex terminal

[<br...@ph...>]

Gaël Varoquaux wrote:

>     It doesn't seem to. I just tried, giving gnuplot two plot commands,
> and when I open the file with gv I can select from the menu "page" the
> entry "next", and switch from my first plot to my second.

That's a bug then, which needs to be fixed.  Off-hand I'd guess it to 
have been introduced when the pslatex and epslatex stuff was merged.

Re: help exit

[Daniel J S. <dan...@ie...>]

Hans-Bernhard Br=F6ker wrote:
> Daniel J Sebald wrote:
>=20
>> Speaking of "help", attached is a short patch for displaying the topic=
=20
>> at the very beginning of help, e.g.,
>>
>> gnuplot> help gnuplot
>> TOPIC: gnuplot-defined variables
>=20
>=20
> Good idea.  But let me point out a detail: at least for gih help,
> some of the node already *have* such a display: it's shown after you=20
> come out of the pager, and it asks you about a choice of sub-node.

Oh yeah.  Some things to worry about there.

> It may not be a major issue if the title is displayed twice, but it may=
=20
> be a better idea to print it always at the same place (which would thus=
=20
> have to be at the end).

Eh, displaying multiple times isn't professional looking.  In some way, w=
e can=20
think of the topic as a "header".  As you point out, we really don't want=
 to=20
print it as part of the help text.  Rather, what would be nice, if possib=
le, is=20
if the pager had it's own kind of header, e.g., at the top of the termina=
l is=20
the header in bold while the text scrolls underneath it.  Is that possibl=
e?

Dan

Re: Replot in epslatex terminal

[V. <gae...@no...>]

On Thu, Jun 23, 2005 at 10:27:48PM +0200, Hans-Bernhard Br=F6ker wrote:
> It cannot act like the X11 terminal.  It has to act like (I hope) png=20
> does: ignore all further plot commands after the first one until the=20
> output is closed and (another one) re-opened.

    It doesn't seem to. I just tried, giving gnuplot two plot commands,
and when I open the file with gv I can select from the menu "page" the
entry "next", and switch from my first plot to my second.

--
    Ga=EBl

Re: Replot in epslatex terminal

[Daniel J S. <dan...@ie...>]

Hans-Bernhard Br=F6ker wrote:
> Ga=EBl Varoquaux wrote:
>=20
>>     Replot in epslatex terminal acts exactly like in postscript termin=
al
>> : it creates a new page of the epsfile.=20
>=20
>=20
> It must not do that.  EPS files by their very definition cannot
> have multiple pages.
>=20
>  > But this page never gets
>=20
>> displayed by latex.=20
>=20
>=20
> It shouldn't --- it's not allowed to be there.
>=20
>  > Wouldn't it be a more consistent option to have
>=20
>> "replot" act like in the x11, or the png terminal ?
>=20
>=20
> It cannot act like the X11 terminal.  It has to act like (I hope) png=20
> does: ignore all further plot commands after the first one until the=20
> output is closed and (another one) re-opened.

How about keep overwriting the old one?  That way, if a person makes a mi=
stake,=20
s/he can retype the command without having go through the "set output; se=
t term=20
...; set output ..." sequence again.

Dan

Re: help exit

[<br...@ph...>]

Daniel J Sebald wrote:

> Speaking of "help", attached is a short patch for displaying the topic 
> at the very beginning of help, e.g.,
> 
> gnuplot> help gnuplot
> TOPIC: gnuplot-defined variables

Good idea.  But let me point out a detail: at least for gih help,
some of the node already *have* such a display: it's shown after you 
come out of the pager, and it asks you about a choice of sub-node.

It may not be a major issue if the title is displayed twice, but it may 
be a better idea to print it always at the same place (which would thus 
have to be at the end).