plplot-devel — Technical discussion about PLplot development

Re: [Plplot-devel] plgriddata() commited

[Rafael L. <lab...@ps...>]

* Joao Cardoso <jc...@fe...> [2003-03-02 02:45]:

> This all works with *hand tuned* makefiles. I tried to automate everything
> but I wasn't able, sorry. Rafael and Alan, I count on you to fix the
> configuration.

I will take a look.

> I also added more capabilities to plenv(), allowing for square plots and 
> minor tick marks.

Have you broken backward compatibility?

> (1) After a fresh checkout, I got this:
> 
> jcard@home:~/tmp/plplot> ./bootstrap.sh 
> Running aclocal (GNU automake) 1.7.2... done
> Running autoheader (GNU Autoconf) 2.57... done
> Running automake (GNU automake) 1.7.2...configure.ac: installing 
> `./install-sh'
> configure.ac: installing `./mkinstalldirs'
> configure.ac: installing `./missing'
> configure.ac:433: installing `./config.guess'
> configure.ac:433: installing `./config.sub'
> configure.ac:433: required file `./ltmain.sh' not found
> Makefile.am:30: required directory ./libltdl does not exist
> bindings/c++/Makefile.am: installing `./depcomp'
> drivers/Makefile.am: installing `./compile'
> Makefile.am:30: required directory ./libltdl does not exist
>  done
> Running libtoolize (GNU libtool) 1.4.3... done
> Running autoconf (GNU Autoconf) 2.57... done
> 
> The second time I run it there where no problems.

I will try to fix this.

-- 
Rafael

[Plplot-devel] plgriddata() commited

[Joao C. <jc...@fe...>]

Hi,

I have committed all that is needed for plgriddata() to work (1)

This includes
src/plgridd.c, where plgriddata() is,
lib/csa with Pavel Sakov libcsa,
lib/nn with a patched Pavel Sakov libnn,
examples/x/x21c.c, a fairly comprehensible plgridadta() demo, see the cvs=
=20
commit message for further details.

This all works with *hand tuned* makefiles.
I tried to automate everything but I wasn't able, sorry.
Rafael and Alan, I count on you to fix the configuration.

Remember when using cvs that the directory "lib", that used to hold fonts=
 and=20
maps now holds two subdirectories each one with a small library. The file=
s=20
that used to be in "lib" are now on the new directory "data".

I also added more capabilities to plenv(), allowing for square plots and=20
minor tick marks.

Joao

(1) After a fresh checkout, I got this:

jcard@home:~/tmp/plplot> ./bootstrap.sh=20
Running aclocal (GNU automake) 1.7.2... done
Running autoheader (GNU Autoconf) 2.57... done
Running automake (GNU automake) 1.7.2...configure.ac: installing=20
`./install-sh'
configure.ac: installing `./mkinstalldirs'
configure.ac: installing `./missing'
configure.ac:433: installing `./config.guess'
configure.ac:433: installing `./config.sub'
configure.ac:433: required file `./ltmain.sh' not found
Makefile.am:30: required directory ./libltdl does not exist
bindings/c++/Makefile.am: installing `./depcomp'
drivers/Makefile.am: installing `./compile'
Makefile.am:30: required directory ./libltdl does not exist
 done
Running libtoolize (GNU libtool) 1.4.3... done
Running autoconf (GNU Autoconf) 2.57... done

The second time I run it there where no problems.

[Plplot-devel] cvs messed and recovered :(

[Joao C. <jc...@fe...>]

Hi,

Sorry, I made a mistake when moving the files from the lib to the new dat=
a=20
directory, but I have corrected it and also changed the cvs log message o=
f=20
Makefile.am and configure.ac accordingly.

Joao

Re: plgriddata and Directory Changes README! [was: Re: [Plplot-devel] Next release]

[Maurice L. <mj...@ga...>]

Rafael Laboissiere writes:
 > * Maurice LeBrun <mj...@ga...> [2003-02-28 15:29]:
 > 
 > > Rafael Laboissiere writes:
 > >  > Is memory space the main concern?  I never knew that!  Why should we care
 > >  > about this anachronic PLFLT thing if computers are delivered today with 3 Gb
 > >  > of RAM, typically? ;-)
 > > 
 > > My laptop only has 256M of memory..
 > > 
 > > Another reason as Alan mentions is to handle legacy codes.
 > 
 > Okay, I got the points.  I was just joking above...  I am looking forward
 > for my next workstation with 3Gb RAM, though.

Me too. :)

-- 
Maurice LeBrun    mj...@ga...
Research Organization for Information Science and Technology of Japan (RIST)

Re: [Plplot-devel] Configuration defaults

[Rafael L. <lab...@ps...>]

* Joao Cardoso <jc...@fe...> [2003-03-01 03:30]:

> [snip]
>
> If the user wants to use those techniques he will ponder and rebuild
> plplot. So, making float the default is not that bad. With a proper
> Announcement, users will be warned from the very beginning and will
> configure accordingly.  (But with a proper Announcement we could also say
> that doubles are now the default.)
> 
> Given all the above, I still think that doubles should be the default.

I am pretty convinced by your arguments.  I thought that it would be
straightforward (and acceptable) to cast back and forth all the data
processed with nn/csa.

> Pity we are not MS. We could make "configure" to send us the configure
> summary so we would collect statistics ;-)

Well, you can always implement this and this would work for non-MS systems
too.  Of course, we should not play the Insidious PLplot Big Brother, but
letting users to choose if they wish to send the informations or not.  There
is a package in Debian, called popularity-contest, that sends regularly
reports about what is installed in the system to a specific email address.
We are using this to collect statistics.

Another possibility would be to make a pool at our website.

-- 
Rafael

Re: [Plplot-devel] Configuration defaults

[Rafael L. <lab...@ps...>]

* Alan W. Irwin <ir...@be...> [2003-02-28 20:30]:

> > 5-Making plgriddata() convert its data from float to doubles is against what
> > the user specified at configure time.
> > 6-Most people who uses floats are likely to be using fortran with legacy code.
> > plgriddata() is not for them.
> 
> Those two arguments do not convince me....;-)

You put a smile at the end of your sentence above, so I hope you are joking.
If that is the case, I am afraid I did not get the joke.  At any rate, both
of Joao's arguments above are perfectly correct.

-- 
Rafael

Re: [Plplot-devel] Configuration defaults

[Alan W. I. <ir...@be...>]

On Sat, 1 Mar 2003, Joao Cardoso wrote:

> 5-Making plgriddata() convert its data from float to doubles is against what
> the user specified at configure time.
> 6-Most people who uses floats are likely to be using fortran with legacy code.
> plgriddata() is not for them.

Those two arguments do not convince me....;-)

Alan
__________________________
Alan W. Irwin
email: ir...@be...
phone: 250-727-2902

Astronomical research affiliation with Department of Physics and Astronomy,
University of Victoria (astrowww.phys.uvic.ca).

Programming affiliations with the Canadian Centre for Climate Modelling and
Analysis (www.cccma.bc.ec.gc.ca) and the PLplot scientific plotting software
package (plplot.org).

__________________________

Linux-powered Science
__________________________

Re: [Plplot-devel] Configuration defaults

[Alan W. I. <ir...@be...>]

On Fri, 28 Feb 2003, Alan W. Irwin wrote:

> However, now that you think double is a better default I am willing to go
> along with that.  If nobody else objects in the next few hours I will
> just go ahead and make the change.
>
> OK.  Unless I hear other objections in the next few hours, I will remove
> with_gcc.
>
> I think the best way to deal with this is to completely remove with_shlib
> and change over to enable_shared logic in sysloc.in.  If nobody objects
> to this change I will do it in the next few hours.

All three changes have now been done.  A complete build and
install from a clean start gave a good plplot-test.sh result.

Enjoy the new default configuration!

BTW, Maurice.  When I was tracking down all the with_gcc stuff it looks like
that variable is used for other things in autotools.  It is possible that
name clash was killing your attempts to use KCC.  Anyhow, now we no longer
attempt to use with_gcc, it is probably worth another try with KCC.  I will
be most interested in the results of that experiment.

Alan
__________________________
Alan W. Irwin
email: ir...@be...
phone: 250-727-2902

Astronomical research affiliation with Department of Physics and Astronomy,
University of Victoria (astrowww.phys.uvic.ca).

Programming affiliations with the Canadian Centre for Climate Modelling and
Analysis (www.cccma.bc.ec.gc.ca) and the PLplot scientific plotting software
package (plplot.org).

__________________________

Linux-powered Science
__________________________

Re: [Plplot-devel] Re: plgriddata and Directory Changes README! [was: Re: [Plplot-devel] Next release]

[Joao C. <jc...@fe...>]

On Friday 28 February 2003 21:20, Alan W. Irwin wrote:
> On Fri, 28 Feb 2003, [iso-8859-1] Jo=E3o Cardoso wrote:
> > | I agree, but you did not answer my question, probably because I mad=
e
> > | a typo. Here I go again: Can't you cast the values returned by nn/c=
sa
> > | with PLFLT?
> >
> > No, data arrays are passed by reference. With nn/csa/qhull array data
> > flows in both directions.
> > If the user wants to use floats because he has huge amounts of float
> > data we would have convert it all back and forth. And the main user
> > concern, memory space, would be jeopardized anyway.
>
> I take your point about memory space, but users may still want
> single-precision to work as a convenience because their application
> programme is inherently single precision.  We do serve the scientific
> plotting community, and there is still a large number of fortran
> single-precision applications around (as I am just re-discovering in my
> current contract). It is extremely convenient to be able to use PLplot
> directly from such a large application without having to do anything
> special with the precision of the PLplot arguments (or the inconvenient
> alternative of writing the results to a file which is then read by a
> special
> double-precision PLplot application).
>
> Thus, I hope very much that you do make an effort to completely support
> both single and double (by allocating the space, copying the arrays, et=
c.)
> just as in the rest of PLplot.

Please see my other e-mail on "configuration defaults". In short, I think=
 that=20
if the user has specified float we must not use doubles.

Joao

>
> Alan
> __________________________
> Alan W. Irwin
> email: ir...@be...
> phone: 250-727-2902
>
> Astronomical research affiliation with Department of Physics and Astron=
omy,
> University of Victoria (astrowww.phys.uvic.ca).
>
> Programming affiliations with the Canadian Centre for Climate Modelling=
 and
> Analysis (www.cccma.bc.ec.gc.ca) and the PLplot scientific plotting
> software package (plplot.org).
>
> __________________________
>
> Linux-powered Science
> __________________________
>
>
>
> -------------------------------------------------------
> This sf.net email is sponsored by:ThinkGeek
> Welcome to geek heaven.
> http://thinkgeek.com/sf
> _______________________________________________
> Plplot-devel mailing list
> Plp...@li...
> https://lists.sourceforge.net/lists/listinfo/plplot-devel

Re: [Plplot-devel] Configuration defaults

[Joao C. <jc...@fe...>]

On Friday 28 February 2003 21:12, Maurice LeBrun wrote:
> Jo=E3o Cardoso writes:
>  > On Friday 28 February 2003 19:51, Alan W. Irwin wrote:
>  > | On Thu, 27 Feb 2003, Maurice LeBrun wrote:
>  > | > These are fine with me personally, yes even with_double.  Althou=
gh
>  > | > in that case you may trip up some people who are used to the old
>  > | > behavior.
>  > |
>  > | OK.  I left with-double=3Dno.
>  >
>  > I don't think that Maurice really meant to keep with-double=3Dno. Af=
ter
>  > all, when we change the API the consequences are much worse, and tha=
t
>  > don't inhibits us from doing it.
>  > And we have the configure summary, that clearly states the float/dou=
ble
>  > status.
>  > Also, with the new nn/csa/qhull libraries, the better is to use doub=
le
>  > by default.
>  > So, I vote for with-double=3Dyes.
>
> I didn't actually say I objected; what I wrote was:
> > These are fine with me personally, yes even with_double.  Although in
> > that case you may trip up some people who are used to the old behavio=
r.
>
> So the latter is just a warning that this is yet another change that ma=
y
> cause some transition problems for some people.  Change is always ok if
> there is a large enough benefit.. is that the case here?  It's easy eno=
ugh
> to specify --with-double[=3Dyes].  Are there other benefits to be gaine=
d,
> other than working with the nn/csa/qhull libraries?

No.

But lets see the other point of view. If users don't specify double, they=
 will=20
loose the best 50% of plgriddata(). And I belive that the majority of use=
rs=20
use doubles, and so the default should be for them.

But this is just a default. We could warn users against using the default=
 !?!.

To sumarise the facts:

1-plgriddata() has 3 native methods that will use either float or doubles=
=2E
2-The 3 other csa/nn/qhul methods are clearly best in most circumstances =
but=20
require doubles.
3-csa is a single file that depends on nothing and that we can "sed" to s=
uit=20
our taste (this is an just an hypothesis).
4-nn/qhull are responsible for the best 2 methods.
   Qhull must exist in the user system for nn to work.
   Qhull is not generally available in most systems.
   Qhull can be compiled for floats, but it not recommended.
   It is not practical to change nn to accept floats.
5-Making plgriddata() convert its data from float to doubles is against w=
hat=20
the user specified at configure time.
6-Most people who uses floats are likely to be using fortran with legacy =
code.=20
plgriddata() is not for them.

Given all point 4 above limitations regarding Qhull, most users will not =
use=20
the plgriddata() related capabilities (but they will feel curious, I hope=
=2E=20
Lets watch the Qhull download statistics when we announce plplot)

Currently users are warned:
   plabort("plgriddata(): you must have Qhull to use GRID_NNI.");
I can also add
   plabort("plgriddata(): you must use doubles to use GRID_NNI.");

If the user wants to use those techniques he will ponder and rebuild plpl=
ot.
So, making float the default is not that bad. With a proper Announcement,=
=20
users will be warned from the very beginning and will configure according=
ly.=20
(But with a proper Announcement we could also say that doubles are now th=
e=20
default.)

Given all the above, I still think that doubles should be the default.

Pity we are not MS. We could make "configure" to send us the configure su=
mmary=20
so we would collect statistics ;-)

>
> Either way it won't affect me, as I always install both float & double
> versions.
>
>  > Maurice, some time ago you argued that we should not drop float supp=
ort.
>  > Could you please "argue" again? Sourceforge mailing lists don't have=
 a
>  > search facility, what makes difficult to find something we know is
>  > there.
>
> If your code uses large floating point arrays and you're developing on =
a
> personal machine the 50% memory reduction gained (and some cpu time gai=
ned)
> by using (float *) rather than (double *) can be very significant.

Yes, 50% is a lot, specially in my salary :)
Joao

Re: plgriddata and Directory Changes README! [was: Re: [Plplot-devel] Next release]

[Rafael L. <lab...@ps...>]

* Maurice LeBrun <mj...@ga...> [2003-02-28 15:29]:

> Rafael Laboissiere writes:
>  > Is memory space the main concern?  I never knew that!  Why should we care
>  > about this anachronic PLFLT thing if computers are delivered today with 3 Gb
>  > of RAM, typically? ;-)
> 
> My laptop only has 256M of memory..
> 
> Another reason as Alan mentions is to handle legacy codes.

Okay, I got the points.  I was just joking above...  I am looking forward
for my next workstation with 3Gb RAM, though.

-- 
Rafael

Re: plgriddata and Directory Changes README! [was: Re: [Plplot-devel] Next release]

[Maurice L. <mj...@ga...>]

Rafael Laboissiere writes:
 > Is memory space the main concern?  I never knew that!  Why should we care
 > about this anachronic PLFLT thing if computers are delivered today with 3 Gb
 > of RAM, typically? ;-)

My laptop only has 256M of memory..

Another reason as Alan mentions is to handle legacy codes.

-- 
Maurice LeBrun    mj...@ga...
Research Organization for Information Science and Technology of Japan (RIST)

[Plplot-devel] Re: plgriddata and Directory Changes README! [was: Re: [Plplot-devel] Next release]

[Alan W. I. <ir...@be...>]

On Fri, 28 Feb 2003, [iso-8859-1] Jo=E3o Cardoso wrote:

> | I agree, but you did not answer my question, probably because I made
> | a typo. Here I go again: Can't you cast the values returned by nn/csa
> | with PLFLT?
>
> No, data arrays are passed by reference. With nn/csa/qhull array data
> flows in both directions.
> If the user wants to use floats because he has huge amounts of float
> data we would have convert it all back and forth. And the main user
> concern, memory space, would be jeopardized anyway.

I take your point about memory space, but users may still want
single-precision to work as a convenience because their application
programme is inherently single precision.  We do serve the scientific
plotting community, and there is still a large number of fortran
single-precision applications around (as I am just re-discovering in my
current contract). It is extremely convenient to be able to use PLplot
directly from such a large application without having to do anything specia=
l
with the precision of the PLplot arguments (or the inconvenient alternative
of writing the results to a file which is then read by a special
double-precision PLplot application).

Thus, I hope very much that you do make an effort to completely support bot=
h
single and double (by allocating the space, copying the arrays, etc.) just
as in the rest of PLplot.

Alan
__________________________
Alan W. Irwin
email: ir...@be...
phone: 250-727-2902

Astronomical research affiliation with Department of Physics and Astronomy,
University of Victoria (astrowww.phys.uvic.ca).

Programming affiliations with the Canadian Centre for Climate Modelling and
Analysis (www.cccma.bc.ec.gc.ca) and the PLplot scientific plotting softwar=
e
package (plplot.org).

__________________________

Linux-powered Science
__________________________

Re: plgriddata and Directory Changes README! [was: Re: [Plplot-devel] Next release]

[Rafael L. <lab...@ps...>]

* João Cardoso <jc...@fe...> [2003-02-28 20:45]:

> No, data arrays are passed by reference. With nn/csa/qhull array data 
> flows in both directions. If the user wants to use floats because he has
> huge amounts of float data we would have convert it all back and forth.

I see.  My proposal was to do the cast transparently in your plgriddata
function, so that the user would never know.

> And the main user concern, memory space, would be jeopardized anyway. 

Is memory space the main concern?  I never knew that!  Why should we care
about this anachronic PLFLT thing if computers are delivered today with 3 Gb
of RAM, typically? ;-)

-- 
Rafael

Re: [Plplot-devel] Configuration defaults

[Maurice L. <mj...@ga...>]

Jo=E3o Cardoso writes:
 > On Friday 28 February 2003 19:51, Alan W. Irwin wrote:
 > | On Thu, 27 Feb 2003, Maurice LeBrun wrote:
 > | > These are fine with me personally, yes even with_double.  Althou=
gh
 > | > in that case you may trip up some people who are used to the old=

 > | > behavior.
 > |
 > | OK.  I left with-double=3Dno.
 >=20
 > I don't think that Maurice really meant to keep with-double=3Dno. Af=
ter=20
 > all, when we change the API the consequences are much worse, and tha=
t=20
 > don't inhibits us from doing it.
 > And we have the configure summary, that clearly states the float/dou=
ble=20
 > status.
 > Also, with the new nn/csa/qhull libraries, the better is to use doub=
le=20
 > by default.
 > So, I vote for with-double=3Dyes.

I didn't actually say I objected; what I wrote was:

> These are fine with me personally, yes even with_double.  Although in=
 that
> case you may trip up some people who are used to the old behavior.

So the latter is just a warning that this is yet another change that ma=
y cause
some transition problems for some people.  Change is always ok if there=
 is a
large enough benefit.. is that the case here?  It's easy enough to spec=
ify
--with-double[=3Dyes].  Are there other benefits to be gained, other th=
an
working with the nn/csa/qhull libraries?

Either way it won't affect me, as I always install both float & double
versions.

 > Maurice, some time ago you argued that we should not drop float supp=
ort.=20
 > Could you please "argue" again? Sourceforge mailing lists don't have=
 a=20
 > search facility, what makes difficult to find something we know is=20=

 > there.

If your code uses large floating point arrays and you're developing on =
a
personal machine the 50% memory reduction gained (and some cpu time gai=
ned) by
using (float *) rather than (double *) can be very significant.

--=20
Maurice LeBrun    mj...@ga...
Research Organization for Information Science and Technology of Japan (=
RIST)

Re: [Plplot-devel] Configuration defaults

[Alan W. I. <ir...@be...>]

On Fri, 28 Feb 2003, [iso-8859-1] Jo=E3o Cardoso wrote:

> On Friday 28 February 2003 19:51, Alan W. Irwin wrote:
> | On Thu, 27 Feb 2003, Maurice LeBrun wrote:
> | > These are fine with me personally, yes even with_double.  Although
> | > in that case you may trip up some people who are used to the old
> | > behavior.
> |
> | OK.  I left with-double=3Dno.
>
> I don't think that Maurice really meant to keep with-double=3Dno.

Yes, from his words above he was clearly willing to go along.  But the only
comment I heard at the point when I made the changes this morning was his
slightly negative one so I decided to be conservative with the change.

However, now that you think double is a better default I am willing to go
along with that.  If nobody else objects in the next few hours I will
just go ahead and make the change.

> | That still leaves the original question about with_gcc.  I think it
> | is completely ignored at the moment so it should probably just be
> | dropped unless somebody wants to maintain it.
>
> Yes, I also think so. gcc must be the default. For other situations
> CC=3Dxx ./configure should do the trick.

OK.  Unless I hear other objections in the next few hours, I will remove
with_gcc.


>
> |
> | > [Out of order...]
> | > with_shlib should probably also stay if it is currently supported
> | > -- does anyone know?
> |
> | I believe the standard --enable-shared completely replaces
> | with_shlib. Thus, I don't think we should maintain a parallel
> | with_shlib. If everyone agrees, I will make the appropriate changes
> | in configure.ac and sysloc.in to replace the with_shlib logic with
> | the corresponding enable-shared logic.
>
> Isn't it the same thing?

No.  There is configure logic in sysloc.in that now depends on with_shlib,
but that option is set by the user independently of --enable-shared.  So
currently it is a mess.

I think the best way to deal with this is to completely remove with_shlib
and change over to enable_shared logic in sysloc.in.  If nobody objects
to this change I will do it in the next few hours.

Alan
__________________________
Alan W. Irwin
email: ir...@be...
phone: 250-727-2902

Astronomical research affiliation with Department of Physics and Astronomy,
University of Victoria (astrowww.phys.uvic.ca).

Programming affiliations with the Canadian Centre for Climate Modelling and
Analysis (www.cccma.bc.ec.gc.ca) and the PLplot scientific plotting softwar=
e
package (plplot.org).

__________________________

Linux-powered Science
__________________________

Re: plgriddata and Directory Changes README! [was: Re: [Plplot-devel] Next release]

[<jc...@fe...>]

On Friday 28 February 2003 20:07, Rafael Laboissiere wrote:
| * Jo=E3o Cardoso <jc...@fe...> [2003-02-28 19:27]:
| > There is already a lib directory, but holds font and map files. We
| > could take the opportunity to change this, as the files in that
| > directory all have cvs version 1.1, and no cvs history will be
| > lost.
| > They could be moved to a new directory called "data", after all
| > they are installed in $(prefix)/...../data/.
| > Only in the main Makefile.am a data subdirectory needs to be added
| > in src_dirs to accomplish this.
| > Does everybody agrees with this?
|
| Okay for me.
|
| > | Is that really necessary?  Can't you cast the values returned by
| > | nn/csa with PLFLOAT?
| >
| > The default plplot configuration will be with-double, as everybody
| > agreed on this (even if by omission).
|
| My recollection is that Maurice objected.

Yes, it slipped from my mind until I saw Alan's last e-mail. I have just=20
now replied to it.

| > If the user specified that he wants floats instead, he must have
| > its own reasons, and must have what he wants (and deserves). Don't
| > you agree?
|
| I agree, but you did not answer my question, probably because I made
| a typo. Here I go again: Can't you cast the values returned by nn/csa
| with PLFLT?

No, data arrays are passed by reference. With nn/csa/qhull array data=20
flows in both directions.
If the user wants to use floats because he has huge amounts of float=20
data we would have convert it all back and forth. And the main user=20
concern, memory space, would be jeopardized anyway.=20

Joao

Re: [Plplot-devel] Configuration defaults

[<jc...@fe...>]

On Friday 28 February 2003 19:51, Alan W. Irwin wrote:
| On Thu, 27 Feb 2003, Maurice LeBrun wrote:
| > These are fine with me personally, yes even with_double.  Although
| > in that case you may trip up some people who are used to the old
| > behavior.
|
| OK.  I left with-double=no.

I don't think that Maurice really meant to keep with-double=no. After 
all, when we change the API the consequences are much worse, and that 
don't inhibits us from doing it.
And we have the configure summary, that clearly states the float/double 
status.
Also, with the new nn/csa/qhull libraries, the better is to use double 
by default.
So, I vote for with-double=yes.

Maurice, some time ago you argued that we should not drop float support. 
Could you please "argue" again? Sourceforge mailing lists don't have a 
search facility, what makes difficult to find something we know is 
there.

|
| However, in my recent commit I changed with_freetype,
| enable_dyndrivers, enable_octave, and enable_ntk to yes.
|
| I also replaced the non-working has_x11 logic with the have_x logic
| that is implemented by the AC_PATH_XTRA.  I tested that new logic by
| trying the standard configure option --without-x which for the first
| time now does the right thing (no tk, no xwin).  I assume from this
| good test, that if anybody ever tries to configure on a system
| without X, that the right thing will be done in that case as well,
| but I don't have access to such a system to test this.
|
| > with_debug
| > with_opt
| > with_profile
| > with_warn
| >
| > Should all stay, as I will get to support them again one of these
| > days. Probably won't be until April.
|
| OK.

And meanwhile should they appear when we do ./configure --help? It will 
only confuse users! I think that they should hibernate until Maurice 
revives them.
./configure --help says to set CC, CFLAGS, etc.

| > with_dbmalloc can go AFAIAC.  These days we have access to better
| > tools (e.g. valgrind) and it has been some years since I've used a
| > debugging malloc.
|
| OK. I removed dbmalloc.
|
| I also removed with_fseek because I couldn't find a trace of any use
| of it.
|
| I left with_pkgdir= alone because apparently it is a part of the tcl
| configuration logic.  (I hope that logic gets cleaned up before
| 5.2.1, but that is a subject for another day.)
|
| That still leaves the original question about with_gcc.  I think it
| is completely ignored at the moment so it should probably just be
| dropped unless somebody wants to maintain it.

Yes, I also think so. gcc must be the default. For other situations 
CC=xx ./configure should do the trick.

|
| > [Out of order...]
| > with_shlib should probably also stay if it is currently supported
| > -- does anyone know?
|
| I believe the standard --enable-shared completely replaces
| with_shlib. Thus, I don't think we should maintain a parallel
| with_shlib. If everyone agrees, I will make the appropriate changes
| in configure.ac and sysloc.in to replace the with_shlib logic with
| the corresponding enable-shared logic.

Isn't it the same thing?

| Does anyone else have additional configure defaults they want to
| change?

Not that I can remember,
Joao

|
| Alan
|
| __________________________
| Alan W. Irwin
| email: ir...@be...
| phone: 250-727-2902
|
| Astronomical research affiliation with Department of Physics and
| Astronomy, University of Victoria (astrowww.phys.uvic.ca).
|
| Programming affiliations with the Canadian Centre for Climate
| Modelling and Analysis (www.cccma.bc.ec.gc.ca) and the PLplot
| scientific plotting software package (plplot.org).
|
| __________________________
|
| Linux-powered Science
| __________________________
|
|
|
| -------------------------------------------------------
| This sf.net email is sponsored by:ThinkGeek
| Welcome to geek heaven.
| http://thinkgeek.com/sf
| _______________________________________________
| Plplot-devel mailing list
| Plp...@li...
| https://lists.sourceforge.net/lists/listinfo/plplot-devel

Re: plgriddata and Directory Changes README! [was: Re: [Plplot-devel] Next release]

[Rafael L. <lab...@ps...>]

* João Cardoso <jc...@fe...> [2003-02-28 19:27]:

> There is already a lib directory, but holds font and map files. We could 
> take the opportunity to change this, as the files in that directory all 
> have cvs version 1.1, and no cvs history will be lost.
> They could be moved to a new directory called "data", after all they are 
> installed in $(prefix)/...../data/.
> Only in the main Makefile.am a data subdirectory needs to be added in 
> src_dirs to accomplish this.
> Does everybody agrees with this?

Okay for me.

> | Is that really necessary?  Can't you cast the values returned by
> | nn/csa with PLFLOAT?
> 
> The default plplot configuration will be with-double, as everybody agreed
> on this (even if by omission).

My recollection is that Maurice objected.

> If the user specified that he wants floats instead, he must have its own
> reasons, and must have what he wants (and deserves). Don't you agree?

I agree, but you did not answer my question, probably because I made a typo.
Here I go again: Can't you cast the values returned by nn/csa with PLFLT?

> | Looking at the License statement in the README file and Pavel's
> | comments above, I think that nn and csa can be considered free
> | software.  If you are really concerned about this, we can ask in the
> | mailing list debian-legal.
> 
> If it's OK for you is fine for me :)

Let us assume that nn/csa is free software.  We must though remind to add a
note at the PLplot copyright file.

-- 
Rafael

Re: [Plplot-devel] Configuration defaults

[Alan W. I. <ir...@be...>]

On Thu, 27 Feb 2003, Maurice LeBrun wrote:

> These are fine with me personally, yes even with_double.  Although in that
> case you may trip up some people who are used to the old behavior.

OK.  I left with-double=no.

However, in my recent commit I changed with_freetype, enable_dyndrivers,
enable_octave, and enable_ntk to yes.

I also replaced the non-working has_x11 logic with the have_x logic that is
implemented by the AC_PATH_XTRA.  I tested that new logic by trying the
standard configure option --without-x which for the first time now does the
right thing (no tk, no xwin).  I assume from this good test, that if anybody
ever tries to configure on a system without X, that the right thing will be
done in that case as well, but I don't have access to such a system to test
this.

>
> with_debug
> with_opt
> with_profile
> with_warn
>
> Should all stay, as I will get to support them again one of these days.
> Probably won't be until April.

OK.

>
> with_dbmalloc can go AFAIAC.  These days we have access to better tools
> (e.g. valgrind) and it has been some years since I've used a debugging
> malloc.

OK. I removed dbmalloc.

I also removed with_fseek because I couldn't find a trace of any use of it.

I left with_pkgdir= alone because apparently it is a part of the tcl
configuration logic.  (I hope that logic gets cleaned up before 5.2.1, but
that is a subject for another day.)

That still leaves the original question about with_gcc.  I think it is
completely ignored at the moment so it should probably just be dropped
unless somebody wants to maintain it.

> [Out of order...]
> with_shlib should probably also stay if it is currently supported -- does
> anyone know?

I believe the standard --enable-shared completely replaces with_shlib. Thus,
I don't think we should maintain a parallel with_shlib. If everyone agrees,
I will make the appropriate changes in configure.ac and sysloc.in to replace
the with_shlib logic with the corresponding enable-shared logic.

Does anyone else have additional configure defaults they want to change?

Alan

__________________________
Alan W. Irwin
email: ir...@be...
phone: 250-727-2902

Astronomical research affiliation with Department of Physics and Astronomy,
University of Victoria (astrowww.phys.uvic.ca).

Programming affiliations with the Canadian Centre for Climate Modelling and
Analysis (www.cccma.bc.ec.gc.ca) and the PLplot scientific plotting software
package (plplot.org).

__________________________

Linux-powered Science
__________________________

plgriddata and Directory Changes README! [was: Re: [Plplot-devel] Next release]

[<jc...@fe...>]

On Friday 28 February 2003 18:18, Rafael Laboissiere wrote:
| * Jo=E3o Cardoso <jc...@fe...> [2003-02-28 15:57]:
| > Rafael, if you have time, could you setup things for this :-?
| > Then I only have to commit src/plgriddata.c, examples/c/x21c.c, and
| > the changes to nn/delauny.c (and the patch file itself).
|
| Please, commit all the necessary files to the CVS repository,

OK, I will do it ASAP.

| including the nn and csa ones (should we create a lib/ directory for
| that in the CVS tree?

There is already a lib directory, but holds font and map files. We could=20
take the opportunity to change this, as the files in that directory all=20
have cvs version 1.1, and no cvs history will be lost.
They could be moved to a new directory called "data", after all they are=20
installed in $(prefix)/...../data/.
Only in the main Makefile.am a data subdirectory needs to be added in=20
src_dirs to accomplish this.
Does everybody agrees with this?

| in this case, you could put the files under
| lib/nn/ and lib/csa/).  I will then make the changes in configure.ac
| and create the necessary Makefile.am's.
|
| > (1)-Another point that I have not digged into is the float/double
| > issue. While Qhull can be compiled for both float or double (we can
| > check for this at configure time), nn and csa use doubles; one
| > might have to disable nn/csa if the user wants to compile plplot
| > with floats.
|
| Is that really necessary?  Can't you cast the values returned by
| nn/csa with PLFLOAT?

The default plplot configuration will be with-double, as everybody=20
agreed on this (even if by omission).
If the user specified that he wants floats instead, he must have its own=20
reasons, and must have what he wants (and deserves).
Don't you agree?

|
| > (2)- Please take a look at the license. It looks a bit confusing
| > for a civilian.  When I ask Pavel (nn/csa author's) for a
| > clarification, he
| >
| > said:
| > >This licence does _not_ explicitely demand
| > >
| > > <BEGIN excerpt from GPL>
| > > b) You must cause any work that you distribute or publish, that
| > > in whole or in part contains or is derived from the Program or
| > > any part thereof, to be licensed as a whole at no charge to all
| > > third parties under the terms of this License.
| > > <END>
| > >
| > > Therefore I would be happy if we quietly implied that "BOTH
| > > SOURCE AND OBJECT CODE ARE MADE FREELY AVAILABLE WITHOUT CHARGE"
| > > refers to the _modified_ code only, not the whole application it
| > > is used in. I _think_ it should be legally safe to use this code
| > > in commercial applications as long as the modified code retains
| > > copyright and is made public.
|
| Looking at the License statement in the README file and Pavel's
| comments above, I think that nn and csa can be considered free
| software.  If you are really concerned about this, we can ask in the
| mailing list debian-legal.

If it's OK for you is fine for me :)

Joao

Re: [Plplot-devel] Next release

[Rafael L. <lab...@ps...>]

* João Cardoso <jc...@fe...> [2003-02-28 15:57]:

> Rafael, if you have time, could you setup things for this :-?
> Then I only have to commit src/plgriddata.c, examples/c/x21c.c, and the 
> changes to nn/delauny.c (and the patch file itself).

Please, commit all the necessary files to the CVS repository, including the
nn and csa ones (should we create a lib/ directory for that in the CVS tree?
in this case, you could put the files under lib/nn/ and lib/csa/).  I will
then make the changes in configure.ac and create the necessary
Makefile.am's.

> (1)-Another point that I have not digged into is the float/double issue. 
> While Qhull can be compiled for both float or double (we can check for 
> this at configure time), nn and csa use doubles; one might have to disable
> nn/csa if the user wants to compile plplot with floats.

Is that really necessary?  Can't you cast the values returned by nn/csa with
PLFLOAT?

> (2)- Please take a look at the license. It looks a bit confusing for a 
> civilian.  When I ask Pavel (nn/csa author's) for a clarification, he 
> said:
> 
> >This licence does _not_ explicitely demand
> >
> > <BEGIN excerpt from GPL>
> > b) You must cause any work that you distribute or publish, that in whole 
> >    or in part contains or is derived from the Program or any part thereof, to
> >    be licensed as a whole at no charge to all third parties under the terms of
> >    this License.
> > <END>
> >
> > Therefore I would be happy if we quietly implied that "BOTH SOURCE AND 
> > OBJECT CODE ARE MADE FREELY AVAILABLE WITHOUT CHARGE"
> > refers to the _modified_ code only, not the whole application it is used
> > in. I _think_ it should be legally safe to use this code in commercial
> > applications as long as the modified code retains copyright and is 
> > made public.

Looking at the License statement in the README file and Pavel's comments
above, I think that nn and csa can be considered free software.  If you are
really concerned about this, we can ask in the mailing list debian-legal.

-- 
Rafael

Re: [Plplot-devel] Next release

[<jc...@fe...>]

On Friday 28 February 2003 15:09, Rafael Laboissiere wrote:
| * Jo=E3o Cardoso <jc...@fe...> [2003-02-28 14:36]:
| > hmm, I was thinking to put the new libraries object files into
| > libplplot, to easy the linking of users apps. But if we install
| > libnn and libcsa the user can use the libraries independently. What
| > do you think about this?
|
| I would prefer to generate separate shared libraries for nn and csa,
| if possible.  In this case, there will be separation both in the
| source tree and in the final result.

OK, but configure should be run only once, checking for Qhull, defining=20
a HAVE_QHULL in config.h, and them making csa and (conditionally) nn.=20
libnn and (conditionally) libcsa and (conditionally) libqhull must then=20
be added to the dependency libs of libplplot, the same that is done=20
with libfreetype(1).

Rafael, if you have time, could you setup things for this :-?
Then I only have to commit src/plgriddata.c, examples/c/x21c.c, and the=20
changes to nn/delauny.c (and the patch file itself).

In my sources, which I don't have at hand, I only use from the original=20
tarball the *.[ch] files plus README, where the license is (2).=20
triangle.* doesn't need (and shouldn't) be committed.

| A "bonus track", as you mention, is the fact that users can link
| separately with the libraries.  For that, I can imagine generating a
| Debian package called, say, plplot-nn-csa.
|
| An aside note: have you came across a "bug" in nnpi.c where hash.c is
| included instead of hash.h?

I don't have the sources at hand. Also, I worked with nn and csa=20
versions which are not the current ones.

Joao

(1)-Another point that I have not digged into is the float/double issue.=20
While Qhull can be compiled for both float or double (we can check for=20
this at configure time), nn and csa use doubles; one might have to=20
disable nn/csa if the user wants to compile plplot with floats.

(2)- Please take a look at the license. It looks a bit confusing for a=20
civilian.  When I ask Pavel (nn/csa author's) for a clarification, he=20
said:

>This licence does _not_ explicitely demand
>
><BEGIN excerpt from GPL>
>b) You must cause any work that you distribute or publish, that in=20
whole=20
>or in part contains or is derived from the Program or any part thereof,=20
>to be licensed as a whole at no charge to all third parties under the=20
>terms of this License.=20
><END>
>
>Therefore I would be happy if we quietly implied that "BOTH SOURCE AND=20
>OBJECT CODE ARE MADE FREELY AVAILABLE WITHOUT CHARGE"
>refers to the _modified_ code only, not the whole application it is=20
used
> in. I _think_ it should be legally safe to use this code in commercial
> applications as long as the modified code retains copyright and is=20
made
> public.

Re: [Plplot-devel] Next release

[Rafael L. <lab...@ps...>]

* João Cardoso <jc...@fe...> [2003-02-28 14:36]:

> hmm, I was thinking to put the new libraries object files into 
> libplplot, to easy the linking of users apps. But if we install libnn 
> and libcsa the user can use the libraries independently. What do you 
> think about this?

I would prefer to generate separate shared libraries for nn and csa, if
possible.  In this case, there will be separation both in the source tree
and in the final result.  

A "bonus track", as you mention, is the fact that users can link separately
with the libraries.  For that, I can imagine generating a Debian package
called, say, plplot-nn-csa.

An aside note: have you came across a "bug" in nnpi.c where hash.c is
included instead of hash.h?

-- 
Rafael

[Plplot-devel] Re: [Plplot-cvs] help asked

[Alan W. I. <ir...@be...>]

On Fri, 28 Feb 2003, mvda wrote:

> I recently installed PLPLOt and tried to get it working in windows 95 and
> absoft 80 environment. However when running one of the "examples programs"
> working I get stopped by the message "Unable to open Font file".
> Has anybody encountered this problem earlier ?
> how was it solved  ?
>
> thanks in advance
>
> mark van der auweraer

I am redirecting this to the correct list.

PLplot requires the font files to be in the installed location.  So I believe
you need to install PLplot (including the fonts) before the examples will work.
For Unix and Linux you install by running the command 'make install'.  I
have no windows experience with PLplot, but I notice some documentation
in sys/win32/msdev that may help you out.

Alan
__________________________
Alan W. Irwin
email: ir...@be...
phone: 250-727-2902

Astronomical research affiliation with Department of Physics and Astronomy,
University of Victoria (astrowww.phys.uvic.ca).

Programming affiliations with the Canadian Centre for Climate Modelling and
Analysis (www.cccma.bc.ec.gc.ca) and the PLplot scientific plotting software
package (plplot.org).

__________________________

Linux-powered Science
__________________________