[Libreoffice] Some questions about the difference between LO to go-oo.

2010-10-15 Thread Nadav Vinik
Hello

Without the change name/website I don't notice any changes.

Are there more full time developers?

Are there more companies that donate developers for that?

Are the releases of the LO depend on Oracle Open Office like LO? (since also
LO is open office version with git of patch differences)

When should LO 3.3 release?

Thanks for your answers

Nadav
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice


Re: [Libreoffice] Error Building, vcl segmentation fault

2010-10-15 Thread René Kjellerup
Hi,

I read on the Linux game publishing blog something about forcing gcc to
resolve of the link library location, unfortunate the blog is down for the
time being. However looking at gcc manual, it could have been the -Bprefix
option.

Hope that it might help.

The Best
René
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice


Re: [Libreoffice] C{,PP}FLAGS ignored in sal

2010-10-15 Thread David Tardon
On Thu, Oct 14, 2010 at 09:33:45AM +0100, Caolán McNamara wrote:
 On Thu, 2010-10-14 at 08:37 +0200, Thomas Klausner wrote:
  On Wed, Oct 13, 2010 at 05:25:59PM +0200, Thomas Klausner wrote:
   I don't see my CFLAGS nor CPPFLAGS, which would make it find boost, in
   the gcc arguments.
   How can I make them used in this part of the build?
 
 I don't think we have any proper support to pass CFLAGS and CPPFLAGS
 down to everything correctly. You can try using ARCH_FLAGS as a route
 to pass extra flags down to the final compiler line.
 

There is ENVCFLAGS for flags to add to $CC and ENVCFLAGSCXX for flags
to add to $CXX .

D.
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice


Re: [Libreoffice] [LibreOffice][Patch] Replace ASSERT and DBG_ASSERT with OSL_ENSURE + some translations, comment removal

2010-10-15 Thread David Tardon
On Thu, Oct 14, 2010 at 09:11:26AM +0100, Caolán McNamara wrote:
 On Wed, 2010-10-13 at 22:34 +0200, Jacek Wolszczak wrote:
  http://wiki.services.openoffice.org/wiki/Debug_Levels. 
  
  Some additional questions. What should I replace DBG_ERROR with?
 
 Hmm, DBG_ERROR takes no condition, and just a message, I suggest that 
 DBG_ERROR(horrible failure);
 can become
 OSL_ENSURE(false, horrible failure);

Yeah, that's the path I've taken in the past. Additionally, I suggest to
turn both DBG_WARNING and DBG_TRACE (not that I remember ever seeing
either of them :) into OSL_TRACE.

 
  Should RTL_LOGFILE_CONTEXT_AUTHOR as in http://bit.ly/buiBlA be updated
  somehow as it contains a lot of hardcoded e-mail addresses?
 
 *shrug*, nah, leave 'em alone I suggest.
 

Given the description of rtl::Logfile:

The intended use for class Logfile is to write time stamp information
for profiling purposes. 

Profiling output should only be generated for a special product version of 
OpenOffice
which is compiled with a defined preprocessor symbol 'TIMELOG'.  
Therefore we have provided a set of macros that uses the class Logfile only if
this symbol is defined.  If the macros are not sufficient, i.e. you need more
then three arguments for a printf style message, then you have to insert an
#ifdef TIMELOG/#endif brace yourself.

Additionally the environment variable RTL_LOGFILE has to be defined in order to 
generate
logging information. If the variable is not empty, it creates a file with the 
name
$(RTL_LOGFILE)_$(PID).log, where $(PID) is the process id of the running 
process.
It can be used as a run time switch for enabling or disabling the logging.
Note that this variable is evaluated only once at the first attempt to write a 
message.

The class LogFile collects runtime data within its constructor and destructor. 
It can be
used for timing whole functions. 
If you want to write timing data without context you can use the 
RTL_LOGFILE_TRACE-macros
which are defined inside rtl/logfile.h.

The class LogFile should not be used directly, instead use the 
RTL_LOGFILE_CONTEXT/
RTL_LOGFILE_TRACE-macros.

, I suggest to scrap them entirely. There are far better ways to profile
code, like gprof or callgrind (esp. combined with kcachegrind).


OT: This comment

// POLICY: Don't use RTL_LOGFILE_PRODUCT_* for your personal logging 
information.
// Be aware that these logs make it into the product shipped to 
customers.
// If you have good reasons for doing this, please contact product 
management.

really amused me :)

D.
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice


Re: [Libreoffice] Error Building, vcl segmentation fault

2010-10-15 Thread René Kjellerup
A slight correction, a bit of further searching and I found the correct
linker option, namely ld's rpath

Used to force the search path of the dynamic linking
(from gcc: $ gcc -Wl,-rpath=path to the local port of the stl )
On Oct 15, 2010 1:04 PM, René Kjellerup rk.katana.st...@gmail.com wrote:
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice


Re: [Libreoffice] Some questions about the difference between LO to go-oo.

2010-10-15 Thread Michael Meeks
Hi Nadav,

On Fri, 2010-10-15 at 09:00 +0200, Nadav Vinik wrote:
 Without the change name/website I don't notice any changes.

Well - that's not really so, we have:

* a different, flattened code-base
+ [ yes we have some legacy patches still but they
   are being slowly merged where appropriate ]
* many more contributers - lots of new volunteers are doing
  great work every day, and more arriving and making a
  difference
* lots of translators, QA, and well-wishers from the
  old OO.o community
* oh - and a brand that 'sucks less'(TM), and the new
  website you noticed ;-)

how deeply are you looking ?

 Are there more full time developers?

Yes, but far more importantly, an extraordinary explosion in volunteer
developers - many of them new to the project: which is most exciting.

 Are there more companies that donate developers for that?

Sure - RedHat is fully involved eg.

 Are the releases of the LO depend on Oracle Open Office like LO?
 (since also LO is open office version with git of patch differences)

We will include code from Oracle if it makes sense; of course - why
not, this is Free Softare right ? :-)

 When should LO 3.3 release?

Well we're working towards a first release in the next handful of
weeks.

HTH,

Michael.

-- 
 michael.me...@novell.com  , Pseudo Engineer, itinerant idiot

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice


Re: [Libreoffice] C{,PP}FLAGS ignored in sal

2010-10-15 Thread Thomas Klausner
On Fri, Oct 15, 2010 at 01:22:46PM +0200, David Tardon wrote:
 There is ENVCFLAGS for flags to add to $CC and ENVCFLAGSCXX for flags
 to add to $CXX .

That's a very good tip, thank you.

Why aren't the environment's CFLAGS and CPPFLAGS variables added
automatically?

Is there a similar LDFLAGS or LIBS variable?
 Thomas
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice


Re: [Libreoffice] [PATCH 1/3] Translate German comments in ure/xml2cmp

2010-10-15 Thread Michael Meeks
Hi Laurent,

Many thanks for the three nice patches :-) just pushed them.

All the best,

Michael.

-- 
 michael.me...@novell.com  , Pseudo Engineer, itinerant idiot

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice


Re: [Libreoffice] LO: Linux distro specific packages vs. universal build

2010-10-15 Thread Petr Mladek
Rene Engelhard píše v Čt 14. 10. 2010 v 16:48 +0200:
 Hi,
 
 On Thu, Oct 14, 2010 at 04:45:57PM +0200, Petr Mladek wrote:
  1. installation root path:
  
  + /opt/libreoffice3 - used by the universal linux build by
default
 
 Well, I don't see why we should keep that version here too but I can
 live with it being there.

You are right. We have already removed the 3-layer structure, so the
version is not longer needed. I am going to remove the it.

(I hope that nobody would shoot me ;-)

  2. user configuration:
  
  + ~/.libreoffice/3 - used by the universal linux build
  + ~/.libreoffice/3-distro - proposed alternative 
 distro-specific path, e.g. ~/.libreoffice/3-suse
  
  I tried to use the same directory but the other build did not
  start because incompatible berkley DB, see the attached
  screenshot
 
 I'd use the same user directory, personally. We use the same bdb as
 universal/vanilla anyways.

I would like to use the same dir as well. Unfortunately, we have
different system bdb on different SUSE distros...

I will try something from the Caolan's suggestions (double check fix for
#i107440, look if the db can be versioned, ...)

Best Regards,
Petr


___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice


Re: [Libreoffice] [PATCH] SAL_N_ELEMENTS changes for components

2010-10-15 Thread Kenneth Venken
Michael Meeks schreef op vr 15-10-2010 om 18:23 [+0100]:
 Hi Kenneth,
Hi Michael,
 
 On Fri, 2010-10-15 at 01:29 +0200, Kenneth Venken wrote:
  this is my first patch for Libreoffice.
  contributed under the LGPLv3+
 
   Welcome ! :-) many thanks for your work, it built nicely. What piece of
 LibreOffice most interests you for hacking on ?
Well, i just want to experience working on a large opensource project.
I'm most interested in improving compatibility with MS Office. But at
the moment i'm just getting a feel of the codebase. I'll start by
eliminating some compiler warnings. 
   Just pushed it,

   Michael.
 


___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice


Re: [Libreoffice] User mailing list

2010-10-15 Thread Gérard Fargeot


Kohei Yoshida wrote:
 
 On Thu, 2010-10-14 at 11:02 -0700, Gérard Fargeot wrote:
 Hi,
 
 Message for Dev.
 
 Please have a look at users list.
 http://the-document-foundation.969070.n3.nabble.com/OFFSET-function-don-t-works-tp1700974p1700974.html
 
 Thanks!  Just filed a bug for this:
 https://bugs.freedesktop.org/show_bug.cgi?id=30900
 
 Yes, I did modify OFFSET function just recently to fix it for external
 references.  I'll take a look at that and see what's going on there.
 
 Kohei
 
 -- 
 Kohei Yoshida, LibreOffice hacker, Calc
 kyosh...@novell.com
 
 ___
 LibreOffice mailing list
 LibreOffice@lists.freedesktop.org
 http://lists.freedesktop.org/mailman/listinfo/libreoffice
 
 
I think it is not only an OFFSET problem :
Array formula don't work https://bugs.freedesktop.org/show_bug.cgi?id=30882
I can't create a drop-down list with Data  Validity  Cell Range.

You may look at the new options :
Formula syntax, Columns and rows separators, Focus on the 1st cell of the
array instead of the last like OOo
and other options intented in make a clone of Microsoft Office.

Gérard.
-- 
View this message in context: 
http://the-document-foundation.969070.n3.nabble.com/User-mailing-list-tp1702874p1710983.html
Sent from the Dev mailing list archive at Nabble.com.
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice


[Libreoffice] [PATCH] Enable testtool (and possibly other stuff) to run with LibO

2010-10-15 Thread Giuseppe Castagno

Hi all,

the following patch should allow testtool to run with LibO.

The first patch (0001-Correct) corrects a symlink created during 
installation, I tested it under universal LibO install (pull +r 
yesterday evening).

What it does should be reviewed by packager, since I don't know what it
does under another packaging but Debian/Ubuntu (Ubuntu 9.10 being my 
current working distro).


The remaining three patches assume meaning only if the first is applied.

The third patch (0003-Changed-symlink...) corrects the behavior of 
testtool, changing a very small part of the script.


I tested testtool on LibO, it started LibO that started the test and 
then it stopped because couldn't find smoketest (not analysed why).


The second (0002-Changed-symlink) and the fourth 
(0004-Changed-symlink) change Java and cpp search path function 
accordingly.
While about Java I'm certain of the use and/or behavior, I'm not so of 
cppuhelper, but the flowing pattern seems similar.


I have some test to do using the SDK and generating an application in Java.
For cpp should be someone else checking it.

Al the patches contributed according LGPLv3+.

All of the above only if you have done nothing on the matter yet :-).

beppec56.
--
Kind Regards,
Giuseppe Castagno
Acca Esse http://www.acca-esse.eu
giuseppe.castagno at acca-esse.eu
beppec56 at openoffice.org

From c8c010c7ee78bb6a9ca317b0ad91c125b7f0314f Mon Sep 17 00:00:00 2001
From: Giuseppe Castagno giuseppe.casta...@acca-esse.eu
Date: Fri, 15 Oct 2010 11:22:59 +0200
Subject: [PATCH] Correct symlink in order to be used in SDK and testtool,
 in order to recognize the application installation path;
 only for Linux and similar platforms.

---
 desktop/scripts/soffice.sh |2 +-
 sysui/desktop/share/create_tree.sh |2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/desktop/scripts/soffice.sh b/desktop/scripts/soffice.sh
index 00c2b5c..5da4b9a 100644
--- a/desktop/scripts/soffice.sh
+++ b/desktop/scripts/soffice.sh
@@ -60,7 +60,7 @@ if [ -e ooenv ] ; then
 . ./ooenv
 fi
 
-sd_binary=`basename $0`.bin
+sd_binary=`basename $0 | sed 's/libreoffice/soffice/g'`.bin 
 
 #collect all bootstrap variables specified on the command line
 #so that they can be passed as arguments to javaldx later on
diff --git a/sysui/desktop/share/create_tree.sh b/sysui/desktop/share/create_tree.sh
index 1ed7da9..d81abf5 100644
--- a/sysui/desktop/share/create_tree.sh
+++ b/sysui/desktop/share/create_tree.sh
@@ -98,7 +98,7 @@ cp openoffice.sh ${DESTDIR}/usr/bin/${PREFIX}
 cp printeradmin.sh ${DESTDIR}/usr/bin/${PREFIX}-printeradmin
 chmod 0755 ${DESTDIR}/usr/bin/${PREFIX} ${DESTDIR}/usr/bin/${PREFIX}-printeradmin
 
-ln -sf /usr/bin/${PREFIX} ${DESTDIR}/usr/bin/libreoffice
+ln -sf ${office_root}/program/soffice ${DESTDIR}/usr/bin/libreoffice
 ln -sf /usr/bin/${PREFIX}-printeradmin ${DESTDIR}/usr/bin/libreoffice-printeradmin
 
 mkdir -p ${DESTDIR}/usr/share/mime/packages
-- 
1.6.3.3

From 017b0bf9e2cdaa78375b8508a54d83593d984f11 Mon Sep 17 00:00:00 2001
From: Giuseppe Castagno giuseppe.casta...@acca-esse.eu
Date: Fri, 15 Oct 2010 19:25:09 +0200
Subject: [PATCH] Changed symlink processing to new application name.
 Now the function will try for libreoffice in the PATH, instead of soffice.

---
 .../sun/star/lib/loader/InstallationFinder.java|2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/odk/source/com/sun/star/lib/loader/InstallationFinder.java b/odk/source/com/sun/star/lib/loader/InstallationFinder.java
index da8117e..f528532 100644
--- a/odk/source/com/sun/star/lib/loader/InstallationFinder.java
+++ b/odk/source/com/sun/star/lib/loader/InstallationFinder.java
@@ -71,7 +71,7 @@ final class InstallationFinder {
 private static final String SYSPROP_NAME =
 com.sun.star.lib.loader.unopath;
 private static final String ENVVAR_NAME = UNO_PATH;
-private static final String SOFFICE = soffice; // Unix/Linux only
+private static final String SOFFICE = libreoffice; // Unix/Linux only
 
 private InstallationFinder() {} // do not instantiate
 
-- 
1.6.3.3

From 3d3ccde3aedebab49b031fbf1126bda5d6c0 Mon Sep 17 00:00:00 2001
From: Giuseppe Castagno giuseppe.casta...@acca-esse.eu
Date: Fri, 15 Oct 2010 15:42:52 +0200
Subject: [PATCH] Changed symlink processing to new application name.

---
 testautomation/global/system/includes/iniinfo.inc |   16 
 1 files changed, 8 insertions(+), 8 deletions(-)

diff --git a/testautomation/global/system/includes/iniinfo.inc b/testautomation/global/system/includes/iniinfo.inc
index 0b83af4..a716a5d 100755
--- a/testautomation/global/system/includes/iniinfo.inc
+++ b/testautomation/global/system/includes/iniinfo.inc
@@ -772,8 +772,8 @@ function getSofficeNative() as string
 if (unx = gPlatGroup) then
 '/// on Linux/Unix systems we need to check/resolv 2 possible links which point to the directory with the 'soffice' executable ///'
 try
-