headers
Jaroslav Hajek | 1 Aug 2008 08:52
Picon

Re: box "off" fails

On Thu, Jul 31, 2008 at 9:01 PM, John W. Eaton <jwe <at> bevo.che.wisc.edu> wrote:
> On 31-Jul-2008, Jaroslav Hajek wrote:
>
> | OK, I transplanted this into the release-3-0-x repo
> | <http://hg.tw-math.de/release-3-0-x/>
>
> It seems odd to me that a change that was just made in this archive is
> not the latest entry in the ChangeLog.  I've been moving the ChangeLog
> entries to the top of the ChangeLog file when I import the changes,
> but maybe that is not really necessary, I don't know.
>

My first thought was to put the dates in order. Since your suggestion
is equally logical and easier, I guess I'll follow it from now on.

> | Any more suggestions?
>
> Here are some recent patches that are candidates for including in
> 3.0.x:
>
>  ea3cd9791703  octMakefile.in (DISTFILES): add "missing" to the list
>  b6d4c644b4b6  Fmkdir: improve compatibility
>  a73b80cd1f10  allow empty matrix by cell (or struct) concatentation
>

I applied all of them. I have also merged the release-3-0-x and
default branches in the repo.

> There may be others.
>
(Continue reading)

Permalink | Reply | 0 comments |
headers
Jaroslav Hajek | 1 Aug 2008 08:57
Picon

Re: box "off" fails

On Thu, Jul 31, 2008 at 9:13 PM, John W. Eaton <jwe <at> bevo.che.wisc.edu> wrote:
> On 31-Jul-2008, Jaroslav Hajek wrote:
>
> | OK, I transplanted this into the release-3-0-x repo
> | <http://hg.tw-math.de/release-3-0-x/>
>
> OK, now that this is available, I will keep a local copy of this
> archive on my system now, and strip the release-3-0-x branch from the
> main development archive that I use.
>
> | I'll start managing release-3-2-x as soon as John tags a suitable
> | initial revision in the main archive (which was his suggestion).
>
> OK.  I assume the release-3-2-x branch will just start as a clone of
> the main development archive at the point when there is a 3.2.0
> release, or perhaps we would decide to create the release branch
> slightly before the actual release, then just merge changes for the
> release in that archive, possibly also copying them to other branches
> as well (main, release-3-0-x, wherever appropriate).
>
> | btw, I did this by first updating my repo (highegg) and then using "hg
> | transplant" to get the changeset into release-3-0-x. Since I intend to
> | use the highegg repo for other stuff, it seems I could use a "pristine
> | clone" repo. (I could, of course, transplant directly from web, but
> | working with local repo is more convenient). I think that Michael and
> | Shai also used a local clone of John's main repo for the graphics
> | development. Perhaps there could be one for everyone?
>
> I'm not sure I understand what you mean by "one for everyone".
>
(Continue reading)

Permalink | Reply | 1 comment |
headers
Jaroslav Hajek | 1 Aug 2008 09:15
Picon

Re: box "off" fails

On Thu, Jul 31, 2008 at 10:01 PM, dbateman <dbateman <at> free.fr> wrote:
>
>
>
> John W. Eaton wrote:
>>
>> Here are some recent patches that are candidates for including in
>> 3.0.x:
>>
>>   ea3cd9791703  octMakefile.in (DISTFILES): add "missing" to the list
>>   b6d4c644b4b6  Fmkdir: improve compatibility
>>   a73b80cd1f10  allow empty matrix by cell (or struct) concatentation
>>
>> There may be others.
>>
>
> I'd also suggest
>
> 736124a4fa3d lasterr, lasterror: unwind-protect error_state
>
> and the patch in
>
> http://www.nabble.com/Re%3A-Bug-492076%3A-octave3.0%3A-incorrect-axis-location-properties-p18672112.html
>

I've applied both, adding a ChangeLog entry to the second one. Now I
see what John meant when he said that ChangeLogs are a constant source
of pain for him. If it weren't for them, all patches would apply
cleanly.
(Continue reading)

Permalink | Reply | 0 comments |
headers
Thomas Weber | 1 Aug 2008 13:09
Picon

Re: box "off" fails

On 01/08/08 08:57 +0200, Jaroslav Hajek wrote:
> On Thu, Jul 31, 2008 at 9:13 PM, John W. Eaton <jwe <at> bevo.che.wisc.edu> wrote:
> Um, sorry - don't mind that, the suggestion was actually dedicated to
> Thomas. I suggested that he kept a pure clone of your main repo
> amongst the other repos, so that people can update their repos from a
> local common source (to save space on the server). But now I see that
> it only makes sense when working directly on the remote server, which
> is not what one will typically do (though I did it yesterday). So
> forget about this.

You mean everyone needs a pristine copy of John's repo on the server? Hmm,
neither space nor bandwidth is currenly a problem. But if the need arises, I
already have a non-public clone of John's repository on the server in my
home directory.

"hg pull -u" could be implemented via a short script that has a set user id
(s) bit, so it's possible without much work.

Just ask me if you think it's usefull.

	Thomas
Permalink | Reply | 0 comments |
headers
David Dean | 4 Aug 2008 05:02
Picon
Gravatar

strtok should default to using all white space characters to match MATLAB

--------
Bug report for Octave 3.0.0 configured for i486-pc-linux-gnu

Description:
-----------

As per title. The default call of strtok only considers spaces to
deliminate tokens. The deliminator should default to be something like
" \t\n" instead of just being " ".

Repeat-By:
---------

> a ="doc\tb"
> [c,d] = strtok(a)
c = doc b
d =
> [c,d] = strtok(a," \t")
c = doc
d =    b

Configuration (please do not edit this section):
-----------------------------------------------

uname output:     Linux olivaw 2.6.24-20-generic #1 SMP Mon Jul 28
13:49:52 UTC 2008 i686 GNU/Linux
configure opts:   '--host=i486-linux-gnu' '--build=i486-linux-gnu'
'--prefix=/usr' '--datadir=/usr/share' '--libdir=/usr/lib'
'--libexecdir=/usr/lib' '--infodir=/usr/share/info'
'--mandir=/usr/share/man' '--with-blas=-lblas-3gf'
(Continue reading)

Permalink | Reply | 3 comments |
headers
Ole Tange | 4 Aug 2008 12:30
Picon

Enhancement: Give error and solution if using make instead of gmake

Host: SunOS solvm 5.10 Generic_118855-14 i86pc i386 i86pc
Source: 3.0.1
Compiler: gcc (GCC) 3.4.5

I have made a fresh install of Solaris 10 to determine the
requirements for compiling Octave 3.0.1.

After I do ./configure I did:

# make

***********************************************************
*
*  To compile Octave, you will need a recent versions of
*  the following software:
*
*    g++ (3.2.x or a more recent version)
*
*    flex (2.5.4 or a more recent version) -- required if
*    you need to recreate lex.cc from lex.l
*
*    bison (1.31 or a more recent version) -- required if
*    you need to recreate parse.cc from parse.y
*
*    gperf (3.0.1 or a more recent version) -- required if
*    you need to recreate oct-gperf.h from octave.gperf
*
*  Now would be a good time to read INSTALL.OCTAVE if
*  you have not done so already.
*
(Continue reading)

Permalink | Reply | 2 comments |
headers
Ben Abbott | 4 Aug 2008 12:50
Picon
Gravatar

Re: [CHANGESET]: strtok should default to using all white space characters to match MATLAB


On Aug 3, 2008, at 11:02 PM, David Dean wrote:

> --------
> Bug report for Octave 3.0.0 configured for i486-pc-linux-gnu
>
> Description:
> -----------
>
> As per title. The default call of strtok only considers spaces to
> deliminate tokens. The deliminator should default to be something like
> " \t\n" instead of just being " ".
>
> Repeat-By:
> ---------
>
>> a ="doc\tb"
>> [c,d] = strtok(a)
> c = doc b
> d =
>> [c,d] = strtok(a," \t")
> c = doc
> d =    b

Tab (" \t") is added to the default delimiters and an additional test  
is included.

Changeset attached.
(Continue reading)

Permalink | Reply | 2 comments |
headers
Ole Tange | 4 Aug 2008 13:40
Picon

Solaris 10: configure does not find a working sed

Host: SunOS solvm 5.10 Generic_118855-14 i86pc i386 i86pc
Source: 3.0.1
Compiler: gcc (GCC) 3.4.5

I have made a fresh install of Solaris 10 to determine the
requirements for compiling Octave 3.0.1.

After I did ./configure I did:

# gmake

***********************************************************
*
*  To compile Octave, you will need a recent versions of
*  the following software:
*
*    g++ (3.2.x or a more recent version)
*
*    flex (2.5.4 or a more recent version) -- required if
*    you need to recreate lex.cc from lex.l
*
*    bison (1.31 or a more recent version) -- required if
*    you need to recreate parse.cc from parse.y
*
*    gperf (3.0.1 or a more recent version) -- required if
*    you need to recreate oct-gperf.h from octave.gperf
*
*  Now would be a good time to read INSTALL.OCTAVE if
*  you have not done so already.
*
(Continue reading)

Permalink | Reply | 2 comments |
headers
John W. Eaton | 4 Aug 2008 16:59
Picon

Re: [CHANGESET]: strtok should default to using all white space characters to match MATLAB

On  4-Aug-2008, Ben Abbott wrote:

| 
| On Aug 3, 2008, at 11:02 PM, David Dean wrote:
| 
| > --------
| > Bug report for Octave 3.0.0 configured for i486-pc-linux-gnu
| >
| > Description:
| > -----------
| >
| > As per title. The default call of strtok only considers spaces to
| > deliminate tokens. The deliminator should default to be something like
| > " \t\n" instead of just being " ".
| >
| > Repeat-By:
| > ---------
| >
| >> a ="doc\tb"
| >> [c,d] = strtok(a)
| > c = doc b
| > d =
| >> [c,d] = strtok(a," \t")
| > c = doc
| > d =    b
| 
| 
| Tab (" \t") is added to the default delimiters and an additional test  
| is included.
|
(Continue reading)

Permalink | Reply | 1 comment |
headers
Ole Tange | 4 Aug 2008 17:44
Picon

Solaris 10: 'gmake check' dumps core

Bug report for Octave 3.0.1 configured for i386-pc-solaris2.10

Description:
-----------

I have made a fresh install of Solaris 10 to determine the
requirements for compiling Octave 3.0.1, so the requirements can be
detected reliably by configure and a failing compilation can be
avoided.

When I tried 'gmake check' it would not even start octave.

run-octave dumps core.

Repeat-By:
---------

I made a fresh install of Solaris 10. Then installed GCC, gmake,
texinfo, and gsed. Then:

./configure; gmake

Then:

gmake check

This fails:

gmake -f octMakefile check
gmake[1]: Entering directory `/octave/octave-3.0.1'
(Continue reading)

Permalink | Reply | 6 comments |

Gmane