Chechu Garguez | 29 Jul 16:41 2014
Picon

Octave 3.8.2rc2 - Linux amd64

Good evening, sorry for the delay.

Does are the results for the debian system we use.
Thanks for the effort.


debian wheezy-stable                 amd64

openblas having lapack stable                     32bit integers

---------
openblas having lapack stable
suitesparse and all other libraries from debian wheezy-stable

octave 3.8.2rc2 with jit and gui                    32bit integers


Summary:

  PASS     11597
  FAIL         0
  XFAIL        7



debian squeeze-oldstable        amd64

openblas having lapack stable
libqt4, llvm and gcc-4.8 backported from debian wheezy
suitesparse and all other libraries from debian squeeze-oldstable

octave 3.8.2rc2 with jit and gui                    32bit integers

Summary:

  PASS     11587
  FAIL         0
  XFAIL        7
  SKIPPED     10 ----> related with FLTK
John D | 29 Jul 22:39 2014
Picon

mxe-octave: ghostscript 9.07, why not 9.10


Message: 6
Date: Tue, 29 Jul 2014 10:09:19 +0200
From: sergey plotnikov <seryozha.plotnikov <at> gmail.com>
To: octave-maintainers <at> gnu.org
Subject: mxe-octave: ghostscript 9.07, why not 9.10
Message-ID: <53D756AF.3070602 <at> gmail.com>
Content-Type: text/plain; charset="iso-8859-1"; Format="flowed"

Hi all,

I've cross-built recent development version of Octave for Windows using
MXE-Octave (also, one of the recent csets) and after some testing get into a
trouble which I'm not sure caused by Octave itself. Sometimes when trying to
print a figure to pdf (print('-dpdf',..)) Octave may hang on waiting before
gs.exe completed writing into a pdf-file. But as a matter of fact gs.exe
doesn't want to finish and continues writing into pdf inflating its size
gradually approaching infinity (last time I forgot to kill gs.exe process my
one-page pdf reached 600Gb).

This bug looks quite similar to GS bug #693591
[http://bugs.ghostscript.com/show_bug.cgi?id=693591] which was fixed in
gs9.10. And according to my recent testing, use of separately installed
gs9.10 instead of default Octave gs.exe helps.

So my questions is rather easy, do you by any chance plan to switch GS
version in MXE-Octave from 9.07 to 9.10? If not, it might worth to apply a
patch fixing this bug in GS source
[http://git.ghostscript.com/?p=ghostpdl.git;a=commit;h=dfd7fc9fe152245256e43
7141dd7780c8313c971].

Sorry, if I overlooked existing discussion on this topic.

Sergey

----

Why not 9.14 ?
Patch works for me under native mingw and cross compiled on my fedora
machine
Attachment (ghost-914.patch): application/octet-stream, 17 KiB
sergey plotnikov | 29 Jul 10:09 2014
Picon

mxe-octave: ghostscript 9.07, why not 9.10

Hi all,

I've cross-built recent development version of Octave for Windows using MXE-Octave (also, one of the recent csets) and after some testing get into a trouble which I'm not sure caused by Octave itself. Sometimes when trying to print a figure to pdf (print('-dpdf',..)) Octave may hang on waiting before gs.exe completed writing into a pdf-file. But as a matter of fact gs.exe doesn't want to finish and continues writing into pdf inflating its size gradually approaching infinity (last time I forgot to kill gs.exe process my one-page pdf reached 600Gb).

This bug looks quite similar to GS bug #693591 [http://bugs.ghostscript.com/show_bug.cgi?id=693591] which was fixed in gs9.10. And according to my recent testing, use of separately installed gs9.10 instead of default Octave gs.exe helps.

So my questions is rather easy, do you by any chance plan to switch GS version in MXE-Octave from 9.07 to 9.10? If not, it might worth to apply a patch fixing this bug in GS source [http://git.ghostscript.com/?p=ghostpdl.git;a=commit;h=dfd7fc9fe152245256e437141dd7780c8313c971].

Sorry, if I overlooked existing discussion on this topic.

Sergey
Rik | 29 Jul 05:47 2014

Grid toggle from Edit menu affects all axes

7/28/14

Andreas,

In a recent changeset the FLTK Edit Menu for plots was changed so that
actions like show grid and hide grid apply to all axes.  Is this
necessary?  When I use the 'g' keyboard shortcut or the 'G' menubar it only
affects the particular axes involved.  I think there are lots of occasions
where multiple plots are displayed and you want to modify a single plot
rather than globally change all plots.

--Rik

Rik | 29 Jul 05:41 2014

Re: Edit-Grid doesn't work

On 07/28/2014 09:01 AM, octave-maintainers-request <at> gnu.org wrote:
Subject:
Re: Overhaul FLTK pan, rotate, zoom
From:
Ben Abbott <bpabbott <at> mac.com>
Date:
07/28/2014 06:34 AM
To:
Andreas Weber <octave <at> tech-chat.de>
CC:
octave-maintainers <at> gnu.org
List-Post:
<mailto:octave-maintainers <at> gnu.org>
Content-Transfer-Encoding:
quoted-printable
Precedence:
list
MIME-Version:
1.0 (Mac OS X Mail 7.3 \(1878.6\))
References:
<53D56250.4060601 <at> tech-chat.de>
In-Reply-To:
<53D56250.4060601 <at> tech-chat.de>
Message-ID:
<DA69DAED-0F8E-4509-A99F-99EDA820E62B <at> mac.com>
Content-Type:
text/plain; charset=us-ascii
Message:
1

On Jul 27, 2014, at 4:34 PM, Andreas Weber <octave <at> tech-chat.de> wrote:
> cset http://hg.savannah.gnu.org/hgweb/octave/rev/52e01aa1fe8b > > Each axes now has its own properties for interactive GUI control of pan, > rotate3d and mouse_wheel_zoom. Now it's possible to have several figures > and set pan for the 2D plot in figure x and rotate3d for the 3D plot in > figure y. There are two new pan modes: "Pan x only" and "Pan y only". > > The toolbar buttons "P" and "R" set pan and rotate3d for the last > clicked axes object or the object below the center of the canvas if none > was clicked yet. > > Try this for example: > > subplot (3, 1, 1); > sombrero (); > set (gca, "rotate3d", "on"); > subplot (3, 1, 2); > plot (sin (0:0.05:10)) > grid > set (gca, "pan", "xon") > subplot (3, 1, 3); > peaks (); > > See how each plot behaves. Try also pan, rotate and zoom. > > The legend axes can now be moved with the mouse. Try > > sombrero(); legend("A sombrero plot") > > and move the legend with the mouse around. > > -- Andy
Andy, Everything works for me ... except that selecting [Edit->Grid] from the menu (i.e. no effect). However if the axes is active, then typing "g" does toggle the grid, as does the "G" button at the bottom of the figure. Ben
Ben, Andreas,

I remarked on this before as well.  You can get it to work by causing the event loop to run.  For example, if you return to the command window and hit <RETURN> the grid will appear.  We need to use the same solution that was used for the 'g'.  In this case, I just added a drawnow () command at the end of the grid_cb function in __add_default_menu__ and it works correctly.

--Rik

David Spies | 28 Jul 18:40 2014
Picon

Sparse matrix logical indexing

I'm not sure what it's worth to send this out to the mailing list, but I've finished with sparse-matrix extraction for logical indexing.

ie
A(B) where B is logical and A is sparse behaves appropriately even when A and B are large.

http://hg.octave.org/octave-dspies/graph/

It's the one labeled: Added special-case logical-indexing function

-----

Note that although this doesn't yet handle assignment with logical indices such as A(B) = C, there's a simple workaround:

A -= A .* B;
[i,j] = find(B);
A += sparse(i,j,C)

gives the same result

Andreas Weber | 27 Jul 22:34 2014
Picon

Overhaul FLTK pan, rotate, zoom

cset http://hg.savannah.gnu.org/hgweb/octave/rev/52e01aa1fe8b

Each axes now has its own properties for interactive GUI control of pan,
rotate3d and mouse_wheel_zoom. Now it's possible to have several figures
and set pan for the 2D plot in figure x and rotate3d for the 3D plot in
figure y. There are two new pan modes: "Pan x only" and "Pan y only".

The toolbar buttons "P" and "R" set pan and rotate3d for the last
clicked axes object or the object below the center of the canvas if none
was clicked yet.

Try this for example:

subplot (3, 1, 1);
sombrero ();
set (gca, "rotate3d", "on");
subplot (3, 1, 2);
plot (sin (0:0.05:10))
grid
set (gca, "pan", "xon")
subplot (3, 1, 3);
peaks ();

See how each plot behaves. Try also pan, rotate and zoom.

The legend axes can now be moved with the mouse. Try

sombrero(); legend("A sombrero plot")

and move the legend with the mouse around.

-- Andy

Richard Crozier | 25 Jul 11:02 2014
Picon
Picon

Re: How to add new files to gui?

On 24/07/14 16:00, John W. Eaton wrote:
> On 07/24/2014 07:18 AM, Richard Crozier wrote:
>> Dear list,
>>
>> How do I go about adding new source files to the GUI?
>>

<snip>

>>
>> It's probably obvious, but could someone point out what I'm missing here
>> please?
>
> Sorry, I'm not sure what the problem is, but I have a small request
> about your change.  Could you please follow the convention we have for
> file names and use hyphen instead of underscore in the names?  I don't
> mean to nitpick, but I'd prefer to keep the naming scheme for Octave
> source files somewhat consistent.
>

yes, I am happy to change this, and there are plenty of other style 
rules etc. to be fixed I'm sure before getting to the patch stage, at 
this point I was just hacking around.

> Also, if you'd post a complete current version of your proposed changes,
> I'd be willing to look at it to see why the build is not working.
>
> Thanks,
>
> jwe
>
>

Thanks I might try and get a patch together next week, at the moment the 
change discussed here is now mixed in with another, which seems to be 
causing my build to fail with a different error, and I'm experimenting 
with mercurial queues to untangle them.

Richard

--

-- 
The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.

c. | 25 Jul 06:38 2014
Picon

octave.sourceforge.net

Hi all,

The website at octave.sourceforge.net seems to have disappeared.

Currently http://octave.sourceforge.net points to the standard placeholder
for project webspace on Sourceforge, so this does not seem a connection problem 
but rather a problem with missing files.

Indeed by connecting to the ebsite I can see the whole "htdocs" folder has disappeared.

Does anyone know what may have caused this? Do we have a backup?

c.

Carnë Draug | 24 Jul 22:42 2014

Octave Forge down (content removed)

Hi

I have, accidentally, removed the web space of Octave Forge. The
package tarballs and repositories are still available for download,
the only thing that disappeared was the online documentation. Since
pkg works by parsing the html of each package documentation, that
won't work either. I am trying to restore it but am dubious that this
will be complete today. I will send a separate mail to the help
mailing list with how to live without the Octave Forge website.

Carnë

Rik | 24 Jul 20:14 2014

Re: regression in statistics() function

On 07/24/2014 10:41 AM, octave-maintainers-request <at> gnu.org wrote:
Subject:
Fwd: [Pkg-octave-devel] Bug#755882: octave: statistics() does not work because of skewness and kurtosis
From:
Juan Pablo Carbajal <ajuanpi+dev <at> gmail.com>
Date:
07/24/2014 09:29 AM
To:
Octave Maintainers List <octave-maintainers <at> octave.org>
List-Post:
<mailto:octave-maintainers <at> gnu.org>
Content-Transfer-Encoding:
quoted-printable
Precedence:
list
MIME-Version:
1.0
References:
<E1XADm4-0000mB-Ls <at> tucano.isti.cnr.it>
In-Reply-To:
<E1XADm4-0000mB-Ls <at> tucano.isti.cnr.it>
Message-ID:
<CABDtPkQL_Jh+eRUkD8Lz1xVoc4cF1+J6XZTpzEhsU8ZJgmcbGQ <at> mail.gmail.com>
Content-Type:
text/plain; charset=UTF-8
Message:
2

>From the debian packaging group. ---------- Forwarded message ---------- From: Francesco Potortì <Potorti <at> isti.cnr.it> Date: Thu, Jul 24, 2014 at 9:49 AM Subject: [Pkg-octave-devel] Bug#755882: octave: statistics() does not work because of skewness and kurtosis To: Debian Bug Tracking System <submit <at> bugs.debian.org> Package: octave Version: 3.8.1-3 Severity: normal signature() and kurtosis() have changed their calling convntions, but statistics.m has not been updated. Here is a patch: --- /usr/share/octave/3.8.1/m/statistics/base/statistics.m~ 2014-05-10 11:24:38.000000000 +0200 +++ /usr/share/octave/3.8.1/m/statistics/base/statistics.m 2014-07-24 09:46:43.197972678 +0200 <at> <at> -61,7 +61,7 <at> <at> function stats = statistics (x, dim) emp_inv = quantile (x, [0.25; 0.5; 0.75], dim, 7); stats = cat (dim, min (x, [], dim), emp_inv, max (x, [], dim), mean (x, dim), - std (x, [], dim), skewness (x, dim), kurtosis (x, dim)); + std (x, [], dim), skewness (x, [], dim), kurtosis (x, [], dim)); endfunction

I made the change on the stable branch and also added a %!test to catch this.  See changeset http://hg.savannah.gnu.org/hgweb/octave/rev/ab5983dc0587.

It's best to put bug fixes on the Savannah tracker though so they don't get lost.

--Rik

Gmane