Re: [Libreoffice] [MHST][Easy Hack] Get current page number for printing
On 07/18/2011 11:07 AM, Andrew Douglas Pitonyak wrote: On 07/16/2011 05:02 AM, Ta Duc Tung wrote: Hi, 1. How to get current page number for printing? We've found a variable mnCurPage from printdlg.cxx but it seems not to be the right one. I think it is current page number for preview in print dialog. If I were writing a macro and I wanted to know, I would probably get this from the view cursor (but that is just a guess). Thank you for replying Andrew. We've managed to get the current page. Cause we want to pass the current page to print dialog so we went to unotxdoc.cxx and get the current page from SwDocShell :) Again, thank you very much. ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] feature/gnumake4
I've done my best I think but there was some error with set_soenv (first patch) Hmm, I looked at the first patch, and I definitely don't think we should re-introduce some Hamburg-specific stuff that we don't understand, want or need. Not even OOo at Apache will (once they have something they can start trying to build) need those _ISBOOTSTRAP, WORK_STAMP etc things. All that (for instance, maybe half (?) of build.pl) is double-plus-obsolete. There were some changes with precompiled headers. But that's not finished. Are they useful for something. I don't understand what that should do and have not asked yet. In theory, pre-compiled headers are quite useful for Windows compilations, they are said to speed up the build significantly. In practice, successfully being able to use pre-compiled headers requires awareness of it and carefulness and following correct (documented somewhere in the OOo wiki, hopefully) procedure while coding, and we certainly have not had any of that. I once tried to use pre-compiled headers when building on Windows, I had to fix lots of stuff here and there in one module, and in another it was just hopeless, so I gave up. I wonder if even Hamburg actually used pre-compiled headers in their Windows builds? I now think that we should just not bother even trying. A key to success is recognizing your limitations. Of course, the above is just my personal opinion, more clever people might disagree. --tml ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] [PATCH] Work around a bash-ism in configure.in
Would it not be simpler to change the -eq to = ? Anyway, good catch! --tml ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] [PATCH] Get rid of SvULongs in calc
Spotted some mistakes in calc patch. From e7db01344f8fef4ca2f2f59868736033d0126b1b Mon Sep 17 00:00:00 2001 From: Maciej Rumianowski maciej.rumianow...@gmail.com Date: Mon, 18 Jul 2011 09:22:57 +0200 Subject: [PATCH] Get rid of SvULongs in calc Instead of SvULongs use ::std::vector sal_uLong --- .../itemsetwrapper/DataPointItemConverter.cxx |7 +-- .../itemsetwrapper/SeriesOptionsItemConverter.cxx |7 +-- chart2/source/view/main/ChartItemPool.cxx |8 +++- sc/source/ui/docshell/impex.cxx| 11 +-- sc/source/ui/view/tabvwsh3.cxx |8 5 files changed, 14 insertions(+), 27 deletions(-) diff --git a/chart2/source/controller/itemsetwrapper/DataPointItemConverter.cxx b/chart2/source/controller/itemsetwrapper/DataPointItemConverter.cxx index 3802ffc..76e7db0 100644 --- a/chart2/source/controller/itemsetwrapper/DataPointItemConverter.cxx +++ b/chart2/source/controller/itemsetwrapper/DataPointItemConverter.cxx @@ -54,8 +54,6 @@ #include editeng/brshitem.hxx //SfxIntegerListItem #include svl/ilstitem.hxx -#define _SVSTDARR_ULONGS -#include svl/svstdarr.hxx #include vcl/graph.hxx #include com/sun/star/graphic/XGraphic.hpp @@ -641,10 +639,7 @@ void DataPointItemConverter::FillSpecialItem( case SCHATTR_DATADESCR_AVAILABLE_PLACEMENTS: { -SvULongs aList; -for ( sal_Int32 nN=0; nNm_aAvailableLabelPlacements.getLength(); nN++ ) -aList.Insert( m_aAvailableLabelPlacements[nN], sal::static_int_cast sal_uInt16 (nN) ); -rOutItemSet.Put( SfxIntegerListItem( nWhichId, aList ) ); +rOutItemSet.Put( SfxIntegerListItem( nWhichId, m_aAvailableLabelPlacements ) ); } break; diff --git a/chart2/source/controller/itemsetwrapper/SeriesOptionsItemConverter.cxx b/chart2/source/controller/itemsetwrapper/SeriesOptionsItemConverter.cxx index a199b66..7df2d61 100644 --- a/chart2/source/controller/itemsetwrapper/SeriesOptionsItemConverter.cxx +++ b/chart2/source/controller/itemsetwrapper/SeriesOptionsItemConverter.cxx @@ -51,8 +51,6 @@ //SfxIntegerListItem #include svl/ilstitem.hxx -#define _SVSTDARR_ULONGS -#include svl/svstdarr.hxx #include rtl/math.hxx #include functional @@ -433,10 +431,7 @@ void SeriesOptionsItemConverter::FillSpecialItem( } case SCHATTR_AVAILABLE_MISSING_VALUE_TREATMENTS: { -SvULongs aList; -for ( sal_Int32 nN=0; nNm_aSupportedMissingValueTreatments.getLength(); nN++ ) -aList.Insert( m_aSupportedMissingValueTreatments[nN], sal::static_int_cast sal_uInt16 (nN) ); -rOutItemSet.Put( SfxIntegerListItem( nWhichId, aList ) ); +rOutItemSet.Put( SfxIntegerListItem( nWhichId, m_aSupportedMissingValueTreatments ) ); break; } case SCHATTR_INCLUDE_HIDDEN_CELLS: diff --git a/chart2/source/view/main/ChartItemPool.cxx b/chart2/source/view/main/ChartItemPool.cxx index 2cdac8f..82df33d 100644 --- a/chart2/source/view/main/ChartItemPool.cxx +++ b/chart2/source/view/main/ChartItemPool.cxx @@ -39,10 +39,9 @@ #include svl/stritem.hxx #include svl/rectitem.hxx #include svl/ilstitem.hxx -#define _SVSTDARR_ULONGS -#include svl/svstdarr.hxx #include editeng/editids.hrc #include svx/svxids.hrc +#include vector #include com/sun/star/chart2/LegendPosition.hpp @@ -63,8 +62,7 @@ ChartItemPool::ChartItemPool(): ppPoolDefaults[SCHATTR_DATADESCR_SHOW_SYMBOL- SCHATTR_START] = new SfxBoolItem(SCHATTR_DATADESCR_SHOW_SYMBOL); ppPoolDefaults[SCHATTR_DATADESCR_SEPARATOR - SCHATTR_START] = new SfxStringItem(SCHATTR_DATADESCR_SEPARATOR,C2U( )); ppPoolDefaults[SCHATTR_DATADESCR_PLACEMENT - SCHATTR_START] = new SfxInt32Item(SCHATTR_DATADESCR_PLACEMENT,0); -SvULongs aTmp; -ppPoolDefaults[SCHATTR_DATADESCR_AVAILABLE_PLACEMENTS - SCHATTR_START] = new SfxIntegerListItem(SCHATTR_DATADESCR_AVAILABLE_PLACEMENTS,aTmp); +ppPoolDefaults[SCHATTR_DATADESCR_AVAILABLE_PLACEMENTS - SCHATTR_START] = new SfxIntegerListItem(SCHATTR_DATADESCR_AVAILABLE_PLACEMENTS, ::std::vector sal_uLong () ); ppPoolDefaults[SCHATTR_DATADESCR_NO_PERCENTVALUE- SCHATTR_START] = new SfxBoolItem(SCHATTR_DATADESCR_NO_PERCENTVALUE); ppPoolDefaults[SCHATTR_PERCENT_NUMBERFORMAT_VALUE - SCHATTR_START] = new SfxUInt32Item(SCHATTR_PERCENT_NUMBERFORMAT_VALUE, 0); ppPoolDefaults[SCHATTR_PERCENT_NUMBERFORMAT_SOURCE - SCHATTR_START] = new SfxBoolItem(SCHATTR_PERCENT_NUMBERFORMAT_SOURCE); @@ -157,7 +155,7 @@ ChartItemPool::ChartItemPool(): ppPoolDefaults[SCHATTR_CLOCKWISE- SCHATTR_START] = new SfxBoolItem( SCHATTR_CLOCKWISE, sal_False ); ppPoolDefaults[SCHATTR_MISSING_VALUE_TREATMENT- SCHATTR_START] = new SfxInt32Item(SCHATTR_MISSING_VALUE_TREATMENT, 0); -ppPoolDefaults[SCHATTR_AVAILABLE_MISSING_VALUE_TREATMENTS - SCHATTR_START] = new
Re: [Libreoffice] feature/gnumake4
Hi Matúš, Hi Tor, On Mon, 18 Jul 2011 00:32:49 -0600 Tor Lillqvist tlillqv...@novell.com wrote: Hmm, I looked at the first patch, and I definitely don't think we should re-introduce some Hamburg-specific stuff that we don't understand, want or need. Not even OOo at Apache will (once they have something they can start trying to build) need those _ISBOOTSTRAP, WORK_STAMP etc things. All that (for instance, maybe half (?) of build.pl) is double-plus-obsolete. Right, but just getting it to build instead of having two half complete conflicting removals of tcsh in the branch is a value of its own. ;) In theory, pre-compiled headers are quite useful for Windows compilations, they are said to speed up the build significantly. Yes, they are. A reason for this is also the badnesss of our codebase: little encapsulation = huge amounts of included headers. In practice, successfully being able to use pre-compiled headers requires awareness of it and carefulness and following correct (documented somewhere in the OOo wiki, hopefully) procedure while coding, and we certainly have not had any of that. Well, IIRC there are two major issues with precompiled headers: - tooltypes vs. windows native types, however this problem should be limited to vcl only mostly after the removetooltypes integration - other namespace conflicts introduced by using namespace statements, the most obnoxious being ::rtl::Reference and ::com::sun::star::uno::Reference conflicts. If somebody uses using namespace ::rtl; using namespace ::com::sun::star::uno; in a cxx because he 'knows' only one of the Reference declarations will get included, thats trouble already by itself. The risk of this blowing up in your face is of course greatly increased by precompiled headers. There is also a related EasyHack still: http://wiki.documentfoundation.org/Development/Easy_Hacks_Complete_List#Remove_.22using_namespace_::rtl.22 I once tried to use pre-compiled headers when building on Windows, I had to fix lots of stuff here and there in one module, and in another it was just hopeless, so I gave up. I wonder if even Hamburg actually used pre-compiled headers in their Windows builds? It was used on all builds and tinderboxes, which is the only way to make that happen. I now think that we should just not bother even trying. A key to success is recognizing your limitations. With the majority of our devs working on *nix, I agree. Instead of spending time get this fragile one platform feature (on gcc it is totally not worth it) working, we should try to fix the root cause, by improving encapsulation (e.g. get to the point where not ~400 .cxx in sw include doc.hxx, which itself includes half of the all the other headers in office). Of course, thats a nut that is a bit harder to crack -- but it will improve so much more than windows build times. Best, Bjoern -- https://launchpad.net/~bjoern-michaelsen ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] [PATCH] Possibly unused OTempFileService class
Hi Caolan, On Sun, 2011-07-17 at 11:22 +0100, Caolán McNamara wrote: Where an extra utl string is to be prefixed to the stock component_getFactory entry point string to give an entry point of Which makes me wonder: does callcatcher cope with eliding foo_component_ prefixed methods we have these days ? ATB, 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] [ANNOUNCE] libreoffice-3.4.2.1 tag created (3.4.2-rc1)
Andras Timar píše v Čt 14. 07. 2011 v 23:00 +0200: Hi Petr, 2011.07.12. 16:02 keltezéssel, Petr Mladek írta: there have been created the libreoffice-3.4.2.1 tag for 3.4.2-rc1 release. The corresponding official builds will be available by the end of the week. See the attached list of changes against 3.4.1-final. I saw changes in the bugfix list that were in 3.4.1 final, so they are now new fixes. Example: + set default min password length in ctor of password dialog (fdo#37771) [Andras Timar] http://cgit.freedesktop.org/libreoffice/libs-core/commit/?h=libreoffice-3-4-1id=0296fabc3ceafb000eb983f1889afb368012056f Maybe you counted all changes after the creation of libreoffice-3.4.1.1 tag. I used the command: bin/lo-commit-stat --log-suffix=release-3.4.2.1 . \ libreoffice-3.4.1.3..libreoffice-3.4.2.1 It got the list of changes using: git log libreoffice-3.4.1.3..libreoffice-3.4.2.1 Hmm, it somehow does not detect fixes that were cherry picked between libreoffice-3-4 and libreoffice-3-4-1 branches. Hmm, the option --cherry-pick does not help. They were either not cherry picked or this option does not work. Does anyone have any idea how to get a better list of changes between the two tags? How to filter the same commits? Best Regards, Petr PS: The above sample commit is from the list-core repo. ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] error in basic.src ?
Hi Matus, On Sat, 2011-07-16 at 23:18 +0200, Matúš Kukan wrote: It seems to be unused also in VCLTestTool. At least I could not grep aVersionString from Ah - but aVersionString is a widget - so this gets displayed in the dialog as a FixedText - without anyone else needing to do anything with that member [ vcl's semantics are a tad opaque in this regard ]. So maybe something like attached patch could be pushed. Someone willing to review and push? So - we did remove the field in this case which is not ideal; then again - this UPDVER doesn't look like anything that useful: 350 eg. I suggest we use: utl::Bootstrap::getBuildIdData instead - which gives a very detailed version; or failing that re-use code from cui/source/dialogs/about.cxx which renders help-about. Feel free to push something like that ;-) we should have a clean run-time method to obtain our version surely somewhere too. HTH, Michael. -- michael.me...@novell.com , Pseudo Engineer, itinerant idiot ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
[Libreoffice] REVIEW: allow to call make distro-pack-install with normal user
Hi, the attached patch fixes two problems found when calling make distro-pack-install as normal user. It would be great to get it into libreoffice-3-4-2 branch. Best Regards, Petr From 64653bcd8c1f1deeb8483e0d3e4ee8768a807b5d Mon Sep 17 00:00:00 2001 From: Petr Mladek pmla...@suse.cz Date: Mon, 18 Jul 2011 12:32:57 +0200 Subject: [PATCH] distro-install-desktop-integration: work with normal user + alllow to overwrite the readonly desktop files + create the compat soffice symlink in DESTDIR --- bin/distro-install-desktop-integration |4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/bin/distro-install-desktop-integration b/bin/distro-install-desktop-integration index c00c8c6..600c54e 100755 --- a/bin/distro-install-desktop-integration +++ b/bin/distro-install-desktop-integration @@ -123,7 +123,7 @@ for file in *.desktop ; do -e s/\($INSTALLDIRNAME\)$PRODUCTVERSION/\1/ \ -e s/\($PRODUCTNAME\) $PRODUCTVERSION/\1/ \ $file $file.new -mv $file.new $file +mv -f $file.new $file done cd - @@ -160,7 +160,7 @@ add_wrapper unopkg unopkg unopkg # you cannot include this link directly into the package # For example, the Novell package mark this symlink as %ghost # and update it in %post and %postun -ln -sf $INSTALLDIR/program/soffice $PREFIXDIR/bin/soffice +ln -sf $INSTALLDIR/program/soffice $DESTDIR$PREFIXDIR/bin/soffice # create bash completion mkdir -p $DESTDIR/etc/bash_completion.d -- 1.7.3.4 ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] [PATCH][REVIEW] SetNamedRanges speedup fdo#35357
Hi Laurent, On Fri, 2011-07-15 at 19:41 +0200, Laurent Godard wrote: Here is my first patch and want at first thanks a lot everybody who helped me starting my build and especially Kohei for his patience and deep knowledge in calc. Kohei is a great mentor :-) First - welcome ! :-) it's great to see your code. The result with the test case in the fdo#35357 is a drop for the 1211 namedRanges from 5.781 seconds to 0.931 seconds. Sounds like a really useful speedup :-) calc is packed with fertile ground for optimisations I think ... it would also be great for me if it could be part of the 3.4.x branch as this would affect a large number of my users instead of being stuck to OOo 3.2.1 I don't see a problem with that; sounds like a nasty bug - but it'd be targetted at 3.4.3 I suspect; wrt. the patch review Kohei or Markus would be able to do that quickest, so it'll be another few hours until they're around. Good stuff :-) 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] REVIEW: allow to call make distro-pack-install with normal user
Hi Petr, On Mon, 2011-07-18 at 12:39 +0200, Petr Mladek wrote: the attached patch fixes two problems found when calling make distro-pack-install as normal user. Both look fine / trivial to me, I've pushed to libreoffice-3-4, two more reviews needed for libreoffice-3-4-2. Thanks, 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] [PATCH] Possibly unused OTempFileService class
On Sun, Jul 17, 2011 at 11:22:45AM +0100, Caolán McNamara wrote: On Sun, 2011-07-17 at 08:13 +0200, Francois Tigeot wrote: It seems like OTempFileService, defined and implemented in libs-gui/unotools/source/ucbhelper is unused. There's no mention of this name elsewhere in the tree, and LibreOffice builds and run fine without the files removed by the attached patch. One of the methods of this class is already in the list of unused functions. I don't see any OTempFileService methods in the unused list ? I must have found a similarly named function and not paid attention to the base class. Howewer, since it involves UNO, this code may be loaded at runtime by an obscure corner of the gui I'm not aware of. unotools/util/utl.component registers that a component called com.sun.star.io.comp.TempFile which provides the com.sun.star.io.TempFile service is available in the unotools lib Where an extra utl string is to be prefixed to the stock component_getFactory entry point string to give an entry point of utl_component_getFactory as the symbol which can be dlsymed and executed with the component name (com.sun.star.io.comp.TempFile) in order to get one. utl_component_getFactory then calls OTempFileService::createServiceFactory_Static - XTempFile_createInstance - XTempFile_createInstance and returns a OTempFileService. So, its used by anything which asks for com.sun.star.io.TempFile e.g. http://opengrok.libreoffice.org/xref/components/package/source/zippackage/ZipPackageStream.cxx#188 where that ..-createInstance(com.sun.star.io.TempFile) call should get mapped to an appropiate call to utl_component_getFactory to basically get a OTempFileService back. OK, this all makes sense. UNO is still a very obscure thing for me and I thought the TempFile references only applied to another class. -- Francois Tigeot ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
[Libreoffice] [REVIEWED one more needed for 3-4-2] allow to call make distro-pack-install with normal user
On Mon, 18 Jul 2011 11:51:29 +0100 Michael Meeks michael.me...@novell.com wrote: Both look fine / trivial to me, I've pushed to libreoffice-3-4, two more reviews needed for libreoffice-3-4-2. Looks fine to me too. One more review needed for 3-4-2. Best, Bjoern -- https://launchpad.net/~bjoern-michaelsen ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] [PATCH] Work around a bash-ism in configure.in
Hi Mike, On Fri, 2011-07-15 at 12:34 -0700, Mike Eberdt wrote: This patch eliminates a test: : bad number diagnostic on FreeBSD. Contributed under LGPL3+/MPL. Thanks; I pushed it, and then updated it to use: if test z$EUID = z0 -a z`uname -o 2/dev/null` = zCygwin; then which Tor suggested IIRC - in general we don't comment previous problems with the code in the new, cleaned-up code - that would belong in the commit message :-) Thanks for your fix ! how are we looking on FreeBSD ? :-) 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] [PATCH] Second patch for Bug 38936
On Wed, 2011-07-13 at 11:17 +0800, jeffrey chang wrote: Hi, Second patch for bug. Now the undo statement for Insert Horizontal Rule replaces Insert horizontal ruler. Hi Jeffrey, Sorry it took so long to attend to your patch. I've applied and pushed to master, but I did change your commit name to properly capitalize your first name. It would be nice to have you correct this in your git config setting. Also, like your previous patch, I've assumed your patch is under LGPLv3 +/MPL 1.1. Please confirm that this is the case. Thanks a lot. Kohei -- Kohei Yoshida, LibreOffice hacker, Calc kyosh...@novell.com ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] [REVIEW 3-4] [PUSHED 3-4] Update mdds to 0.5.3
On Thu, 2011-07-14 at 10:31 +0200, David Tardon wrote: Nitpick, unrelated to the review: you may want to disable operator= for storage_filled_linear (and storage_filled_nested_array too), because its use would cause double delete. Good point. Done (on master for the next release). Thanks for the nit-pick. Kohei -- Kohei Yoshida, LibreOffice hacker, Calc kyosh...@novell.com ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] [PATCH] Fix 33898 - import/export xrefs to numbered paras
Hello Troy, Digging mailing-list archive and just glancing at your patch, isn't this patch against fdo#33960 ? (https://bugs.freedesktop.org/show_bug.cgi?id=33960) IMHO it isn't likely to be 33898. Best Regards, -- Korrawit Pruegsanusak ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] Major mode of Emacs for editing LibreOffice BASIC
Hi Michael, Thank you for your suggestion! On Mon, 18 Jul 2011 11:22:38 +0100, Michael Meeks michael.me...@novell.com wrote: On Mon, 2011-07-18 at 02:04 +0900, Takeshi Abe wrote: I would like to introduce my pet project libo-basic-mode: https://github.com/tabe/libo-basic-mode so that it can obtain your kind review. Nice :-) I wonder how best to distribute these developer tools - the python gdb bindings are another rather key feature that I still don't have installed (somehow) ;-) It'd perhaps be good to create a new devtools output in the scp2 that packages this stuff up (?) to encourage distributors to package it - any chance of something like that ? OK, I just learned what scp2 is for, from: http://wiki.services.openoffice.org/wiki/Extensions_Integration_into_Installation_Set If we can create a dedicated directory for installation of elisp, share/emacs/site-lisp/ sounds conventional and so nice to me. Cheers, -- Takeshi Abe ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] [PATCH][REVIEW] SetNamedRanges speedup fdo#35357
Hi Laurent, On Fri, 2011-07-15 at 19:41 +0200, Laurent Godard wrote: thanks in advance for your review and thank a lot again to Kohei Oh, thanks a lot for working on this performance optimization! :-) As I've already been involved in reviewing your changes giving you code pointers etc, your patch already has my blessing. I've just checked this in and pushed to master (as I promised in IRC). I'd also like to push this to the -3-4 branch, but let me just do a quick test on ODS import of named ranges to make sure if works, then I'll push. Kohei -- Kohei Yoshida, LibreOffice hacker, Calc kyosh...@novell.com ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] Major mode of Emacs for editing LibreOffice BASIC
Hi Takeshi, On Mon, 2011-07-18 at 23:22 +0900, Takeshi Abe wrote: Thank you for your suggestion! Hey - no problem; I think once we have a place for misc. system tools, it'll be much easier to get them widely packaged, and add to them. On Mon, 18 Jul 2011 11:22:38 +0100, Michael Meeks michael.me...@novell.com wrote: OK, I just learned what scp2 is for, from: http://wiki.services.openoffice.org/wiki/Extensions_Integration_into_Installation_Set If we can create a dedicated directory for installation of elisp, share/emacs/site-lisp/ sounds conventional and so nice to me. Right :-) sounds good; the trick is to play with it I guess; the problem is that our normal installs go into a prefix; so what we're really targetting is a unix / system package where we can whack things in /usr/share - Petr would be the right guy to help you work out how to get all those useful dev tools into a real package at the other end of the pipe :-) 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] [REVIEW 3-4] [PUSHED 3-4] Update mdds to 0.5.3
Hi there, On Fri, 2011-07-15 at 00:55 +0200, Markus Mohrhard wrote: So one review needed to get it into 3.4.2. It is an quite important fix as we got several reports where people reported they thought that calc froze and switched back to 3.3.2. So: git diff --numstat 0_5_2..0_5_3 # inside mdds shows: 9 1 include/mdds/mixed_type_matrix.hpp 22 3 include/mdds/mixed_type_matrix_def.inl 113 652 include/mdds/mixed_type_matrix_storage.hpp 701 0 include/mdds/mixed_type_matrix_storage_filled_linear.inl 374 0 include/mdds/mixed_type_matrix_storage_filled_nested_array.inl 376 0 include/mdds/mixed_type_matrix_storage_sparse.inl 2 244 include/mdds/rectangle_set.hpp 274 0 include/mdds/rectangle_set_def.inl 2 2 misc/mdds.spec.in 11 0 multidimalgorithm.vpj 81 0 src/array_creation_perf_test.cpp 75 8 src/mixed_type_matrix_test.cpp 1 0 src/test_global.hpp I feel extremely uncomfortable (personally) introducing a couple of thousands of LOC change between RC1 and final for 3.4.2. We're in a triple review mode because our changes will not get much testing before they are unleashed on tens of thousands of new people per day. I'd prefer to see this slip into 3.4.3 (personally) - shame if the problem is there, but the chance of regression seems high - even though the code is nice and clean of course :-) HTH, Michael. -- michael.me...@novell.com , Pseudo Engineer, itinerant idiot ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
[Libreoffice] [REVIEW] more distro-pack-install fixes
Hi, the below patches should fix two more problems reported by Andy. I would like to get them into libreoffice-3-4-2 branch. Best Regards, Petr From 7b202653f352a6c1bd18066bc71c93b1a5b33c13 Mon Sep 17 00:00:00 2001 From: Petr Mladek pmla...@suse.cz Date: Mon, 18 Jul 2011 16:48:49 +0200 Subject: [PATCH 2/3] distro-install-desktop-integration: do not pack libreoffice-base.desktop twice --- bin/distro-install-desktop-integration |2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/bin/distro-install-desktop-integration b/bin/distro-install-desktop-integration index 600c54e..9e3b499 100755 --- a/bin/distro-install-desktop-integration +++ b/bin/distro-install-desktop-integration @@ -148,7 +148,7 @@ add_wrapper lomath soffice --math libreoffice libreoffice-mat add_wrapper loimpress soffice --impresslibreoffice libreoffice-impress.desktop gid_Module_Brand_Prg_Impress add_wrapper loweb soffice --weblibreoffice gid_Module_Brand_Prg_Wrt add_wrapper lowriter soffice --writer libreoffice libreoffice-writer.desktop gid_Module_Brand_Prg_Wrt -add_wrapper lofromtemplate soffice .uno:NewDoc libreoffice libreoffice-base.desktopgid_Module_Root_Brand +add_wrapper lofromtemplate soffice .uno:NewDoc libreoffice gid_Module_Root_Brand add_wrapper libreofficesoffice libreoffice libreoffice-startcenter.desktop gid_Module_Root_Brand add_wrapper lofficesoffice libreoffice gid_Module_Root_Brand add_wrapper unopkg unopkg unopkg gid_Module_Root_Brand -- 1.7.3.4 From 72d5c4c3f7d2a5d329eb74fa1d1f69c1fb8c15b9 Mon Sep 17 00:00:00 2001 From: Petr Mladek pmla...@suse.cz Date: Mon, 18 Jul 2011 16:50:18 +0200 Subject: [PATCH 3/3] distro-install: fix en-US-only installation generate en-US file list when --with-distro= sigh, we should set WITH_LANG=en-US when no language is selected; the empty string is pretty ugly; unforrunately, many makefile tests check for this empty variable --- configure.in |3 +++ 1 files changed, 3 insertions(+), 0 deletions(-) diff --git a/configure.in b/configure.in index 18f93d2..65813cb 100755 --- a/configure.in +++ b/configure.in @@ -7906,6 +7906,9 @@ for lang in $WITH_LANG ; do done # list with substituted ALL WITH_LANG_LIST=`echo $WITH_LANG | sed s/ALL/$ALL_LANGS/` +# this variable is used only by bin/distro-install-* helper scripts +# they need a real list of languages +test -z $WITH_LANG_LIST WITH_LANG_LIST=en-US AC_SUBST(ALL_LANGS) AC_SUBST(WITH_LANG) AC_SUBST(WITH_LANG_LIST) -- 1.7.3.4 ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
[Libreoffice] [REVIEW] starmath: check for existing edit window
Hi guys, On Sat, 2011-07-16 at 01:08 +0200, Jonas Finnemann Jensen wrote: Please let me know if it looks problematic Looks great :) I'd like to get Miklos' crasher fix pair (just a couple of null pointer checks) back-ported from libreoffice-3-4 into libreoffice-3-4-2, so far Jonas, Miklos myself have approved them. I'd like one more approval (they are somewhat trivial) and someone to pick to libreoffice-3-4-2; patches appended. Thanks, Michael. commit aac4ab2af1ea151839e8a41ee9a9ec4d74558367 Author: Miklos Vajna vmik...@frugalware.org Date: Fri Jul 15 18:11:43 2011 +0200 starmath: check for edit window when inline editing is enabled Signed-off-by: Michael Meeks michael.me...@novell.com diff --git a/starmath/source/view.cxx b/starmath/source/view.cxx index 662911e..c924e31 100644 --- a/starmath/source/view.cxx +++ b/starmath/source/view.cxx @@ -224,7 +224,8 @@ void SmGraphicWindow::GetFocus() { if (!IsInlineEditEnabled()) return; -pViewShell-GetEditWindow()-Flush(); +if (pViewShell-GetEditWindow()) +pViewShell-GetEditWindow()-Flush(); //Let view shell know what insertions should be done in visual editor pViewShell-SetInsertIntoEditWindow(false); SetIsCursorVisible(true); commit b8c8e7627e0ffd75744049228207be344ddb7118 Author: Miklos Vajna vmik...@frugalware.org Date: Thu Jul 14 16:44:32 2011 +0200 starmath: check for existing edit window Signed-off-by: Michael Meeks michael.me...@novell.com diff --git a/starmath/source/view.cxx b/starmath/source/view.cxx index 3a099c3..662911e 100644 --- a/starmath/source/view.cxx +++ b/starmath/source/view.cxx @@ -195,6 +195,8 @@ void SmGraphicWindow::MouseButtonDown(const MouseEvent rMEvt) if (pNode) { SmEditWindow *pEdit = pViewShell-GetEditWindow(); +if (!pEdit) +return; const SmToken aToken (pNode-GetToken()); // set selection to the beginning of the token -- michael.me...@novell.com , Pseudo Engineer, itinerant idiot ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] small problems with --enable-split-app-modules
Andreas Radke píše v Ne 17. 07. 2011 v 21:26 +0200: 1) I have to pass yes to the install process due to some overwritten files: yes | make DESTDIR=${srcdir}/fakeinstall distro-pack-install Installer finished [...] Copying GNOME icons .. chmod: cannot operate on dangling symlink `/build/src/fakeinstall/usr/bin/libreoffice' /build/src/build mv: try to overwrite `base.desktop', overriding mode 0444 (r--r--r--)? j mv: try to overwrite `calc.desktop', overriding mode 0444 (r--r--r--)? j mv: try to overwrite `draw.desktop', overriding mode 0444 (r--r--r--)? j This should get fixed by the attached 0001-distro-install-desktop-integration-work-with-normal-.patch. Heh, this problem was not visible when I tested it as root. 2) the /usr/share/applications/base.desktop file goes not only into the base filelist but also into the common pkg file list. I had to remove it from the common pkg file list. Great catch! This gets fixed by the attached 0002-distro-install-desktop-integration-do-not-pack-libre.patch 3) I'm running into a build issue when using --with-lang=: Error: ../unxlngx6.pro/misc/sysui/dummy/localize.sdf 0 Bytes! Error: 1 damaged files encountered I am unable toi reproduce this but I do not have a clean build tree at hands. Anyway, --with-lang= is needed for the en-US build on many other locations, so it is the right way to go. I'm removing the --with-lang or add --with-lang=en-US now to get en-US langpack included. But there's no full langpack/helpfile file list created. Only the splitted files in orig/gid_Modules_... are there. It should work better with the attached 0003-distro-install-fix-en-US-only-installation.patch This probably does not fix the build problem in sysui. Though, it should generate lang_en_US_list.txt file list even when using --with-lang= Without any langpack installed I get this error that I can solve with installing any other langpack from the rpms, but there's now en-US langpack: [andyrtr@workstation64 trunk]$ libreoffice Missing vcl resource. This indicates that files vital to localization are missing. You might have a corrupt installation. terminate called after throwing an instance of 'com::sun::star::uno::RuntimeException' Yup, it means that you do not have installed any lang pack files. Beside this this splitting looks working well, also all desktop integration seems to work. Great to hear this. I expect that more problems might be found over time. It was added too late, so some patches might be needed on top of the final libreoffice-3.4.2.x source tarballs. Anyway, it is good to have a common code base again. Best Regards, Petr From 64653bcd8c1f1deeb8483e0d3e4ee8768a807b5d Mon Sep 17 00:00:00 2001 From: Petr Mladek pmla...@suse.cz Date: Mon, 18 Jul 2011 12:32:57 +0200 Subject: [PATCH 1/3] distro-install-desktop-integration: work with normal user + alllow to overwrite the readonly desktop files + create the compat soffice symlink in DESTDIR --- bin/distro-install-desktop-integration |4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/bin/distro-install-desktop-integration b/bin/distro-install-desktop-integration index c00c8c6..600c54e 100755 --- a/bin/distro-install-desktop-integration +++ b/bin/distro-install-desktop-integration @@ -123,7 +123,7 @@ for file in *.desktop ; do -e s/\($INSTALLDIRNAME\)$PRODUCTVERSION/\1/ \ -e s/\($PRODUCTNAME\) $PRODUCTVERSION/\1/ \ $file $file.new -mv $file.new $file +mv -f $file.new $file done cd - @@ -160,7 +160,7 @@ add_wrapper unopkg unopkg unopkg # you cannot include this link directly into the package # For example, the Novell package mark this symlink as %ghost # and update it in %post and %postun -ln -sf $INSTALLDIR/program/soffice $PREFIXDIR/bin/soffice +ln -sf $INSTALLDIR/program/soffice $DESTDIR$PREFIXDIR/bin/soffice # create bash completion mkdir -p $DESTDIR/etc/bash_completion.d -- 1.7.3.4 From 7b202653f352a6c1bd18066bc71c93b1a5b33c13 Mon Sep 17 00:00:00 2001 From: Petr Mladek pmla...@suse.cz Date: Mon, 18 Jul 2011 16:48:49 +0200 Subject: [PATCH 2/3] distro-install-desktop-integration: do not pack libreoffice-base.desktop twice --- bin/distro-install-desktop-integration |2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/bin/distro-install-desktop-integration b/bin/distro-install-desktop-integration index 600c54e..9e3b499 100755 --- a/bin/distro-install-desktop-integration +++ b/bin/distro-install-desktop-integration @@ -148,7 +148,7 @@ add_wrapper lomath soffice --math libreoffice libreoffice-mat add_wrapper loimpress soffice --impresslibreoffice libreoffice-impress.desktop gid_Module_Brand_Prg_Impress add_wrapper loweb soffice --weblibreoffice gid_Module_Brand_Prg_Wrt add_wrapper lowriter soffice --writer libreoffice libreoffice-writer.desktop
Re: [Libreoffice] [litmus] - buildid: tag no longer exists in 3.4 help - About LibreOffice
Yifan Jiang píše v Po 18. 07. 2011 v 10:46 +0800: Hi Petr, Is there a constant rule for checking the exact version from UI directly? For example, Help-About LibreOffice 3.4 340m1(Build:103) At the moment, it looks the Build:103 means 3.4.1 RC 3. Does it reliable enough? It is reliable and enough for the announced builds (Betas, RCs). Though, the build id is not updated for daily builds. Caolan has already made some improvements in master for LO-3.5. See also https://bugs.freedesktop.org/show_bug.cgi?id=39160 AFAIK, we want to distinguish the builds by the last commit id once we switch to single git repo. Best Regards, Petr ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
[Libreoffice] [REVIEWED 3.4.2] The number of paragraphs always zerro
On Wed, 2011-07-13 at 21:32 +0100, Noel Power wrote: +1 and applied to 3.4 ( and master ) branch(s) And a last time - and pushed for 3.4.2. Thanks, 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] [REVIEW] Fix for fdo#37767
Hi Kohei, On Fri, 2011-07-15 at 09:48 +0100, Noel Power wrote: looks fine to me +1, one more needed for 3.4.2 Ditto here; much preferred to creating that new temporary DPObject :-) I've merged to libreoffice-3-4-2. I closed the bug too ;-) Thanks, 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] [PATCH] Possibly unused OTempFileService class
On Mon, 2011-07-18 at 11:25 +0100, Michael Meeks wrote: Hi Caolan, On Sun, 2011-07-17 at 11:22 +0100, Caolán McNamara wrote: Where an extra utl string is to be prefixed to the stock component_getFactory entry point string to give an entry point of Which makes me wonder: does callcatcher cope with eliding foo_component_ prefixed methods we have these days ? callcatcher itself can't really know about dlsym-ed symbols. But for the make findunusedcode I take an ultra-conservative approach and grep ::.* \( on the real list, which hides both the extern C methods which are plausibly dlsymabled, and also the C++ non-class functions which aren't, for the moment. I had/have a whitelist in callcatcher itself, but that's gone a bit stale, if/when we hit some low count in the easy list I might get back to extracting/updating the whitelist. C. ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] [PATCH][REVIEW] SetNamedRanges speedup fdo#35357
On Mon, 2011-07-18 at 10:25 -0400, Kohei Yoshida wrote: I'd also like to push this to the -3-4 branch, but let me just do a quick test on ODS import of named ranges to make sure if works, then I'll push. And I'm glad I did. This change unfortunately causes a problem on global named range import. It incorrectly imports them as external references. I'm looking into this at the moment. Kohei -- Kohei Yoshida, LibreOffice hacker, Calc kyosh...@novell.com ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] [REVIEW][PUSHED:master; 3-4; 3-4-2] starmath: check for existing edit window
Michael Meeks píše v Po 18. 07. 2011 v 15:59 +0100: Hi guys, On Sat, 2011-07-16 at 01:08 +0200, Jonas Finnemann Jensen wrote: Please let me know if it looks problematic Looks great :) I'd like to get Miklos' crasher fix pair (just a couple of null pointer checks) back-ported from libreoffice-3-4 into libreoffice-3-4-2, so far Jonas, Miklos myself have approved them. I'd like one more approval (they are somewhat trivial) and someone to pick to libreoffice-3-4-2; patches appended. Looked fine = pushed both changes into 3-4-2 branch. Best Regards, Petr ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
[Libreoffice] REMINDER: Release 3.4.2-rc2 from libreoffice-3-4-2 branch
Hi, please note that the commit deadline for 3.4.2-rc2 is today, Jul 18, 2011. It will be used as LO-3.4.2 final if no blocker is reported. See also http://wiki.documentfoundation.org/ReleasePlan#3.4_release http://wiki.documentfoundation.org/Release_Criteria http://wiki.documentfoundation.org/Development/Branches Best Regards, Petr ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] REVIEW: allow to call make distro-pack-install with normal user
Am Mon, 18 Jul 2011 12:39:26 +0200 schrieb Petr Mladek pmla...@suse.cz: Hi, the attached patch fixes two problems found when calling make distro-pack-install as normal user. It would be great to get it into libreoffice-3-4-2 branch. Best Regards, Petr Booth issues seem fixed for me. Thanks. -Andy ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
[Libreoffice] A thought about fdo#38095: Character classification for Western or Asian text font
Hi all, According to a discussion on the Japanese local mailing list, the following issue may prevent some users from migrating to 3.4.x series: https://bugs.freedesktop.org/show_bug.cgi?id=38095 I am not yet sure how it works in 3.3.x, but at least for Calc on master, there seems a gap around determining a font from the Unicode Script Property of content of a cell. In ScDocument::GetStringScriptType() at sc/source/core/data/documen6.cxx, a breakiterator of type i18n::ScriptType::WEAK is just ignored, while BreakIteratorImpl::getScriptClass() at i18npool/source/breakiterator/breakiteratorImpl.cxx maps entries of Common value of the script property to WEAK. More worse, DIGIT ZERO..DIGIT NINE and FULLWIDTH DIGIT ZERO..FULLWIDTH DIGIT NINE are both in Common [2], and, both of LATIN SMALL LETTER A..LATIN SMALL LETTER Z and FULLWIDTH LATIN SMALL LETTER A..FULLWIDTH LATIN SMALL LETTER Z are in Latin, so it can not distinguish fullwidth ones from ascii ones. I found [1] mentions a similar situaion, like: Note that while it is necessary to include Latin in the preceding expression to ensure that it can cover the typical script use found in many Japanese texts, doing so would make it difficult to isolate a run of Japanese inside an English document, for example. Of course next I should check the 3.3.x way, anyway it would be great if someone give me a hint for how to deal with the case. [1] http://unicode.org/reports/tr24/ [2] http://unicode.org/Public/UNIDATA/Scripts.txt Cheers, -- Takeshi Abe ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] [REVIEW] more distro-pack-install fixes
Am Mon, 18 Jul 2011 16:58:03 +0200 schrieb Petr Mladek pmla...@suse.cz: Hi, the below patches should fix two more problems reported by Andy. I would like to get them into libreoffice-3-4-2 branch. Best Regards, Petr The first patch works as expected. The second patch doesn't fix the file list creation for me. I'm using --with-lang= in configure that says: checking which languages to be built... en-US that's all I get here: [andyrtr@workstation64 trunk]$ ls -1 ~/arch64/chroots/testing-x86_64/andyrtr/build/src/build/file-lists/ base_list.txt calc_list.txt common_list.txt draw_list.txt gnome_list.txt impress_list.txt kde4_list.txt kde_list.txt math_list.txt mono_list.txt orig sdk_doc_list.txt sdk_list.txt writer_list.txt in orig folder there are : gid_Module_Helppack_Help_en_US gid_Module_Langpack_Base_en_US gid_Module_Langpack_Basis_en_US gid_Module_Langpack_Brand_en_US gid_Module_Langpack_Calc_en_US gid_Module_Langpack_Math_en_US gid_Module_Langpack_Resource_en_US gid_Module_Langpack_Writer_en_US gid_Module_Oo_Linguistic Do you need some further infos? ** Successful packaging process! *** Installer finished Installation finished, you can now execute: /usr/lib/libreoffice/program/soffice ./bin/distro-install-clean-up Cleaning up ... Fixing permissions... Checking for DESTDIR inside installed files... ./bin/distro-install-desktop-integration Copying icons .. Copying GNOME icons .. chmod: cannot operate on dangling symlink `/build/src/fakeinstall/usr/bin/libreoffice' /build/src/build /build/src/build Install /usr/bin/lobase Install /usr/share/man/man1/lobase.1.gz Install /usr/bin/localc Install /usr/share/man/man1/localc.1.gz Install /usr/bin/lodraw Install /usr/share/man/man1/lodraw.1.gz Install /usr/bin/lomath Install /usr/share/man/man1/lomath.1.gz Install /usr/bin/loimpress Install /usr/share/man/man1/loimpress.1.gz Install /usr/bin/loweb Install /usr/share/man/man1/loweb.1.gz Install /usr/bin/lowriter Install /usr/share/man/man1/lowriter.1.gz Install /usr/bin/lofromtemplate Install /usr/share/man/man1/lofromtemplate.1.gz Install /usr/bin/libreoffice Install /usr/share/man/man1/libreoffice.1.gz Install /usr/bin/loffice Install /usr/share/man/man1/loffice.1.gz Install /usr/bin/unopkg Install /usr/share/man/man1/unopkg.1.gz Install /basis/program/java-set-classpath ./bin/distro-install-sdk SDK installation clean up ./bin/distro-install-file-lists Generating package file lists for ArchLinux... -Andy ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] [REVIEW][PUSHED:3-4;3-4-2 Fix for fdo#37767
Michael Meeks píše v Po 18. 07. 2011 v 16:36 +0100: Hi Kohei, On Fri, 2011-07-15 at 09:48 +0100, Noel Power wrote: looks fine to me +1, one more needed for 3.4.2 Ditto here; much preferred to creating that new temporary DPObject :-) I've merged to libreoffice-3-4-2. I closed the bug too ;-) Just updated the subject accordingly ;-) Best Regards, Petr ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] [PATCH][REVIEW] SetNamedRanges speedup fdo#35357
On Mon, 2011-07-18 at 11:47 -0400, Kohei Yoshida wrote: On Mon, 2011-07-18 at 10:25 -0400, Kohei Yoshida wrote: I'd also like to push this to the -3-4 branch, but let me just do a quick test on ODS import of named ranges to make sure if works, then I'll push. And I'm glad I did. This change unfortunately causes a problem on global named range import. It incorrectly imports them as external references. I'm looking into this at the moment. This commit fixes it http://cgit.freedesktop.org/libreoffice/calc/commit/?id=ce470bf9f95ae53c29356976b2c5e44cd6c73fed There was a bug in the RangeNameInserter function object which was exposed by your change. Apparently LibO / OOo sometimes write range names this way: $Sheet1.$A$1:.$A$3 (the leading dot in the end range address) which was supposed to be corrected to $Sheet1.$A$1:$A$3 (without the dot in the end range address) before passing it to a range name object. But the code incorrectly used the original, uncorrected version of the range string, which caused the problem. Could someone please give this change a quick review so we can cherry-pick this to the -3-4 branch along with Laurent's original patch? I can sign off on Laurent's original patch but I can't self-review this one. Thanks! Kohei -- Kohei Yoshida, LibreOffice hacker, Calc kyosh...@novell.com ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] [libreoffice-l10n] [ANN] LibreOffice 3.4.2 RC1 available
Nino Novak píše v Čt 14. 07. 2011 v 22:48 +0200: Am Donnerstag, 14. Juli 2011, um 14:29:53 schrieb Florian Effenberger: The list of fixed bugs in this release is here: http://dev-builds.libreoffice.org/pre-releases/src/bugfixes-libreoffi ce-3-4-release-3.4.2.1.log for a clickable version, look at http://wiki.documentfoundation.org/User:Nnino/Drafts/bugfixes3.4.2 Looks great. I wonder how you generated it. LeMoyne has been working on some scripting. Do you have any script as well? ;-) Best Regards, Petr ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
[Libreoffice] [GSOC] Report #8. Wizard
Hello, What have I done this week? - Work on the agenda wizard. I wanted it to be done about last weekend but there's a bug ( which affects the java wizard on master too ) and it took me a while to find it out. I reported the bug here: https://bugs.freedesktop.org/show_bug.cgi?id=39313. I don't think it affects the 3.4 branch but it would be nice if someone can try it and check that the wizard runs properly ( File Wizards Agenda ) because it crashes Libo on master. - Take a look to this bug: https://bugs.freedesktop.org/show_bug.cgi?id=38451 . I already know how to fix it but I'm waiting for cédric's answer ( I think he's on holidays) - As I said in my previous report, there is something ( i haven't been able to figure it out ) that breaks Adress Data Source. Here some information: http://www.mail-archive.com/libreoffice@lists.freedesktop.org/msg13235.html . Should i create a new bug ? Greetings ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] [REVIEW][PUSHED:3-4;3-4-2] Fix for fdo#36688
Noel Power píše v Pá 15. 07. 2011 v 09:44 +0100: On 15/07/11 08:27, Tor Lillqvist wrote: Review, sign-off etc appreciated. /me raises hand. One more needed for the 3-4-2 branch. pushed to 3.4.2 Heh. the above line was quite hidden ;-) Just updated the subject accordingly. Best Regards, Petr ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] REVIEW: allow to call make distro-pack-install with normal user
Am Mon, 18 Jul 2011 18:11:00 +0200 schrieb Andreas Radke a.ra...@arcor.de: Am Mon, 18 Jul 2011 12:39:26 +0200 schrieb Petr Mladek pmla...@suse.cz: Hi, the attached patch fixes two problems found when calling make distro-pack-install as normal user. It would be great to get it into libreoffice-3-4-2 branch. Best Regards, Petr Booth issues seem fixed for me. Thanks. -Andy Hm. Maybe too fast answered. How do you build your packages? I'm doing a fake install with: make DESTDIR=${srcdir}/fakeinstall distro-pack-install and the I move the files from ${srcdir}/fakeinstall into the splitted packaging directories using block per pkg this: # create directories from *list.txt file for directory in `grep ^%dir ${srcdir}/build/file-lists/common_list.txt`; do install -dm755 ${pkgdir}/${directory/\%dir/} done # install files into the pkg from fakeinstall dir for file in `grep -v ^%dir $srcdir/build/file-lists/common_list.txt`; do dirname=`dirname $file` # check if directory has been already been created - some are missing like manpages [ -d ${pkgdir}/$dirname ] || install -dm755 ${pkgdir}/$dirname # mv file from fakeinstall to pkgdir mv ${srcdir}/fakeinstall${file} ${pkgdir}$file done That doesn't move the /usr/bin/soffice symlink into the common pkg. No big deal. I can recreate it but maybe I would miss other links in future builds as well. Do you have a more elegant way that prevents this? -Andy /bin/distro-install-clean-up Cleaning up ... Fixing permissions... Checking for DESTDIR inside installed files... ./bin/distro-install-desktop-integration Copying icons .. Copying GNOME icons .. chmod: cannot operate on dangling symlink `/build/src/fakeinstall/usr/bin/libreoffice' ??? /build/src/build /build/src/build Install /usr/bin/lobase ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] [PATCH] Work around a bash-ism in configure.in
On 07/17/11 23:47, Tor Lillqvist wrote: Would it not be simpler to change the -eq to = ? Anyway, good catch! --tml I guess I didn't go that route because my brain tends to avoid the zFoo = zBar construct if possible. But yes, I should have written it in the form that Michael Meeks checked in. Sorry 'bout that. ME ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
[Libreoffice] PM / was: Re: [libreoffice-l10n] [ANN] LibreOffice 3.4.2 RC1 available
Hi Petr, Am Montag, 18. Juli 2011, um 18:33:06 schrieben Sie: Nino Novak píše v Čt 14. 07. 2011 v 22:48 +0200: Am Donnerstag, 14. Juli 2011, um 14:29:53 schrieb Florian Effenberger: The list of fixed bugs in this release is here: http://dev-builds.libreoffice.org/pre-releases/src/bugfixes-libre offi ce-3-4-release-3.4.2.1.log for a clickable version, look at http://wiki.documentfoundation.org/User:Nnino/Drafts/bugfixes3.4.2 Looks great. I wonder how you generated it. LeMoyne has been working on some scripting. Do you have any script as well? ;-) just piped it through the following sed script to obtain Wiki markup: s/fdo#[0-9]*/{{\0}}/g s/bnc#[0-9]*/{{\0}}/g s/i#[0-9]*/{{\0}}/g s/i#/issue#/g s/^\+ \(.*\)$/== \1 ==/ s/^. /* / Nino ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] [REVIEW] more distro-pack-install fixes
Petr Mladek píše v Po 18. 07. 2011 v 16:58 +0200: Hi, the below patches should fix two more problems reported by Andy. I would like to get them into libreoffice-3-4-2 branch. One more fix. We finally install /usr/bin/soffice into $DESTDIR. We need to put it into the file list as well. Best Regards, Petr From 2046eaeaaabbc58326066ee37f53cc97ebf679bb Mon Sep 17 00:00:00 2001 From: Petr Mladek pmla...@suse.cz Date: Mon, 18 Jul 2011 19:47:39 +0200 Subject: [PATCH 4/4] distro-install-desktop-integration: put $PREFIXDIR/bin/soffice into file list --- bin/distro-install-desktop-integration |1 + 1 files changed, 1 insertions(+), 0 deletions(-) diff --git a/bin/distro-install-desktop-integration b/bin/distro-install-desktop-integration index 9e3b499..77ccf3a 100755 --- a/bin/distro-install-desktop-integration +++ b/bin/distro-install-desktop-integration @@ -161,6 +161,7 @@ add_wrapper unopkg unopkg unopkg # For example, the Novell package mark this symlink as %ghost # and update it in %post and %postun ln -sf $INSTALLDIR/program/soffice $DESTDIR$PREFIXDIR/bin/soffice +test -f $DESTDIR/gid_Module_Root_Brand echo $PREFIXDIR/bin/soffice $DESTDIR/gid_Module_Root_Brand # create bash completion mkdir -p $DESTDIR/etc/bash_completion.d -- 1.7.3.4 ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] REVIEW: allow to call make distro-pack-install with normal user
Andreas Radke píše v Po 18. 07. 2011 v 18:40 +0200: Am Mon, 18 Jul 2011 18:11:00 +0200 schrieb Andreas Radke a.ra...@arcor.de: Am Mon, 18 Jul 2011 12:39:26 +0200 schrieb Petr Mladek pmla...@suse.cz: Hi, the attached patch fixes two problems found when calling make distro-pack-install as normal user. It would be great to get it into libreoffice-3-4-2 branch. Best Regards, Petr Booth issues seem fixed for me. Thanks. -Andy Hm. Maybe too fast answered. How do you build your packages? I'm doing a fake install with: make DESTDIR=${srcdir}/fakeinstall distro-pack-install I use make DESTDIR=$RPM_BUILD_ROOT distro-pack-install. The variable RPM_BUILD_ROOT points to /vat/tmp/packname-version-arch. The rpm tool creates the packages directly from there using the file lists. So, I do not need to move the files after they are installed. That doesn't move the /usr/bin/soffice symlink into the common pkg. No big deal. I can recreate it but maybe I would miss other links in future builds as well. Do you have a more elegant way that prevents this? Ah, the file was not mentioned in the file list. It should get fixed by the attached patch. Thanks for testing. Best Regards, Petr From 2046eaeaaabbc58326066ee37f53cc97ebf679bb Mon Sep 17 00:00:00 2001 From: Petr Mladek pmla...@suse.cz Date: Mon, 18 Jul 2011 19:47:39 +0200 Subject: [PATCH 4/4] distro-install-desktop-integration: put $PREFIXDIR/bin/soffice into file list --- bin/distro-install-desktop-integration |1 + 1 files changed, 1 insertions(+), 0 deletions(-) diff --git a/bin/distro-install-desktop-integration b/bin/distro-install-desktop-integration index 9e3b499..77ccf3a 100755 --- a/bin/distro-install-desktop-integration +++ b/bin/distro-install-desktop-integration @@ -161,6 +161,7 @@ add_wrapper unopkg unopkg unopkg # For example, the Novell package mark this symlink as %ghost # and update it in %post and %postun ln -sf $INSTALLDIR/program/soffice $DESTDIR$PREFIXDIR/bin/soffice +test -f $DESTDIR/gid_Module_Root_Brand echo $PREFIXDIR/bin/soffice $DESTDIR/gid_Module_Root_Brand # create bash completion mkdir -p $DESTDIR/etc/bash_completion.d -- 1.7.3.4 ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] permission 444/555 of installation files
Andreas Radke píše v Po 18. 07. 2011 v 18:26 +0200: For a long time now we see lots of files installed with permission 444 and 555 in ArchLinux. We are not sure if this is caused by a bug in our chroot environment using fakeroot or due to the LibO installer. Because of the missing write permission our build tools fail to strip the symbols. So we apply this: #fix http://bugs.archlinux.org/task/17656 find ${pkgdir} -perm 444 -exec ls -lh {} \; find ${pkgdir} -perm 444 -exec chmod 644 {} \; find ${pkgdir} -perm 555 -exec ls -lh {} \; find ${pkgdir} -perm 555 -exec chmod 755 {} \; I was always curious about the too strict permissions. I havn't seen it in other software. The attached patch should change it. I am not sure if it is acceptable for others, though. Best Regards, Petr PS: If you want to test it, please, remove scp2/unxlng* and rebuild/deliver it. Just for sure ;-) --- scp2/inc/macros.inc.old 2011-05-18 17:51:47.0 +0200 +++ scp2/inc/macros.inc 2011-07-18 19:58:27.0 +0200 @@ -92,10 +92,10 @@ #endif #define BIN_FILE_BODY \ -UnixRights = 555 +UnixRights = 755 #define TXT_FILE_BODY \ -UnixRights = 444 +UnixRights = 644 #define USER_FILE_BODY \ UnixRights = 644 ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] [PATCH] Get rid of SvULongs in calc
Hi Maciej, First of all, thanks for submitting your patches. I've done my review of both of your patches, and I'll leave my comments below. First, for your svl patch: 1) in SfxIntegerListItem::GetList(), you do +rList.insert( aIt, m_aList[n] ); but you can simply call push_back here to append new data to the end of the vector. Also, let's use pre-increment on iterators in the for loop there i.e. instead of doing aIt++, do ++aIt. Post-increment on iterators is generally more expensive than pre-increment, so it's best to stick with pre-increments. 2) Your new constructors take arrays of different integer type; one takes std::vectorsal_uLong while the other takes uno::Sequencesal_Int32. Let's just stick with one type and use sal_Int32 in both cases since that's the integer type that this class uses internally. And for your sc patch: 3) In ScTabViewShell::Execute() you do -SvULongs aIndexList( 4, 4 ); +::std::vector sal_uLong aIndexList( 4, 4 ); but this actually changes the behavior of the code. The arguments to SvULong's constructor controls how the internal array is laid out during initialization but it doesn't populate the container with data. Vector's ctor arguments OTOH populate the container with data. So, your new code initializes aIndexList with 4 elements of value 4, which is not what the code intends to do. You can simply do without passing any arguments to aIndexList there. 4) Last but not least, please state that you are submitting your patches under LGPLv3+/MPL 1.1. That's all I can think of. Let's get these points revised, and I'll push the changes to master. Best, Kohei -- Kohei Yoshida, LibreOffice hacker, Calc kyosh...@novell.com ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] PM / was: Re: [libreoffice-l10n] [ANN] LibreOffice 3.4.2 RC1 available
On Mon, 2011-07-18 at 19:39 +0200, Nino Novak wrote: just piped it through the following sed script to obtain Wiki markup: s/fdo#[0-9]*/{{\0}}/g s/bnc#[0-9]*/{{\0}}/g s/i#[0-9]*/{{\0}}/g s/i#/issue#/g s/^\+ \(.*\)$/== \1 ==/ s/^. /* / I've been using rhbz# for Fedora/RHEL related bugs FWIW, not sure that are any fresh fixes for any of those in 3.4.2 however C. ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] A thought about fdo#38095: Character classification for Western or Asian text font
On Tue, 2011-07-19 at 01:14 +0900, Takeshi Abe wrote: Hi all, According to a discussion on the Japanese local mailing list, the following issue may prevent some users from migrating to 3.4.x series: https://bugs.freedesktop.org/show_bug.cgi?id=38095 I am not yet sure how it works in 3.3.x, but at least for Calc on master, there seems a gap around determining a font from the Unicode Script Property of content of a cell. This will be an easy fix, I'll take it. I had seen the change in passing and stuck it as a to-do to examine it, but hadn't realized it wasn't just an obscure edge-case that these number-forms were no longer categorized the same as they had been historically. IIRC they used to be WEAK chars but are now being categorized as LATIN like the numbers/letters they resemble. i18npool/source/breakiterator/breakiteratorImpl.cxx will be the right place. C. ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] [REVIEW 3-4] [PUSHED 3-4] Update mdds to 0.5.3
Hello Michael, Kohei, I feel extremely uncomfortable (personally) introducing a couple of thousands of LOC change between RC1 and final for 3.4.2. We're in a triple review mode because our changes will not get much testing before they are unleashed on tens of thousands of new people per day. I'd prefer to see this slip into 3.4.3 (personally) - shame if the problem is there, but the chance of regression seems high - even though the code is nice and clean of course :-) I understand your concerns but in my opinion the current situation is even worse. I think we will call 3.4.2 ready for enterprise use but calc is limited to simple documents because complex documents need some minutes to load and most people won't expect that. Even worse I read several comments on the german user list reporting that they think calc freezes when you load complex documents and suggest to stick with 3.3. I think we can't expect from an enterprise ready version that you can't load complex documents. I know that we may introduce problems in some corner cases but it will be even worse if we can not work in normal use cases with complex documents. So I think the question is how important it is to be absolutely safe that we don't include any bug when we fix a serious problem. In my opinion in this case the fix for the bug is more important but others might have different opinions here. Regards, Markus ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] [REVIEW 3-4] [PUSHED 3-4] Update mdds to 0.5.3
On Mon, 2011-07-18 at 15:53 +0100, Michael Meeks wrote: Hi there, On Fri, 2011-07-15 at 00:55 +0200, Markus Mohrhard wrote: So one review needed to get it into 3.4.2. It is an quite important fix as we got several reports where people reported they thought that calc froze and switched back to 3.3.2. So: git diff --numstat 0_5_2..0_5_3 # inside mdds shows: 9 1 include/mdds/mixed_type_matrix.hpp 22 3 include/mdds/mixed_type_matrix_def.inl 113 652 include/mdds/mixed_type_matrix_storage.hpp 701 0 include/mdds/mixed_type_matrix_storage_filled_linear.inl 374 0 include/mdds/mixed_type_matrix_storage_filled_nested_array.inl 376 0 include/mdds/mixed_type_matrix_storage_sparse.inl 2 244 include/mdds/rectangle_set.hpp 274 0 include/mdds/rectangle_set_def.inl 2 2 misc/mdds.spec.in 11 0 multidimalgorithm.vpj 81 0 src/array_creation_perf_test.cpp 75 8 src/mixed_type_matrix_test.cpp 1 0 src/test_global.hpp I feel extremely uncomfortable (personally) introducing a couple of thousands of LOC change between RC1 and final for 3.4.2. We're in a triple review mode because our changes will not get much testing before they are unleashed on tens of thousands of new people per day. Just to tone down this statistics a bit. The vast majority of changes are in fact code relocations with no changes e.g. the change in rectangle_set is purely moving code from one file to another. Also, we should ignore the changes in the src directory since those files are only for internal unit tests, that are not even accessible from the consumer code. There are other files that are not relevant to the actual code change that affects the LibO code, such as rpm spec file, my editor's project file etc. With that, the real change that directly affects LibreOffice code is much much smaller than a couple of thousands of LOC that you say. I'm just mentioning this so that we can make a decision based on more meaning analysis than simply running stats on all changes when in fact not all changes are relevant to our code base. Peace, Kohei -- Kohei Yoshida, LibreOffice hacker, Calc kyosh...@novell.com ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
[Libreoffice] Make RPMs
Hello all, How can I build *rpm installation set under CentOS 5.6? I know that ./autogen.sh --with-distro=LibreOfficeLinux --with-lang=vi --with-max-jobs=20 --with-num-cpus=5 --with-package-format=rpm deb --without-junit --with-gcc-speedup=ccache --prefix=$HOME make make install make dev-install will create execution files but I don't know how to create rpms files. # It seems that--with-package-format=rpm deb doesn't help. -- Nguyen Vu Hung (Nguyye^~n Vu~ Hu+ng) LibreOffice, Vietnamese Translation Team. ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] PM / was: Re: [libreoffice-l10n] [ANN] LibreOffice 3.4.2 RC1 available
On Mon, 18 Jul 2011 19:39:31 +0200 Nino Novak nn.l...@kflog.org wrote: s/fdo#[0-9]*/{{\0}}/g s/bnc#[0-9]*/{{\0}}/g s/i#[0-9]*/{{\0}}/g s/i#/issue#/g s/^\+ \(.*\)$/== \1 ==/ s/^. /* / FWIW, I use /lp#[0-9]*/ for the rare launchpad fix. Best, Bjoern -- https://launchpad.net/~bjoern-michaelsen ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] permission 444/555 of installation files
Nice, you've almost fixed all of them expect some *rc files: -r--r--r-- 1 root root 126 Jul 18 22:01 /build/pkg/libreoffice-common/usr/lib/libreoffice/ure/bin/versionrc -r--r--r-- 1 root root 1.1K Jul 18 22:01 /build/pkg/libreoffice-common/usr/lib/libreoffice/basis3.4/program/unorc -r--r--r-- 1 root root 44 Jul 18 22:01 /build/pkg/libreoffice-common/usr/lib/libreoffice/basis3.4/program/pythonloader.unorc -r--r--r-- 1 root root 1.5K Jul 18 22:01 /build/pkg/libreoffice-common/usr/lib/libreoffice/basis3.4/program/fundamentalbasisrc -r--r--r-- 1 root root 150 Jul 18 22:01 /build/pkg/libreoffice-common/usr/lib/libreoffice/basis3.4/program/versionrc -r--r--r-- 1 root root 50 Jul 18 22:01 /build/pkg/libreoffice-common/usr/lib/libreoffice/program/redirectrc -r--r--r-- 1 root root 56 Jul 18 22:01 /build/pkg/libreoffice-common/usr/lib/libreoffice/program/setuprc -r--r--r-- 1 root root 1.5K Jul 18 22:01 /build/pkg/libreoffice-common/usr/lib/libreoffice/program/fundamentalrc -r--r--r-- 1 root root 421 Jul 18 22:01 /build/pkg/libreoffice-common/usr/lib/libreoffice/program/versionrc -Andy ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] [PATCH] Work around a bash-ism in configure.in
Hi Mike, On 07/18/11 04:23, Michael Meeks wrote: Hi Mike, On Fri, 2011-07-15 at 12:34 -0700, Mike Eberdt wrote: This patch eliminates a test: : bad number diagnostic on FreeBSD. Contributed under LGPL3+/MPL. Thanks; I pushed it, and then updated it to use: if test z$EUID = z0 -a z`uname -o 2/dev/null` = zCygwin; then which Tor suggested IIRC - in general we don't comment previous problems with the code in the new, cleaned-up code - that would belong in the commit message :-) Oops, sorry; I'll try to put comments in the right place going forward. Thanks for your fix ! how are we looking on FreeBSD ? :-) The official FreeBSD port (currently at LO 3.3.3 + some platform-specific patches), builds fine. I'm not involved in the maintenance of that port at all, so I don't know if an upgrade to LO 3.4.x is presenting problems or is even in the works. However, I thought it would be nice (at least for casual hacking purposes) to be able to build LO HEAD outside of the FreeBSD Ports infrastructure, using the 'bootstrap' repo (the official port uses the 'build' tarball). That approach doesn't get far at all: it dies linking 'registry/tools/regmerge' because it can't find one of its own shared libs. I've attached build_error.log and FreeBSDAMDEnv.Set.sh in case someone who understands the build system wants to take a look. My 'dmake' fu is basically non-existent, so I haven't poked at this seriously yet. And as I understand it, a transition is underway to GNU make, so I guess I've been a bit reluctant to put a lot of time into learning the old build system. However, I didn't notice any info in the Developers section of the Wiki about using the new build system, or helping with the migration, etc. Is this something that mere mortals can contribute to, or should I just wait for the grand unveiling? :) Thanx, ME # # # LibreOffice 350 build environment file for: x86_64-unknown-freebsd8.2. # Generated on: Mon Jul 18 15:00:57 PDT 2011 # Source this file to set up the build environment. # 1. exec sh/bash/ksh # 2. source FreeBSDAMDEnv.Set.sh # # SRC_ROOT=/hacking/libreoffice/src OOO_SHELL=/usr/local/bin/bash JAVA_HOME=NO_JAVA_HOME L10N_MODULE=/hacking/libreoffice/src/translations # # Platform independent constant values. # OOODMAKEMODE=YES CALL_CDECL=TRUE NO_REC_RES=TRUE SO3=TRUE SOLAR_PLUGIN=TRUE UPDATER=YES VCL=TRUE ENABLEUNICODE=TRUE NO_BSYMBOLIC=TRUE ENABLE_DEBUG_STL=FALSE EXTERNAL_WARNINGS_NOT_ERRORS=TRUE PRODUCT=full PROFULLSWITCH=product=full PROEXT=.pro ENABLE_NSS_MODULE=NO BUILD_VER_STRING=built on FreeBSD 8.2 DEFAULT_TO_ENGLISH_FOR_PACKING=yes ENABLE_GTK=TRUE GTK_CFLAGS=-D_THREAD_SAFE -D_REENTRANT -I/usr/local/include/gtk-2.0 -I/usr/local/lib/gtk-2.0/include -I/usr/local/include/atk-1.0 -I/usr/local/include/cairo -I/usr/local/include/gdk-pixbuf-2.0 -I/usr/local/include/pango-1.0 -I/usr/local/include/gio-unix-2.0/ -I/usr/local/include -I/usr/local/include/glib-2.0 -I/usr/local/lib/glib-2.0/include -I/usr/local/include/pixman-1 -I/usr/local/include/freetype2 -I/usr/local/include/drm GTK_LIBS=-pthread -L/usr/local/lib -lgtk-x11-2.0 -lgdk-x11-2.0 -latk-1.0 -lpangocairo-1.0 -lXext -lXrender -lXinerama -lXi -lXrandr -lXcursor -lXcomposite -lXdamage -lpangoft2-1.0 -lgio-2.0 -lXfixes -lcairo -lX11 -lpango-1.0 -lfreetype -lfontconfig -lgdk_pixbuf_xlib-2.0 -lgdk_pixbuf-2.0 -lgobject-2.0 -lgmodule-2.0 -lpng -lz -lm -lgthread-2.0 -lglib-2.0 GTHREAD_CFLAGS=-D_REENTRANT -I/usr/local/include/glib-2.0 -I/usr/local/lib/glib-2.0/include GTHREAD_LIBS=-pthread -L/usr/local/lib -lgthread-2.0 -lglib-2.0 ENABLE_STATIC_GTK=FALSE ENABLE_OPENGL=TRUE ENABLE_PDFIMPORT=NO ENABLE_MINIMIZER=NO ENABLE_PRESENTER_SCREEN=NO ENABLE_REPORTBUILDER=NO ENABLE_SCRIPTING_BEANSHELL=NO ENABLE_SCRIPTING_JAVASCRIPT=NO ENABLE_SCRIPTING_PYTHON=YES ENABLE_QUICKSTART_LIBPNG=TRUE SYSTEM_JFREEREPORT=NO WITH_THEMES= default crystal hicontrast oxygen tango WITH_HELPPACK_INTEGRATION=YES WITH_EXTENSION_INTEGRATION=NO ENABLE_MEDIAWIKI=NO ENABLE_GSTREAMER=TRUE MOC=moc MOC4=moc PSPRINT=TRUE MKDEPENDSOLVER=TRUE TARFILE_LOCATION=/hacking/libreoffice/src/../downld GIT_REPO_NAMES=artwork base calc components extensions extras filters help impress libs-core libs-extern libs-extern-sys libs-gui postprocess sdk testing ure writer # # Platform dependent constant values. # COM=GCC COMPATH=/usr CPU=X CPUNAME=X86_64 GUI=UNX GUIBASE=unx GVER=VCL OS=FREEBSD OSVERSION=802508 PTHREAD_LIBS=-pthread OUTPATH=unxfbsdx INPATH=unxfbsdx.pro PATH_SEPERATOR=: # # Variable values. # PERL=/usr/local/bin/perl GNUMAKE=gmake GNUTAR=tar UPD=350 WORK_STAMP=OOO350 SOLARVER=/hacking/libreoffice/src/solver/350 SOLARVERSION=$SOLARVER WORKDIR=$SOLARVER/$INPATH/workdir OUTDIR=$SOLARVER/$INPATH SOLARENV=/hacking/libreoffice/src/solenv SOLARDEFIMG=/hacking/libreoffice/src/default_images
Re: [Libreoffice] [PATCH] Work around a bash-ism in configure.in
Hi Christian, On 07/18/11 04:56, Christian Lohmaier wrote: Hi MIke, *, On Fri, Jul 15, 2011 at 9:34 PM, Mike Eberdtlibr...@comcast.net wrote: Configure.in eventually gets parsed by /bin/sh, and on Free/Net/OpenBSD /bin/sh is not synonymous with bash. Therefore, bash-isms in configure.in can be problematic. Oops, I mistyped; I meant to say that it is 'configure' (not 'configure.in') that's getting parsed by /bin/sh. I don't understand your patch then. -if test $EUID -eq 0 -a z`uname -o 2/dev/null` = zCygwin ; then +# $EUID is a bash-ism, so we can't assume its existence. [...] +if test z`uname -o 2/dev/null` = zCygwin test $EUID -eq 0; then So while you converted test and link to, the $EUID that according to your comment is a bashism still is used. So how does this solve anything? While /bin/sh on FreeBSD always evaluates both X and Y in if test X -a Y, it does correctly skip Y when X is false in if X Y. So as long as the OS check is first, we don't evaluate $EUID except on Cygwin. However, the form that Michael Meeks checked in is much better, since it avoids the subtlety in the ordering of the 2 tests, while still using test -a like in the rest of the file. (Besides that, configure is created with shebang of /bin/bash here) Hmm. For me, 'configure' begins with #!/bin/sh on Linux as well as on FreeBSD. ME ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] A thought about fdo#38095: Character classification for Western or Asian text font
Hi Caolán, On Mon, 18 Jul 2011 21:20:02 +0100, Caolán McNamara caol...@redhat.com wrote: On Tue, 2011-07-19 at 01:14 +0900, Takeshi Abe wrote: Hi all, According to a discussion on the Japanese local mailing list, the following issue may prevent some users from migrating to 3.4.x series: https://bugs.freedesktop.org/show_bug.cgi?id=38095 I am not yet sure how it works in 3.3.x, but at least for Calc on master, there seems a gap around determining a font from the Unicode Script Property of content of a cell. This will be an easy fix, I'll take it. I had seen the change in passing and stuck it as a to-do to examine it, but hadn't realized it wasn't just an obscure edge-case that these number-forms were no longer categorized the same as they had been historically. IIRC they used to be WEAK chars but are now being categorized as LATIN like the numbers/letters they resemble. Hm, that sounds likely. Thanks a lot for your attention. Cheers, -- Takeshi Abe i18npool/source/breakiterator/breakiteratorImpl.cxx will be the right place. C. ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
[Libreoffice] [PATCH] Replace List with std::vector ImplBmpObj*
I'd like someone doing a Unix build to review this for me. I compile Mac and this is Unix only code so I don't just want to push and hope... Thanks, Joe P. 0001-Replace-List-with-std-vector-ImplBmpObj.patch Description: Binary data ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] [MHST][Easy Hack] Get current page number for printing
Hi TaTung, all! Sorry for answering so late ... Am Samstag, den 16.07.2011, 21:40 +0700 schrieb Ta Duc Tung: On 07/16/2011 07:29 PM, Christoph Noack wrote: [...] We are trying to hack into this easy hack https://bugs.freedesktop.org/show_bug.cgi?id=38830 We're still studying source code. We have 2 questions: Great! [...] 2. How to add another RadioButton in to print dialog? We can enable and disable existing RadioButton like All, Pages, or Selection but it's still mystery for adding new RadioButton. We want to add a Current Page radio button just below Pages radio button. Anyone please help us :) Is there a specific reason to use another radio button? OOo once added the current page number to the Pages field - which worked okay for many users. I think it will be more intuitive and easier to use if we have a Current Page radio button there. As a beginner, I won't aware that the number in the text box is current page because normally, I scarcely care which page I was. Okay, that's a good point - when we've designed the today's printing dialog, we tried hard to get rid of many of the old options to make it both useful and clean (although current page was never part of that). I've noticed that you already provided a patch for that - thanks! And here I'd like to ask you to help me to get some understanding ... reason: I lack C++ coding skills. So ... * Where is Current page located? Between Pages and Selection, I assume, or? * Please take care that the dialog does not grow in height (too much), since it won't be usable on netbooks. * Does this patch also provide Current Page for Draw/Impress? This would be essential for consistency. If yes, then the terminology for Impress only) is different: Current Slide. Only for Calc and Math it is not needed or makes less sense. Could you please provide some hints to me? Thanks! Mmh, looking at a recent LibO version, I notice that somebody moved ... the reverse order option to a weird position (makes no sense, breaks spacing etc.) But that's another story. Cheers, Christoph ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] [PATCH] Work around a bash-ism in configure.in
So as long as the OS check is first, we don't evaluate $EUID except on Cygwin. But it wasn't as such the existence or not of $EUID that was the problem here, was it? The problem was the use of -eq operator when the left-hand side of the operator, $EUID , was an empty string and thus not a number, no? BTW, the z prefixes that Michael added are not needed. Surely all modern shells (or standalone test commands) handle test = 0 just fine. The convention to avoid nonempty arguments to test is just hypercorrectness by now. (For any platform where one would seriously consider building LibreOffice, at least.) The configure.in script uses test $foo =already. --tml ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] [PATCH][REVIEW][PUSHED 3-4] SetNamedRanges speedup fdo#35357
On Mon, Jul 18, 2011 at 12:28:55PM -0400, Kohei Yoshida wrote: On Mon, 2011-07-18 at 11:47 -0400, Kohei Yoshida wrote: On Mon, 2011-07-18 at 10:25 -0400, Kohei Yoshida wrote: I'd also like to push this to the -3-4 branch, but let me just do a quick test on ODS import of named ranges to make sure if works, then I'll push. And I'm glad I did. This change unfortunately causes a problem on global named range import. It incorrectly imports them as external references. I'm looking into this at the moment. This commit fixes it http://cgit.freedesktop.org/libreoffice/calc/commit/?id=ce470bf9f95ae53c29356976b2c5e44cd6c73fed There was a bug in the RangeNameInserter function object which was exposed by your change. Apparently LibO / OOo sometimes write range names this way: $Sheet1.$A$1:.$A$3 (the leading dot in the end range address) which was supposed to be corrected to $Sheet1.$A$1:$A$3 (without the dot in the end range address) before passing it to a range name object. But the code incorrectly used the original, uncorrected version of the range string, which caused the problem. Could someone please give this change a quick review so we can cherry-pick this to the -3-4 branch along with Laurent's original patch? I can sign off on Laurent's original patch but I can't self-review this one. +1 from me. (I haven't pushed Laurent's patch.) D. ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] [PATCH] Work around a bash-ism in configure.in
My 'dmake' fu is basically non-existent, so I haven't poked at this seriously yet. [...] a transition is underway to GNU make Oh, (unfortunately) you have it backwards regarding the complexity;) dmake is mostly a normal make with few special features. Understanding how things work in the old build mechanism is not that hard, it does not require re-adjusting your perception of what makefiles are and how they look. Basically, it uses plain makefiles, that include a set of common headers with makefile snippets. There are relatively normal-looking rules where targets depend on pre-requisites with a recipe of commands following. Sure, there are nests of .IF/.ELSE/.ENDIF conditionals. Sure, the actual rules are in the headers and use variables defined in the outermost directory-specific makefile.mk. But still... It's the new GNU Make based system that is much harder to understand... because there the makefiles don't look normal at all, as all the actual rules are generated dynamically by eval and call function invocations. Sure, looking at the top level of a makefile for a deliverable in a module you can guess what the stuff means, I am not saying *that* is hard. But understanding how it actually works, and doing changes to that, can be hard. --tml ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] [PATCH] Work around a bash-ism in configure.in
On Mon, Jul 18, 2011 at 05:33:56PM -0700, Mike Eberdt wrote: Hi Christian, On 07/18/11 04:56, Christian Lohmaier wrote: Hi MIke, *, On Fri, Jul 15, 2011 at 9:34 PM, Mike Eberdtlibr...@comcast.net wrote: Configure.in eventually gets parsed by /bin/sh, and on Free/Net/OpenBSD /bin/sh is not synonymous with bash. Therefore, bash-isms in configure.in can be problematic. Oops, I mistyped; I meant to say that it is 'configure' (not 'configure.in') that's getting parsed by /bin/sh. I don't understand your patch then. -if test $EUID -eq 0 -a z`uname -o 2/dev/null` = zCygwin ; then +# $EUID is a bash-ism, so we can't assume its existence. [...] +if test z`uname -o 2/dev/null` = zCygwin test $EUID -eq 0; then So while you converted test and link to, the $EUID that according to your comment is a bashism still is used. So how does this solve anything? While /bin/sh on FreeBSD always evaluates both X and Y in if test X -a Y, it does correctly skip Y when X is false in if X Y. So as long as the OS check is first, we don't evaluate $EUID except on Cygwin. However, the form that Michael Meeks checked in is much better, since it avoids the subtlety in the ordering of the 2 tests, while still using test -a like in the rest of the file. Why don't we use `id -u` instead of $EUID? That would work even in the old form (with -a and -eq) and AFAIK it is portable :) D. ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice