[dev] Re: fixing postgresql sdbc connector
On Thu, Jul 21, 2011 at 02:40:30PM +0430, Jan wrote: The build instructions on http://dba.openoffice.org/drivers/postgresql/index.html seem totally outdated since they use cvs, not hg. Yeah, because that driver *is* in cvs, not in hg. Separate project etc, never merged into the official code. You need to get that stuff and add it to your hg tree. http://cgit.freedesktop.org/libreoffice/build/tree/patches/postgresql contains patches you can add on top of your hg and do stuff(tm). (Icreated those patches and used them until discovering that bug after which it was disabled - see http://bugs.debian.org/620100) Grüße/Regards, René -- - To unsubscribe send email to dev-unsubscr...@openoffice.org For additional commands send email to sy...@openoffice.org with Subject: help
[dev] Re: Off the records (Ubuntu's LO) ... (Re: Re: Genuine OOo in distributions ? (Re: OOo installation packages for Linux, a few (easy) questions
Hi, On Tue, Jun 07, 2011 at 09:56:35AM +0200, Rene Engelhard wrote: rene@frodo:~$ find /usr/lib/libreoffice/ -name unoinfo* /usr/lib/libreoffice/program/unoinfo rene@frodo:~$ less /usr/lib/libreoffice/program/unoinfo in Debian, is that what you mean? Good that this script isn't documented at all and has no help and you need to read the code, but it even works: $ /usr/lib/libreoffice/program/unoinfo c++ /usr/lib/libreoffice/program/../basis-link/ure-link/librene@frodo:~$ (Note that it doesn't have a linebreak...) $ /usr/lib/libreoffice/program/unoinfo java 0/usr/lib/libreoffice/program/../basis-link/ure-link/share/java/ridl.jar/usr/lib/libreoffice/program/../basis-link/ure-link/share/java/jurt.jar/usr/lib/libreoffice/program/../basis-link/ure-link/share/java/juh.jar/usr/lib/libreoffice/program/../basis-link/program/classes/unoil.jar/usr/lib/libreoffice/programrene@frodo:~$ (No linebreak here either.) Grüße/Regards, René -- - To unsubscribe send email to dev-unsubscr...@openoffice.org For additional commands send email to sy...@openoffice.org with Subject: help
[dev] Re: Off the records (Ubuntu's LO) ... (Re: Re: Genuine OOo in distributions ? (Re: OOo installation packages for Linux, a few (easy) questions
On Tue, Jun 07, 2011 at 11:03:00AM +0200, rony wrote: unoinfo java and interestingly unoil.jar is missing! Did a locate unoil.jar after an updatedb, but it is not installed. So this is Obviously totally untrue. I don't think Ubuntu removed unoil.jar, that would break everything Java-ish. Did you install libreoffice-java-common? (Reading helps.) That one contains the common Java atuff (except the URE Java components, which are in URE) (a apt-file search or a search on packages.debian.org/.ubuntu.com would have shown you where unoil.jar is) As Java is not a must-have, it was, is, and always will be in an optional package. probably the culprit on Ubuntu (assuming that your Debian installation package has it), breaking the office. I am 100% sure Ubuntu also has it. What breaks here is that you cannot properly use your package manager. BTW, René, where could one download your version of the deb package to test it? Debian. Where else? Grüße/Regards, René -- - To unsubscribe send email to dev-unsubscr...@openoffice.org For additional commands send email to sy...@openoffice.org with Subject: help
[dev] Re: Off the records (Ubuntu's LO) ... (Re: Re: Genuine OOo in distributions ? (Re: OOo installation packages for Linux, a few (easy) questions
Hi, On Tue, Jun 07, 2011 at 12:28:18PM +0200, rony wrote: This is the story for the current Ubuntu 11.04, plain-vanilla installation, after having gone further: * Ubuntu 11.04 does *not* install any of the Java classes in /usr/lib/libreoffice/basis-link/program/classes, though it installs the Java support in /usr/lib/libreoffice/basis-link/ure-link/share/java The latter is URE stuff, the other one is LibO stuff Did you install libreoffice-java-common? No, am I supposed to do so on Ubuntu? Where would I learn this from? Package descriptions? And if you look for unoil.jar you could have searched for it and you'd have seen it. Where am I told to look up for reading materials on Ubuntu, if Ubuntu deviates so badly from the reference package of OOo/LO? It doesn't. It just packages it up so that people don't get useless stuff if they don't need it. If they want, they can just install it. And if clueless people just install libreoffice (or somethig needing Java liuke Base, they of course get -java-comon) How would I know? By using your brain. [But thanks for the pointer at apt-file.] The reference installations from OOo/LO work. Why should I expect that Ubuntu (and maybe Debian) cripple OOo/LO? We don't. It's just packaged differently. You can install whatever you need. Just install the goddamn package. *Who* defines that Java is *not a must-have* in todays world? Everytime the Java thing came up, anyone claims that Java ins not necessary for functioning of the office. Except Base, which depends on it. If you need Java functionality, instakl it (And note we patch LibO to say install libreoffice-java-common when Java doesn't work because the necessary stuff is not there). You get told. Maybe not when you just do extension stuff (but then you should be able to use your package system) but for normal use. If you do not install it, you cripple OOo/LO bad time! E.g. one cannot use scripts written in JavaScript and much more (like databse, xsl/xml, report True. The Database of course depends on the Java stuff. Get your facts straight, read, think. Same for the filters you mention. They also depends on -java-common. If you have no clue, you should shut up... etc.). This stance may be o.k. for amateurish ..., but is not o.k. for professionals living in todays world and who have been constantly witnessing the changes and raising needs and taking advantage of them. I am not arguing against java per se. Read again. Would you care for an URL? No. Grüße/Regards, René -- - To unsubscribe send email to dev-unsubscr...@openoffice.org For additional commands send email to sy...@openoffice.org with Subject: help
[dev] Re: Off the records (Ubuntu's LO) ... (Re: Re: Genuine OOo in distributions ? (Re: OOo installation packages for Linux, a few (easy) questions
Hi, On Tue, Jun 07, 2011 at 12:45:09PM +0200, rony wrote: Thank you very much for the link! In the meantime further information has become available, thanks to Renés comments, so the installation seems to be (very unfortunate!) intentional (to cripple LO)! :( There IS NOTHING CRIPPELD. Just install the correct packages and you're done. That's the same with the normal install. You also would call your OOO crippeled if you didn't install -xstfilter or so? No, you just install that package. Do install your packages here, too and be done. (And note a default apt-get install libreoffice at least on Debian *will* give you also -java-common, because it picks up Base which picks up libreoffice-java-common) You obviously have no clue at all and resist to even look up basic stuff. Grüße/Regards, René -- - To unsubscribe send email to dev-unsubscr...@openoffice.org For additional commands send email to sy...@openoffice.org with Subject: help
[dev] Re: Off the records (Ubuntu's LO) ... (Re: Re: Genuine OOo in distributions ? (Re: OOo installation packages for Linux, a few (easy) questions
Hi, On Tue, Jun 07, 2011 at 01:52:41PM +0200, rony wrote: If packages deviate from the reference packages (OOo or LO), they get crippled, like it or not. Nonsense. They just ha a bit other structure, the functionality is supposed to be there. Just install the one package you miss. Insulting people does not change facts and reality either, but is quite You don't? Aha. I see insults from you too. You cripple XYZ. When we don't. (And you say we do that intentionally, which is also a insult) Wer im Glashaus sitzt... Grüße/Regards, René -- - To unsubscribe send email to dev-unsubscr...@openoffice.org For additional commands send email to sy...@openoffice.org with Subject: help
[dev] Re: Off the records (Ubuntu's LO) ... (Re: Re: Genuine OOo in distributions ? (Re: OOo installation packages for Linux, a few (easy) questions
Hi, On Tue, Jun 07, 2011 at 02:28:53PM +0200, rony wrote: On 07.06.2011 14:07, Rene Engelhard wrote: Insulting people does not change facts and reality either, but is quite You don't? Aha. I see insults from you too. You cripple XYZ. When we don't. (And you say we do that intentionally, which is also a insult) O.K., I take this one on my shoulders, sorry, if you - as a person - took this as an insult, it was never meant to address you. (From your mailings it is clear that you have not removed functionality from the reference OOo or LO packages from the respective Debian package, therefore you have never been meant to be addressed.) No? P.S.: In the case of Ubuntu I would expect that they at least advertise very prominently that their OOo/LO distribution is not a standard one, They base on our packages, we have the same structure. Thus I am defending this. but one that lacks features compared to the standard OOo/LO packages. Nonsense. Again, install the correct package. You also have to install the correct packages when you install the standard OOo/LO packages. If you leave one of those out, you also loose something. Would cou call that crippled? No. This way, everyone would be aware of this fact and if the information was given, how to add the missing pieces (sudo apt-get install librefoffice), end-users would become able to upgrade the special Ubuntu OOo/LO to the full package on their own. Which missing pieces? What did you install in the first place? Any clueless user should use libreoffice as the start point, when Ubuntus install does something else, ok, but then again, people can install stuff afterwards. (That said, if you install Debian and choose desktop you get libreoffice installed, no idea what Ubuntu does) Would you also argue it's a crippling of someone installed a standard OOo/LO package but left -xsltfilter out? Which he could? And wonders why those don't work? No. Same story here. But we are going in cirlces, I grok that you don't care about proper packaging and making optional stuff choosable. Thankfully distros are not following every stupidity and make life of their users easier by allowing them to get rid of stuff THEY don't want/need. Grüße/Regards, René -- - To unsubscribe send email to dev-unsubscr...@openoffice.org For additional commands send email to sy...@openoffice.org with Subject: help
Re: [dev] EducOOo donated code for ARM Linux port
On Sat, Feb 19, 2011 at 08:41:43AM +0100, eric b wrote: Hi, For your information : http://www.openoffice.org/issues/show_bug.cgi?id=117017 EducOOo donated code for ARM Linux port makes it sound you did the port. Wrong. Stop your propaganda, please. Now to the technical points: ARM_TARGET=7 ehm, no. *If* you want to set a minimal baseline, use armv4t. Assuming anyone has the latest and greatest ARM is bogus. And +if test x$with_arm_target x6 + thenAC_MSG_RESULT([arm target is armv5 or inferior]) + ARM_TARGET=5 aha, let's break even more older ARMs... Intreresting you even dropped the correct list from your CCs (dev@porting). What does marketing to do here instead? Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
[dev] Re: [porting-dev] Re: [dev] EducOOo donated code for ARM Linux port
Hi, On Sat, Feb 19, 2011 at 01:51:02PM +0100, Rene Engelhard wrote: ARM_TARGET=7 Oh, and this: +# Default is armv7 mini. If you build +# for an older version (could be slow), +# please adapt to your needs + +.IF $(ARM_TARGET) == ARMV6 +ARCH_FLAGS+=-march=armv6 -mfloat-abi=softfp -D__SOFTFP__ +CDEFS+=-DARMV6 +.ENDIF + +.IF $(ARM_TARGET) == ARMV7 +ARCH_FLAGS+=-march=armv7-a -mtune=cortex-a8 -mfpu=neon -mfloat-abi=softfp -D__SOFTFP__ +CDEFS+=-DARMV7 +.ENDIF + is broken, too. *If* you want to optimize for speed, you want hardfp, and IMHO you don't need to specify -mfloat-abi=softfp -D__SOFTFP__ manually in either case, that's the job of your toolchain. Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
[dev] Re: [porting-dev] Re: [dev] EducOOo donated code for ARM Linux port
On Sat, Feb 19, 2011 at 01:55:50PM +0100, Rene Engelhard wrote: +.IF $(ARM_TARGET) == ARMV6 This is broken, too, afaics. In configure you set ARM_TARGET to e.g. 6 (WITHOUT ARMV) but there you use it WITH ARMV... Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Re: [porting-dev] Re: [dev] EducOOo donated code for ARM Linux port
On Sat, Feb 19, 2011 at 02:37:39PM +0100, Rene Engelhard wrote:
On Sat, Feb 19, 2011 at 01:55:50PM +0100, Rene Engelhard wrote:
+.IF $(ARM_TARGET) == ARMV6
This is broken, too, afaics.
In configure you set ARM_TARGET to e.g. 6 (WITHOUT ARMV)
but there you use it WITH ARMV...
I fixed some more (even syntax errors) in the patch, can't attach it
right now since IZ is down, but here it is.
Will attach a new patch, which
- defaults to armv4t
- fixes syntax errors in the patch (test has no and but -lt and -gt,
-a test is broken (either or -a without test, missing $ before build_cpu)
- remove the x$something nonsense for test which doesn't play well with
test and -lt/-gt
- you should not assume the build_cpu is stricly arm. on my debian sid on a
sheevaplug the triplet is armv5tel-unknown-linux-gnu when not specifying
arm-* manually on the configure line. There's a reason we have arm* in
configure.ins case statements :)
The original patch obviously wasn't tested at all...
Grüße/Regards,
René
diff -r 67e476e04669 configure.in
--- a/configure.in Fri Nov 12 20:02:04 2010 +0100
+++ b/configure.in Fri Feb 18 21:47:58 2011 +0100
@@ -828,6 +828,16 @@
AC_ARG_ENABLE(dependency-tracking,
[ --disable-dependency-tracking Disables generation of dependency information.
],,)
+AC_ARG_WITH(arm-target,
+[ --arm-targetThe minimal targeted arm processor
+ used for the build environment.
+ Cases :
+arm-target 6 : armv4t compatibility
+arm-target = 6 : exact armv6 compatibility
+ arm-target 6 : armv7-a compatibility
+
+ Usage: --with-arm-target=7
+],with_arm_target=$withval,with_arm_target=4)
BUILD_TYPE=OOo
@@ -5803,6 +5813,24 @@
AC_SUBST(GSTREAMER_CFLAGS)
AC_SUBST(GSTREAMER_LIBS)
+dnl ===
+dnl Check the ARM target
+dnl ===
+
+if test $_os = Linux echo $build_cpu | $GREP -q arm; then
+# default value
+ARM_TARGET=ARMV4T
+AC_MSG_CHECKING([which ARM processor optimization to use])
+if test $with_arm_target -lt 6; then
+ ARM_TARGET=ARMV4T
+elif test $with_arm_target = 6; then
+ ARM_TARGET=ARMV6
+elif test $with_arm_target -gt 6; then
+ ARM_TARGET=ARMV7
+fi
+AC_MSG_RESULT([$ARM_TARGET])
+AC_SUBST(ARM_TARGET)
+fi
dnl ===
dnl Check whether the Cairo libraries are available.
diff -r 67e476e04669 sal/osl/unx/interlck.c
--- a/sal/osl/unx/interlck.c Fri Nov 12 20:02:04 2010 +0100
+++ b/sal/osl/unx/interlck.c Fri Feb 18 21:47:58 2011 +0100
@@ -134,6 +134,54 @@
return nCount;
}
+#elif defined ( GCC ) defined ( ARM )
+
+/*/
+/* osl_incrementInterlockedCount */
+/*/
+oslInterlockedCount SAL_CALL osl_incrementInterlockedCount(oslInterlockedCount* pCount)
+{
+#if defined( ARMV7 ) || defined( ARMV6 )
+register oslInterlockedCount nCount __asm__ (r1);
+int nResult;
+
+__asm__ __volatile__ (
+1: ldrex %0, [%3]\n
+ add %0, %0, #1\n
+ strex %1, %0, [%3]\n
+ teq %1, #0\n
+ bne 1b
+: =r (nCount), =r (nResult), =m (*pCount)
+: r (pCount)
+: memory);
+
+return nCount;
+#else
+return __sync_add_and_fetch( pCount, 1 );
+#endif
+}
+
+oslInterlockedCount SAL_CALL osl_decrementInterlockedCount(oslInterlockedCount* pCount)
+{
+#if defined( ARMV7 ) || defined( ARMV6 )
+register oslInterlockedCount nCount __asm__ (r1);
+int nResult;
+
+__asm__ __volatile__ (
+0: ldrex %0, [%3]\n
+ sub %0, %0, #1\n
+ strex %1, %0, [%3]\n
+ teq %1, #0\n
+ bne 0b
+: =r (nCount), =r (nResult), =m (*pCount)
+: r (pCount)
+: memory);
+return nCount;
+#else
+return __sync_sub_and_fetch( pCount, 1 );
+#endif
+}
+
#else
/* use only if nothing else works, expensive due to single mutex for all reference counts */
diff -r 67e476e04669 set_soenv.in
--- a/set_soenv.in Fri Nov 12 20:02:04 2010 +0100
+++ b/set_soenv.in Fri Feb 18 21:47:58 2011 +0100
@@ -1694,6 +1694,7 @@
ToFile( SET_EXCEPTIONS,$SET_EXCEPTIONS,e );
ToFile( use_shl_versions, $use_shl_versions, e );
ToFile( FLIPCMD, $FLIPCMD, e );
+ToFile( ARM_TARGET, @ARM_TARGET@, e );
#
# Writing the variables to file.
# (c = comment, e = environment variable, a = alias, n = newline )
diff -r 67e476e04669 solenv/inc/unxlngr.mk
--- a/solenv/inc/unxlngr.mk Fri Nov 12 20:02:04 2010 +0100
+++ b/solenv/inc/unxlngr.mk Fri Feb 18 21:47:58 2011 +0100
@@ -32,3 +32,19 @@
CDEFS+=-DARM32
CFLAGS+=-fno-omit-frame-pointer
DLLPOSTFIX=lr
+
+.IF $(ARM_TARGET) == ARMV4T
+ARCH_FLAGS+=-march=armv4t
+CDEFS+=-DARMV4T
+.ENDIF
Re: [dev] EducOOo donated code for ARM Linux port
On Sat, Feb 19, 2011 at 06:02:44AM -0700, Tor Lillqvist wrote: EducOOo donated code for ARM Linux port makes it sound you did the port. Me, I didn't understand if EducOOo is the subject or object here... (And with issuezilla being down, I couldn't check the link either.) Was it donated to EducOOo, or did EducOOo donate it to somebody else? And what does donate mean, exactly, in the context of Open Source software? Or is this not Open Source? According to the issue they did do some arm6/armv7 optimizations and donated the change back (Though the patch was buggy and not-working, see my later comments) but the comment made it look that they donated (the) code for ARM Linux port. That was what I objected to. Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] ARM port
Hi, On Tue, Feb 15, 2011 at 10:38:15AM +0100, eric b wrote: OpenOffice.org is already ported to Linux ARM, oabi and eabi, le and be. Sure. But the current port is generic and needs some improvements at several places anyway, plus, the current OpenOffice.org is unusable on such machines. Wrong. To be honest, only a stripped down version like OOoLight or OOo4Kids is today - sort of- correct to my eyes. No, there's powerful multi-core ARMs. And there's armhf even mitigating the floating point thing. Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] ARM port
On Tue, Feb 15, 2011 at 12:08:47PM +0100, eric b wrote: Sorry, but I tend to disagree : can you please tell me the arm version choosen by default in OOo source code ? I'm just working at Whatever default your compiler does. $ file /usr/lib/ure/lib/libuno_sal.so.3 /usr/lib/ure/lib/libuno_sal.so.3: ELF 32-bit LSB shared object, ARM, version 1 (SYSV), dynamically linked, stripped Nothing per default, yes, though. (Debioan armel targets armv4t anyway, afaicr) the optimization and what I see, tells me it's not optimized at all. But I can be wrong, and you'll tell me then ... Even if you were right, how does it matter? OOo also is optimized only for pentiumpro on i386. And if you wished, you can optimize for whatever ARM you see fit. It is obvious, that the current OpenOffice.org can run (I know) on current ARM, but will be slow as hell on the machine anybody can buy, today. Nobody will use that, and a lot of work is mandatory. Wrong. I tried it on my SheevaPlug, and it's quite usable. Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Segmentatiefout in OOo -3.2.1 (Dutch version)
On Sat, Oct 23, 2010 at 06:47:42PM +0200, Ben Engbers wrote:
I have used the insturcions from
http://wiki.services.openoffice.org/wik ...
to install van LibreOffice 3.3. beta. Great!
This is not a libreoffice discussion list (unfortunately)
The relavan part of the output from 'strace ./soffice' is as far as I
can see:
No, it's not.
open(/usr/share/locale/nl/LC_MESSAGES/libc.mo, O_RDONLY) = 3
fstat64(3, {st_mode=S_IFREG|0644, st_size=134850, ...}) = 0
mmap2(NULL, 134850, PROT_READ, MAP_PRIVATE, 3, 0) = 0xb755d000
close(3) = 0
write(2, ./soffice: regel 127: 5127 Segm..., 102./soffice: regel 127:
5127 Segmentatiefout (geheugendump gemaakt) $sd_prog/$sd_binary $@
) = 102
It finds it, see the last open. Otherwise you wouzld have gotten
the correct Segmentaton fault instead of the nl translation of it :)
Does this nformation help?
No.
Grüße/Regards,
René
-
To unsubscribe, e-mail: dev-unsubscr...@openoffice.org
For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] macro run error
On Thu, Sep 30, 2010 at 09:11:53AM +0200, Stephan Bergmann wrote: Strange. Which platform are you on, and where did you get OOo from? (The lr in deploymnetlr.uno.so indicates a platform, but it is none of the standard ones.) What might give a clue is to call Linux/ARM. Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Classpath management in Java UNO extensions
Hi, On Mon, Sep 20, 2010 at 09:56:50AM -0400, Benson Margulies wrote: If our UNO extension depends on some additional JAR files, do we need to manifest them somehow, or just pack them into the OXT? You can do either of them. Class-Path: works, too if you for whatever reason don't want to embed the jar. (But then you of course need the dependent jar present outside of the oxt) Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Missing vcl resource
On Thu, Sep 09, 2010 at 08:58:17AM +0200, Malte Timmermann wrote: 3.2.1.4 sounds like ooo-build, not vanilla OOo, so please use the ooo-build mailing list for such questions... He already got told, but as we already notice for other questions, he can't read and does not follow advice. That said, ooo-build 3.2.1.4 of course works fine also when installed, so he (unsurprsingly) must have either done something wrong there on install or did bogus things in the build/after the install. Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] no module named uno error
Hi, On Mon, Sep 06, 2010 at 08:19:01PM -0700, Soohong Min wrote: First time, I've built Openoffice without --with-system-python, but the result is same. You obviously didn't read completely. I said that ooo-build defaults to using sytstem-python so you *of course* get no internal python. That's called logic. I found interesting point that I've installed Openoffice package on Ubuntu, by the way I also cannot find internal python in usr/lib/openoffice/program. As proven here. I already said that disros usually build with system python. I think these are related. I wonder why installed Openoffice package on Ubuntu doesn't include python. I said that already in my initial reply: Because Ubuntus OOo puts uno.py etc. in the system python path (and uses system python) Learn to read. And do that. Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] no module named uno error
Hi, On Mon, Sep 06, 2010 at 09:59:36AM +0200, Stephan Bergmann wrote: On 09/06/10 01:01, Rene Engelhard wrote: But a vanilla OpenOffice.org install doesn't do that. And the only way (unless you put all the .pys and the needed .so files into pythons site-packages) way to do that is to include OpenOffice.orgs internal python copy. ...and actually use it, i.e., call .../openoffice.org3/program/python from the OOo installation instead of system python (in case this was not obvious). Actually that was what I meant, just bogusly wrote included than run ;) Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] no module named uno error
On Mon, Sep 06, 2010 at 11:29:29AM -0700, Soohong Min wrote: I tried to do it --with-system-python. But it is not included python in program directory. Of course. --with-system-python is the switch to NOT include it. So I tried to build python separately and I got this message. Building module python Entering /ooo-build-3.2.1.4/build/ooo320-m19/python/ An already available installation of python should exist on your system. Therefore the version provided here does not need to be built in addition. rmdir /tmp/CZ5Q6SRH95 Of course. (see above) How can I build python in Openoffice to include internal python? Do a build WITHOUT --with-system-python. (if you really want that...) Note that --with-system-python is default on ooo-build for all Linuxes. Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] no module named uno error
On Sun, Sep 05, 2010 at 03:44:00PM -0700, Soohong Min wrote: import uno, unohelper Import Error: No module named uno. Jup. After install package for Openoffice 3.2.1.4 on Ubuntu, I tried these scripts on Ubuntru, I run it successfully. Because Ubuntus OOo puts uno.py etc. in the system python path (and uses system python) But after building Openoffice, I cannot run it. (Openoffice is running well) But a vanilla OpenOffice.org install doesn't do that. And the only way (unless you put all the .pys and the needed .so files into pythons site-packages) way to do that is to include OpenOffice.orgs internal python copy. Yes, counterintuitive and (IMHO) broken, so distros of course use system-python. Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Porting addon from Linux to Windows
On Thto) u, Sep 02, 2010 at 10:47:13AM +0430, Jan private wrote: Microsoft Visual C. Does this mean that I must use MSVC to compile my addon? That would be a BIG problem since I am using the GiNaC library wh ich does not exist in a MSVC version and does not willingly compile with it. I believe that to stay binary-compatible (and you have to because you would link to OpenOffice.orgs libraries) you won't find a way around this. Using Mingw/gcc would be easy. The mingw32 port is quite fragile, and I don't think it's exactly ABI-compatible... But that better should be on d...@porting mailinglist anyways. (CC#ed, M-F-T set) Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Print format changes for Linux
On Thu, Sep 02, 2010 at 01:31:51PM +0200, Philipp Lohmann wrote: the CUPS print system that is used on Linux and other Unix operating systems is switching its file format from PostScript to PDF. As part of this OpenOffice.org should switch its print output file format to PDF, too. This was implemented in OOo for issue 94173. When is CUPS switching/did it switch? Do I read it right in the issue that CUPS = 1.2 uses PDF, whereas older ones will use PS? This will be OK, as 1.2 is so ancient. But if a newer CUPS switches, do we need some new versioned dependency? instead of PostScript; also you can switch the behavior back to the old one in either the (not well known) spadmin utility or per print job: in ^ ... and not to be used when using CUPS ... Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] patch error
On Mon, Aug 30, 2010 at 04:16:53AM -0700, Soohong Min wrote: Thanks, I tried to do again, but I still have some patch error as follows: [snip] How can I fix this problem? Please let me know. No. Not unless you start reading and following the hints/solutions we give you. (No, you don't currently, obviously - as we see with your attempt to STILL cross-compile OOo) Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Re: building openoffice problem
On Fri, Aug 27, 2010 at 12:53:48PM -0700, Soohong Min wrote: But, I wonder what is difference between /usr/lib/openoffice/basis3.1/program/libavmediali.so and above file. The one is the system-version of it (of OOo 3.1!) and for Intel, and what you compiled is 3.2.1 and arm. Obvious, no? Because this directory, /usr/lib/openoffice/basis3.1/program/, has same file name compared to No, it doesn't. Not if one if Linux/Intel (li) and you compiled a Linux/Arm one (lr) /ooo-build-3.2.1.4/build/ooo320-m19/solver/320/unxlngr.pro/lib But the contents of files are different. Of course. Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Re: building openoffice problem
On Fri, Aug 27, 2010 at 02:43:22PM -0700, Soohong Min wrote: Because this directory, /usr/lib/openoffice/basis3.1/program/, has same file name compared to There is some error. My directory is /usr/oo-311/basis3.1/program/ includes (*lr) files So you already built 3.1.1 for ARM? BTW, I wonder after bin/ooinstall, I see the files (li) is searched in That is weird, yes. You didn't try cross-compiling by chance? Doesn't work. They cannot find *lr files so I modified original (li) into (lr). Thus, solver/320/lib/unxlngr.pro/lib includes li and lr at the same time. That doesn't make sense. It is quite difficult to understand my condition and I don't know how to fix them. Please let me know. You probably should do a clean build on an ARM machine. Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Re: building openoffice problem
On Fri, Aug 27, 2010 at 03:19:32PM -0700, Soohong Min wrote: BTW, I wonder after bin/ooinstall, I see the files (li) is searched in That is weird, yes. Is there way to search *lr instead of lr while processing bin/ooinstall ? You just do a correct build. The OOo install does not have that problem here when building *fully* on a arm machine. As you know, It takes a lot of times to finish compiling Openoffice. Yes, and? That is even more true if you do braindead (sorry) things to trick the buildsystem/file naming conventions. You probably should do a clean build on an ARM machine. How can I do 'clean build' on ARM ? I am crosscompiling Openoffice in qemu. Clean build means no crosscompiling.. If you need to use qemu, use a complete environment inside it. No qemu-arm wrappers etc. Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] external module stax and java baseline
Hi, On Wed, Aug 25, 2010 at 11:46:20AM +0200, Martin Hollmichel wrote: I'm just wondering wether the external module stax is still needed at all anymore, it seems only needed if using JDK 1.5 and lower and not needed if using gcj. My suggestion is to raise Java baseline to 1.6 and remove external module stax, Is gcj capable of doing 1.6? I wouldn't be that sure. But if you meant 1.6 when not using gcj and removing stax, ok :) And yes, we still need gcj for various archs where we don't have an other Java compiler/interpreter (neither Sun nor OpenJDK) Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] change log of jpeg library 8b
On Wed, Aug 25, 2010 at 01:31:54PM +0200, Martin Hollmichel wrote: we're still using version 6b, this is the change log since that version, maybe some expert want to recommend an update ? Actually I'd recommend *against* it. I don't see versioned symbols there, and thus mixing 8b and 6b will be interesting(tm) with random crashes. And note you link against system gtk etc which will most probably use 6b. Debian had that problem when trying to switch, it was rolled back for now. (And the LSB mandates 6b anyways...) Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Re: Genuine OOo in distributions ? (Re: [dev] OOo installation packages for Linux, a few (easy) questions
Hi, On Tue, Aug 24, 2010 at 08:50:13AM +0200, Stephan Bergmann wrote: The static NativeLibraryLoader.loadLibrary(PipeConnection.class.getClassLoader(), jpipe); in class com.sun.star.lib.connections.pip.PipeConnection within jurt.jar effectively depends on finding a jpipe dynamic library (libjpipe.so on Linux) through the Class-Path in the jurt.jar META-INF/MANIFEST.MF, ridl.jar unoloader.jar ../../lib/ ../bin/. Ah, OK. This works in standard OOo installations, where the jpipe dynamic library is found relative to jurt.jar within the URE installation either in ../../lib (Unix) or ../bin (Windows). This fails if you either use a jurt.jar copied elsewhere, or the given OOo installation does not preserve the above requirement. r...@frodo:~$ cd /usr/lib/ure/share/java/ r...@frodo:/usr/lib/ure/share/java$ ls -l ../../lib/libj libjava_unolibjuh.so libjvmfwk.so.3 libjava_uno.so libjuhx.so libjpipe.solibjvmaccessgcc3.so.3 r...@frodo:/usr/lib/ure/share/java$ ls -l ../../lib/libjpipe.so -rw-r--r-- 1 root root 9280 23. Aug 21:19 ../../lib/libjpipe.so But: r...@frodo:/usr/lib/ure/share/java$ ls -l jurt.jar lrwxrwxrwx 1 root root 31 23. Aug 21:58 jurt.jar - ../../../../share/java/jurt.jar So the symlink is the problem? Grmpf. Why can't you for one time just handle symlinks correctly Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Re: Genuine OOo in distributions ? (Re: [dev] OOo installation packages for Linux, a few (easy) questions
Hi, On Tue, Aug 24, 2010 at 08:50:13AM +0200, Stephan Bergmann wrote: This works in standard OOo installations, where the jpipe dynamic library is found relative to jurt.jar within the URE installation either in ../../lib (Unix) or ../bin (Windows). This fails if you either use a jurt.jar copied elsewhere, or the given OOo installation does not preserve the above requirement. rony, OK please try $ cd /usr/lib/ure/share/java $ ln -f ../../../../share/java/openoffice/jurt.jar jurt.jar (my paste in the last mail is from 3.3 packages where I cleaned up the paths a bit and got rid of openoffice/ as mandated by the Debian Java policy) Looks I need to use a hardlink here in the package... :( (But the underlying problem still should be fixed. Symlinks should work..) Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Re: Genuine OOo in distributions ? (Re: [dev] OOo installation packages for Linux, a few (easy) questions
On Tue, Aug 24, 2010 at 10:26:58AM +0200, Rene Engelhard wrote: Hi, On Tue, Aug 24, 2010 at 08:50:13AM +0200, Stephan Bergmann wrote: The static NativeLibraryLoader.loadLibrary(PipeConnection.class.getClassLoader(), jpipe); in class com.sun.star.lib.connections.pip.PipeConnection within jurt.jar effectively depends on finding a jpipe dynamic library (libjpipe.so on Linux) through the Class-Path in the jurt.jar META-INF/MANIFEST.MF, ridl.jar unoloader.jar ../../lib/ ../bin/. Ah, OK. [...] So the symlink is the problem? Grmpf. Why can't you for one time just handle symlinks correctly Hmm. Or I just fix the Class-Path: there. Stephan, can I just put a /usr/lib/ure/lib there (in addition to ../../lib/) or will there be other stuff still not working then? Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Re: Genuine OOo in distributions ? (Re: [dev] OOo installation packages for Linux, a few (easy) questions
On Mon, Aug 23, 2010 at 08:53:26PM +0200, rony wrote: So there is at least a problem in the Ubuntu distribution with the setup somewhere. As we already found out in this thread. It's a problem in Debian and Ubuntu because we move the Java libs to a standardized path with symlinks and OOo cannot handle the symlink... Now I am waiting for Stephan to tell me whether my proposed solution works. Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Re: Genuine OOo in distributions ? (Re: [dev] OOo installation packages for Linux, a few (easy) questions
Hi, On Tue, Aug 24, 2010 at 12:02:55PM +0200, Rene Engelhard wrote: On Mon, Aug 23, 2010 at 08:53:26PM +0200, rony wrote: So there is at least a problem in the Ubuntu distribution with the setup somewhere. As we already found out in this thread. It's a problem in Debian and Ubuntu because we move the Java libs to a standardized path with symlinks and OOo cannot handle the symlink... Now I am waiting for Stephan to tell me whether my proposed solution works. Fixed (for 1:3.2.1-7) with adding /usr/lib/ure/lib to Class-Path: + # fix up Class-Path of jurt.jar to be able to find libjpipe.so. + # See http://www.openoffice.org/servlets/ReadMsg?list=devmsgNo=27765 + t=`mktemp -q -d`; \ + mkdir -p $$t cd $$t \ + unzip $(CURDIR)/debian/ure/usr/share/java/$(OODIRNAME)/jurt.jar \ + META-INF/MANIFEST.MF \ + perl -pi -e 's#../../lib/#../../lib/ /$(OOUREDIR)/lib/#g' \ + META-INF/MANIFEST.MF \ + zip -r $(CURDIR)/debian/ure/usr/share/java/$(OODIRNAME)/jurt.jar \ + META-INF/MANIFEST.MF \ + cd $(CURDIR) rm -rf $$t No idea whether I'll still be allowed to have the fix in squeezes 3.2.1 (we're in freeze in Debian) but already merged that into 3.3, too, so it surely will be fixed there. Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Genuine OOo in distributions ? (Re: [dev] OOo installation packages for Linux, a few (easy) questions
Hi, On Sat, Aug 21, 2010 at 11:57:26PM +0200, rony wrote: [...] Uninstalling the Ubuntu OOo and instead installing the genuine OOo, downloaed from http://OpenOffice.org/download, installing it and running the very same program works without an error! We (at least Debian, can't speak of Ubuntu) don't patch anything in that area, and we use OpenJDK (which should be sufficiently compatible and doesn't cripple stuff for building). Using Sun JDK is no option, no. So sorry, no idea. What I *can* say, though is that not veretyhing exposed on distros is a bug *caused* by those distros. We had many bugs in the last years where it just occurred with some circumstances, other compilers, ... In the case it matters, here is the CLASSPATH setting for the Ubuntu OOo: /opt/BSF4ooRexx/bsf-v400-20090910.jar:/opt/BSF4ooRexx/bsf-rexx-engine.jar:.::/usr/lib/openoffice/program/../basis-link/ure-link/share/java/ridl.jar:/usr/lib/openoffice/program/../basis-link/ure-link/share/java/jurt.jar:/usr/lib/openoffice/program/../basis-link/ure-link/share/java/juh.jar:/usr/lib/openoffice/program/../basis-link/program/classes/unoil.jar:/usr/lib/openoffice/progra Looks OK. Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Re: Genuine OOo in distributions ? (Re: [dev] OOo installation packages for Linux, a few (easy) questions
On Mon, Aug 23, 2010 at 11:42:22AM +0200, Michael Stahl wrote: i think i remember this error... it is caused by not finding some URE dynamic libraries, like libjpipe.so. the Java UNO bridge apparently uses native code via JNI for some things. Then that is a bug in the bridge or the extension itself IMHO ... on a Ubuntu box here the libraries seems to be in /usr/lib/ure/lib/libjpipe.so so try adding /usr/lib/ure/lib/ to CLASSPATH, see if that helps. ... because ure-link is exactly what points to that /usr/lib/ure thing. Anything which assumes that the ure is inside the OOo dir is wrong; the only valid assumption is that *ure-link* is. That's how the three-layewr OOo interface was defined. Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Re: Genuine OOo in distributions ? (Re: [dev] OOo installation packages for Linux, a few (easy) questions
Michael Stahl wrote. so try adding /usr/lib/ure/lib/ to CLASSPATH, see if that helps. Or LD_LIBRARY_PATH (no idea whether that helps for JNI Linkage but it can be tried, too) Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Instance of OOo running after unopkg ?
Hi, On Mon, Aug 23, 2010 at 04:01:45PM +0200, Rony G. Flatscher wrote: for installing an extension (under Ubuntu in this particular case) in an installation script the command-line tool unopkg is used. When starting OOo thereafter a warning comes up that an instance of OOo would be running, and if one would like to proceed. Is this an expected behaviour? Nope. I don't see that in all the packages here which use unopkg in their preinst/postinst/prerm. Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Re: Genuine OOo in distributions ? (Re: [dev] OOo installation packages for Linux, a few (easy) questions
On Mon, Aug 23, 2010 at 03:29:47PM +0200, rony wrote: (Or do I have to go through Synaptic manager and check all sort of Eww, synaptic. modules (with the risk that I am overlooking an important one, given that the OOo related modules seem to be quite dispersed.) apt-get install openoffice.org. Which is a dummy package which is supposed (unless Ubuntu patched somthing) there install all modules + Java suppoer (including officebean), but of course without SDK) (or select openoffice.org in synaptic). Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Genuine OOo in distributions ? (Re: [dev] OOo installation packages for Linux, a few (easy) questions
Hi, On Mon, Aug 16, 2010 at 06:49:24PM +0200, rony wrote: (Due to a package that excercises the Java-UNO-bridge I have stumbled over Ubuntu's distro which seems to be broken in that area and read Where exactly? (Ubuntu borrows my packages and breaks them at times, but what you experience could also be bug in Debian, so...) about other distributions that at least used to cripple OOo by removing some if not all of the genuine Java support). used to is right, I think nowadays all ship it compiled with OpenJDK at least on mainstream architectures. (Or use Sun JDK, which truly free distros of course cannot use). Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Build order
Hi, On Wed, Jul 28, 2010 at 07:29:50PM -0700, Gregoire Gentil wrote: Which script are you talking about? I can't find any script that answers my question. Thanks in advance for any tip, whateverEnvSet.sh Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Re: Building ooo
On Thu, Jul 29, 2010 at 07:50:45PM +0200, Michael Stahl wrote: On 29/07/2010 17:46, Rick Kautick wrote: Hi all, I'm not sure if I have come to the right place, but here goes nothing. I recently decided that I wanted to get involved with open source development, and thought that fixing some small bugs or creating basic enhancements to oo would be a good place to start. So, naturally, the first thing I did was download the source from http://openoffice.mirrorbrain.org/files/stable/3.2.1/OOo_3.2.1_src_core.tar.bz2. So far so good. Not really. dmake: makefile.mk: line 61: Error: -- Include file /home/zeppelin/Desktop/OOO320_m19/solver/320/unxlngi6.pro/inc/external/neon/version.mk, not found ERROR: Error 65280 occurred while making /home/zeppelin/Desktop/OOO320_m19/ucb/source/ucp/webdav rmdir /tmp/c8b1jjIqx5 As Michael says, this file is delivered from the neon module. Which you don't have. Because you didn'tunpack the _system tarball. OOo includes (almost) all the libs it needs in its code :-(. Either you want _system unpacked or you want to use --with-system-$stuff for some stuff. Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Build error in helpcontent2 (Lucene)
Hi, On Wed, Jun 23, 2010 at 10:54:53PM +0200, Pavel Laštovička wrote: citing from issue #103354: During the massive parallel build of helpcontent2 the lucene indexer generates index files which are renamed. This behavior breaks the MSP pack process, were no files have to be removed. So renamed index files (cfs) are really a problem only on Sun builds? And Yes. what is the MSP pack process? :-) Microsoft Patch (has to google myself when I read that in the issue). As no one else is building them it's basically a Sun-only problem (but that's why my initial patch made those checks on for Windows, but you can do that yourself if you wished by setting the new variable anyways). Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Re: FASTBOOL macro vs bool - decrease memory usage
Hi,
On Thu, Jun 24, 2010 at 02:15:29PM +0200, Michael Stahl wrote:
isn't bool ususally (or at least sometimes) 4 bytes in size?
$ cat test.cxx
#include stdio.h
int main() {
printf(%d\n, sizeof(bool));
}
$ g++ -o lala ./test.cxx
$ ./lala
1
Grüße/Regards,
René
-
To unsubscribe, e-mail: dev-unsubscr...@openoffice.org
For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Re: Build Open office 3.20 on window XP
Hi, On Mon, Jun 21, 2010 at 11:47:07AM +0200, Michael Stahl wrote: On 20/06/2010 14:25, Lê Việt Quang wrote: Dear sir , I try to build OOO320_m12 on window XP , but it fail , and take this error , what is it ? any reason why you want to build such an outdated milestone? Because he wants to build 3.2.0? Because he has that as a Base or needs to provide fixes to it? iirc latest OOO320 is at m19... Which is a different release to 3.2.0. And given the environemnt, a upgrade to 3.2.1 isn't a drop-in, build due to various changes in that micro release. Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Build error in helpcontent2 (Lucene)
On Mon, Jun 21, 2010 at 06:13:15PM +0200, Pavel Lastovicka wrote: when doing a parallel build of 3.2.1 on Cygwin I hit the following error: Checking cfs file _0.cfs: Not found cfs file check failed, terminating... See http://www.openoffice.org/issues/show_bug.cgi?id=112339 for a similar (NOT the same issue). Is fixed in 3.3 snapshots by making that check Sun-only. See http://hg.services.openoffice.org/DEV300/rev/f505f7499559 Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Password protect extension
Hi, On Tue, Jun 08, 2010 at 02:01:57PM +0200, Jean-Baptiste Faure wrote: Le 08.06.2010 11:47, Kai Sommerfeld a écrit : Hi, On 08.06.10 10:37, Christian Heise wrote: Hi everybody, Is it possible to protect an extension in a way that disabling or uninstallation have to be confirmed by entering a password? No, that is not possible. - Kai. But if you install this extension in shared mode, you have to launch OpenOffice.org as admin to disable or uninstall this extension. And on all serious OS you have to enter a password to open an admin session or launch a software as admin. No ? Correct. (Actually I had a similar reply in my postponed but refrained from sending it as that's so obvious that the OP probably is not OK with that possibility and just wants to restrict usage of the extension itself) Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Re: ant error while compiling on win7
Hi, On Thu, Jun 03, 2010 at 12:40:59PM +0200, Michael Stahl wrote: ant -version Unable to locate tools.jar. Expected to find it in /usr/lib/jvm/java-6-openjdk/lib/tools.jar Apache Ant version 1.7.1 compiled on March 5 2010 JAVA_HOME='/usr/lib/jvm/java-6-sun' ant -version Apache Ant version 1.7.1 compiled on March 5 2010 yeah. so if ant is somehow unhappy it will print extra junk on stdout. the test in configure.in does not expect that. apparently JAVA_HOME is set up before the ant test in configure.in. And so it should work, as we see above.. But we of course can make it more safe... Rene, any idea how to prevent this? maybe a grep for version in the check would help? Yeah, probably. And it also would make sense to move the whether ant works before that check, as otherwise the check here fails with the same weird symptom if ant itself fails (like with a bus error or whatever) Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Re: ant error while compiling on win7
Hi, On Thu, Jun 03, 2010 at 09:14:32PM +0800, xiaofen...@iscas.ac.cn wrote: Are you replying to my question? No, I replied to Michael. It seems that what you quoted isn't from mine. Please read: On Thu, Jun 03, 2010 at 12:40:59PM +0200, Michael Stahl wrote: Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] subsequenttests
On Mon, May 31, 2010 at 01:25:26PM +0200, Stephan Bergmann wrote: On 05/31/10 11:55, Bernd Eilers wrote: what X Server do the subsequenttests use by the way xvfbd or xvnc or something else? subsequenttests and below sets up nothing. The started soffice instances simply use whatever DISPLAY points to. svp not possible? The smoketest at least works with it. Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] ant error while compiling on win7
On Mon, May 31, 2010 at 09:17:24PM +0800, xiaofen...@iscas.ac.cn wrote: The error is: ... checking for jakarta-ant... no checking for ant... /cygdrive/c/ant/bin/ant checking whether ant is = 1.6.0... ./configure: line 28460: test: tools: integer expression expected Are you sure your ant executed properly from configure? What does config.log say? (And yes, the error probably should be more meaningful there, but...) Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] subsequenttests
On Mon, May 31, 2010 at 03:47:00PM +0200, Stephan Bergmann wrote: svp not possible? The smoketest at least works with it. No idea what you mean. SAL_USE_VCLPLUGIN=svp aka. headless. As said, works for me for the smoketest in 3.2.x. Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] CMake module for finding OpenOffice.org SDK installation
On Tue, May 25, 2010 at 01:59:56PM +0400, Konstantin Tokarev wrote: File is attached. License is BSD. Any fixes/addition are welcomed. No. ;) (And you better file an issue with the patch anyway) Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: Re: [dev] CMake module for finding OpenOffice.org SDK installation
[ please follow Netiquette and wrote proper lines with proper linebreaks ] Hi, On Tue, May 25, 2010 at 05:20:49PM +0400, Konstantin Tokarev wrote: Actually, no, I disagree. As it's just a data file, we can just include it in the SDK. Actually, this script is not OOo side integration. Wrong. It is. It is a file for loooking where OOos SDK is. It's intended to be included into projects using CMake, it will not work from from SDK location (as, for example, pkg-config files work). Err., last I looked cmake had a dir where it looked for .cmake files: r...@frodo:/usr/share/cmake-2.8$ ls -la insgesamt 80 drwxr-xr-x 5 root root 4096 20. Nov 2009 . drwxr-xr-x 416 root root 12288 24. Mai 23:52 .. drwxr-xr-x 2 root root 4096 17. Mai 12:59 include drwxr-xr-x 6 root root 53248 17. Mai 12:59 Modules drwxr-xr-x 2 root root 4096 17. Mai 12:59 Templates What about this? Alternatively, it could be submitted to CMake, placed in its directory it will be available without copying to project. I don't think that's a good idea. Such stuff belongs to the project wheere it#s supposed to check for (so it can be adapted there for changes) Like stuff for the 3 SDKs won't work with = 3 SDKs. Of course, I can provide an example of usage, but if you want to include CMake integration into SDK, it will require a bit more effort. A bit, not much. (And yeah, I don't believe that without tight system integration it will work, but if you read closer, I didn't claim that anyways) Also, I'd like to add that this file could be potentially used not only for extensions, but for any project which needs OpenOffice.org files for any purpose (but it may require some improvements in it). Exactly. One more point why it should be in SDK (and be it for reference or for copying it in your tree - like the infamous unowinreg.dll) Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] New build system
Hi, On Tue, May 11, 2010 at 12:07:04PM +0400, Konstantin Tokarev wrote: Also, with autotools you'll depend on their version. With CMake, you'll set version number, and newer CMake versions will run in compatibility mode with that version. You're off-topic, noone here said anything about full autotoolizing. Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] New build system
On Tue, May 11, 2010 at 11:54:24AM +0400, Konstantin Tokarev wrote: * much faster installation - up-to-date binaries are not updated As we (unfortunately) don't have proper make install (but run make_installer anyway, which has already some sort of caching of the installset created) this point is moot. Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Re: [releases] recommended stlport settings?
Hi, On Mon, Apr 26, 2010 at 09:43:28AM +0100, Caolán McNamara wrote: On Sun, 2010-04-25 at 19:44 +0200, Rene Engelhard wrote: We can only improve things here when we eventually drop the STLport-requirement (and become URE-incompatible on the affected platforms). (Note that I didn't say that) If we continue to build and package into the install sets stlport on Linux x86, but not actually build OOo itself against it, its quite likely that stuff will work out ok (i.e. legacy binary x86 extensions that link against stlport will continue to work) as there's no explicit use of stl types in the public ure interfaces. Not 100% certain about Yeah. Or that. (As AFAIR SuSE does already) Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
[dev] Re: [releases] recommended stlport settings?
[ releases is not the correct list for that, X-posting to dev@ ] Hi, On Sun, Apr 25, 2010 at 04:24:51PM +0200, Andreas Radke wrote: So far i686 needed --with-stlport to be able to use 3rd party extensions. DEV300_m77 introduced cppunit requirement and now i686 configure fails: Not really. Was there before too, just that there always the internal hacked version was used. Now sb118 introduced the possinbility to use system-cppunit. (And note the default is still internal cppunit) checking which cppunit to use... external OK, so you use --with-system-cppunit. (or --with-system-libs which of course implies --with-system-cppunit).. checking for CPPUNIT... yes checking STL compatibility... configure: error: to use system cppunit you need to use --without-stlport What the recommended way now? --without-stlport now for all architectures? I build x86_64 and i686 for ArchLinux. You can use internal cppunit. Either only on i386 or everywhere. Though I don't like this either. See Issue 109791, we can't apply the same workarounds here like for mysqlc and graphite. To quote sb: --- snip --- Wrt STLport: no way, if you really have a STL conflict (like we ha(d/ve) with graphite and mysqlc) apply a workaround like the one caolan did for graphite already (see issue 106157); it is bad to use system-cppunit on everything except i386. Better all or none You unfortunately cannot work around that here. The interface of CppUnit is full of std-types, so the OOo test code calling CppUnit and CppUnit itself have to use the same STL implementation. However, the OOo test code also shall access OOo code that has std-types in its respective interface, so those two need to use the same STL implementation, too. We can only improve things here when we eventually drop the STLport-requirement (and become URE-incompatible on the affected platforms). --- snip --- An other reason why not dropping STLport when we could have (last major release AKA 3.0.0) was a seriously bad thing... Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Re: Re: OpenOffice.org Product Development
Hi, On Wed, Mar 31, 2010 at 11:36:16AM +0200, Juergen Schmidt wrote: mmh, i have still one problem. One motivation for me to go into the CC was and still is that i will try to work for a stronger OpenOffice.org community and against derivated work. And in consequence of this i see a stronger community means actively involving hughe parts of the community. which you don't. there's still many sun-internal decision just posed to the community as a fact everyxone has to live with. There's still bugs not fixed or cared about because it doesn't affect *you* (but only people outside) Or patches lingering around for no reason because no one cares about. a conflict with somebody in the CC who is nowadays the driving force behind a spin-off (wording of go-oo, i call it fork others derivated work). It might be not so obvious for many community members what go-oo really is. And one reason for this is that our project rules are not accepted as they are. Or only partly. That's not true. At least not for all cases where stuff ends up in go-oo. be a product development representative for OOo he has to accept the OOo project rules. says the one who for a long time didn't accept those either. Should I remember you about non-free stuff of yours in the code despite this being a open source (and I do take the official open source definition) project? Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Re: Re: OpenOffice.org Product Development
Hi, On Wed, Mar 31, 2010 at 01:32:56PM +0200, Cor Nouws wrote: mmh, i have still one problem. One motivation for me to go into the CC was and still is that i will try to work for a stronger OpenOffice.org community and against derivated work. And in consequence of this i see a stronger community means actively involving hughe parts of the community. I think this is an interesting point. Getting people that work on a 'derived product' more involved is IMO positive. On the other hand, there is the question of credibility. The point is: with better behaving/cooperation from you many patches would not need to be only in derived products but can also be in vanilla... Or patches lingering around for no reason because no one cares about. [IMO OT, but I doubt that there are no aspects as time and complexity involved. Anyway, I ám interested in this. So expect continued contact from me about this.] Please do :) a conflict with somebody in the CC who is nowadays the driving force behind a spin-off (wording of go-oo, i call it fork others derivated work). It might be not so obvious for many community members what go-oo really is. And one reason for this is that our project rules are not accepted as they are. Or only partly. That's not true. At least not for all cases where stuff ends up in go-oo. partly = at least not for all cases True. But jsc suggests that's the main cause of patches not being upstream. Whereas many bugfixes are not upstream because the get not integrated - even when the rules are OK to follow. be a product development representative for OOo he has to accept the OOo project rules. says the one who for a long time didn't accept those either. Should I remember you about non-free stuff of yours in the code despite this being a open source (and I do take the official open source definition) project? When you write 'for a long time didn't', the problem seems solved? After years of fighting, yes (And it got fixes by just removing it from the code and it's just only in wiki now, still non-free) (There's still some non-free or legally questionable or even non-legally distributable stuff in the source code, but not in jscs area) But also this is not relevant in considering the situation we are in now, IMO. No, it is relevant. How is it not relevant when people accusing other people of not caring about rules does not do it himself? Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Re: Re: OpenOffice.org Product Development
Hi, On Wed, Mar 31, 2010 at 01:30:12PM +0200, bjoern michaelsen - Sun Microsystems - Hamburg Germany wrote: On Wed, 31 Mar 2010 12:52:18 +0200 Rene Engelhard r...@openoffice.org wrote: a stronger community means actively involving hughe parts of the community. which you don't. there's still many sun-internal decision just posed to the community as a fact everyxone has to live with. Who is you here? Sun contributors are not the Borg and in a community Sun. they should be treated according to their individual merits and vices and not be hold in kin liability. True. As is the other way round. There's still bugs not fixed or cared about because it doesn't affect *you* (but only people outside) So you want to have control over the way that Sun manages its resources? Thats not how it works in _any_ OSS project. If you would stomp into #gentoo, #debian or #ubuntu and request your personal itch Neither of those are development channels :) to be scratched you would be laughed out the door. Well, on ubuntu you might be offered a support contract, but that is different from what you are suggesting here. Ubuntu has to die. :) Or patches lingering around for no reason because no one cares about. That would be a different issue. But there are also non-Sun domain developers, so this is not specific to Sun contributors -- its plain old lack of ressources. Wrong. How is answering a simple queston so that someone outside Sun can finally do th ecws and get it integrated lack of resources? How is patch -pX lack of resources? How is telling people how they need to change stuff to get a patch (which works for everyone else except Sun and is needed there) finalized to get integrated a resource problem? (The fix I have in mind is a triviality for someone who knows how to add/hide stuff in the Options dialogs) Finally, I do not like statements implying You are at Sun and thus not part of the community.. Actually, that is quite an unfair discrimination. True. But for some people at Sun this is the case. Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Re: Re: OpenOffice.org Product Development
On Wed, Mar 31, 2010 at 01:30:12PM +0200, bjoern michaelsen - Sun Microsystems - Hamburg Germany wrote: On Wed, 31 Mar 2010 12:52:18 +0200 Rene Engelhard r...@openoffice.org wrote: a stronger community means actively involving hughe parts of the community. which you don't. there's still many sun-internal decision just posed to the community as a fact everyxone has to live with. Who is you here? Sun contributors are not the Borg and in a community they should be treated according to their individual merits and vices and not be hold in kin liability. Oh, what I forgot in my last reply. So we should accept decisions imposed on anyone which even violate your so well-vbalued project rules? Like features in a bugfix release? Happened in OOO320_m14. Or restructuing without even discussing with people in the outside. Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Re: Re: OpenOffice.org Product Development
Hi, On Wed, Mar 31, 2010 at 02:25:12PM +0200, bjoern michaelsen - Sun Microsystems - Hamburg Germany wrote: On Wed, 31 Mar 2010 13:52:11 +0200 Rene Engelhard r...@openoffice.org wrote: How is patch -pX lack of resources? How is telling people how they need to change stuff to get a patch (which works for everyone else except Sun and is needed there) finalized to get integrated a resource problem? (The fix I have in mind is a triviality for someone who knows how to add/hide stuff in the Options dialogs) First the issue must be understood(*), then the patch need to be The issue was understood. reviewed, [...] The patch was reviewed, that's were they asked me to change it. Just that noone wants to tell me (who doesn't know the code nor wants to go such deep into the code) *how* to change. Instead, they just block. Or for the other example issue in mind: it's clearly understand there, too. Patch is there and builds (maybe not without warnings on !Linux but that is not jls problem but the cws owners). Still no answer thus effectively delaying it. it must build warning-free on all platforms against baseline and be QA'ed. True. But that is a different step then from applying them. Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] problem with compiling ooo-build
Hi, On Sun, Mar 28, 2010 at 06:48:26PM +0200, Heiko L. wrote: |diff --git a/sc/inc/sc.hrc b/sc/inc/sc.hrc - pwd != patchdir - use -p1 Nonsense. If you looked at the actual patch: --- sc/inc/sc.hrc +++ sc/inc/sc.hrc Go figure. (The sbuild tree of theones who tries building is just fucked up and doesn't have sc (any probably many more) for whatever reason). And that patch just is the first one ever applied. Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] new dependency to ant version 1.7.1
On Wed, Feb 10, 2010 at 11:29:15AM +0100, Juergen Schmidt wrote: i would like to increase the ant dependency to version 1.7.1. The reason is that newer NetBeans versions create build.xml scripts that make use of features from ant 1.7.1 Then they should be fixed not to use them. It should be possible to build such extensions (that are generated with NB) in the OO build env. The idea behind this is to make it possible for developers to easy bring their extensions in the OO build env and take advantage of the i18n process for example. It should be possible to add usable build.xmls, yes. Is there anybody who see problems with this increased dependency to ant 1.7.1? If yes please let me know. Same story as as it came up the last time. There's distros out there not shipping 1.7.1 yet. You'd break the build on those. See e.g. Debian: $ rmadison ant [...] ant |1.7.0-6 |stable | source, all ant |1.7.1-4 | testing | source, all ant |1.7.1-4 | unstable | source, all The other question is where i have to make changes? configure.cmd configure.in No. *If* you want 1.7.1, make it a check for those extensions who need it. It makes no sense to break the build for the others which have a sufficiently uptodate ant otherwise) currently we have: ant_minver=1.6.0 # update for more extensions... if test $ENABLE_MEDIAWIKI = YES; then ant_minver=1.7.0 fi configure No change needed, except running autoconf. Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Using JUnit 4 in OOo build environment
Hi, On Wed, Feb 10, 2010 at 02:57:24PM +0100, Stephan Bergmann wrote: 1 Include JUnit 4 as an external module in the OOo code base, allowing some configure --with-system-junit switch to override it. We could include either Good. 1.1 http://sourceforge.net/projects/junit/files/junit/4.8.1/junit-4.8.1.jar/download which, due to being platform-independent, can directly be used without any compilation (and appears to run fine on our Java 5 baseline), or No way. No binary only stuff from somewhere in the tree. Should you do that, be sure I'll immediately file a P1. 1.2 http://sourceforge.net/projects/junit/files/junit/4.8.1/junit4.8.1.zip/download and build it. That's the way to go if you want to include it in the build. 2 Require JUnit 4 to be installed as a prerequisite, checked by configure. We could either 2.1 try to find it on CLASSPATH (which would imply that set_soenv.in must no longer resets the CLASSPATH; which is probably dubious, anyway---as dubious as the additional XCLASSPATH), or Which probably will break a lost of system* stuff, so I'd be wary.. Of course you shoudl try your best (like with the SRB libs, the tomcat libs or the apache-commons libs) to find out where a junit4 is. [ /usr/share/java/junit4.jar is a safe bet. ] 2.2 require that an appropriate configure --with-junit=... switch is always passed (which would be optionally available in 2.1, too, to explicitly specify the JUnit 4 to be used). Only when the automatic detection (see above) fails. Grüße/Regards, René - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] OO SDK GLIBCXX_3.4.9 not found in Linux Ubuntu
Hi, On Mon, Jan 11, 2010 at 10:33:08AM +0100, Stephan Bergmann wrote: r...@catullus:~# /opt/openoffice.org/basis3.1/sdk/bin/uno-skeletonmaker /opt/openoffice.org/basis3.1/sdk/bin/uno-skeletonmaker: /opt/openoffice.org/basis3.1/sdk/bin/../../ure-link/lib/libstdc++.so.6: ^^^ version `GLIBCXX_3.4.9' not found (required by /opt/openoffice.org/basis3.1/sdk/bin/uno-skeletonmaker) r...@catullus:~# Are the packages of OOo itself (incl. file /opt/openoffice.org/ure/lib/libstdc++.so.6) and the OOo SDK package (/opt/openoffice.org/basis3.1/sdk) coming from the same source (either both from your Ubuntu installation, or both from the OOo download site)? (Obviously) both are from the download site? Ubuntu doesn't have OOo installed in /opt, and neither does it ship a copy of libstdc++6. Grüße/Regards, Rene - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Re: Build error
Hi, On Tue, Dec 22, 2009 at 02:26:08PM +0100, Björn Michaelsen wrote: so simple now with mercurial), but there is a also testautomation tarball, maybe you need to add that one? If that is the case, we should indeed update the docs to reflect that. No, we should fix the build to *NOT* require it if you don't need it. Grüße/Regards, Rene - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Re: Build error
Hi, On Tue, Dec 22, 2009 at 03:08:09PM +0100, Björn Michaelsen wrote: Am Tue, 22 Dec 2009 14:39:51 +0100 schrieb Rene Engelhard r...@openoffice.org: No, we should fix the build to *NOT* require it if you don't need it. Actually, there is a lot of work going on in that area. see: http://wiki.services.openoffice.org/wiki/Build_Environment_Effort/Split_Build That's not remotely relevant for this discussion. Even now we do not (unless something changes recently which should be fixed) testautomatiopn for a simple build without running tests. Grüße/Regards, Rene - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] disabling module xmlsecurity when not building module nss
On Thu, Nov 26, 2009 at 04:17:05PM +0100, Joachim Lingner - Sun Germany Software Engineer - ham02 - Hamburg wrote: As a consequence I will disable building of xmlsecurity when building of the nss module is disabled. Please do not forget system-nss (aka system-mozilla) and that this of course should keep xmlsecurity enabled (even though the nss module is not built). Just in case Grüße/Regards, Rene - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Library requirements (freetype2 = 2.0 ) not met . . .
Hi, On Tue, Nov 17, 2009 at 11:32:11AM -0500, Albretch Mueller wrote: Thank you and I think error messages should be a bit more explicit. They are explicit enough. It's common that you need -dev for the libs. r...@knoppix:/media/hdb2/inst/sw/OO/source/ooxpkgs# dpkg -l | grep curl ii curl 7.15.5-1etch3 Get a file from an HTTP, HTTPS, FTP or GOPHE ii libcurl3 7.15.5-1 Multi-protocol file transfer library ii libcurl3-gnutls 7.15.5-1 Multi-protocol file transfer library So no libcurl3-dev? (or some variant of it)? qed. What more curl(ing) stuff do I need? Obviusly the curl development stuff, like we already told you for freetype. Grüße/Regards, Rene - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Build with Ubuntu 9.10
Hi, On Thu, Nov 19, 2009 at 11:24:55AM +0800, Zhu Lihua wrote: In file mkdepend/def.h:157 char *getline(); But in file /usr/include/stdio.h:651 extern _IO_ssize_t getline (char **__restrict __lineptr, size_t *__restrict __n, FILE *__restrict __stream) __wur; Why this happened? Becausse there is a copnflict with that getline and glibcs. As the error says: [...] def.h:157: error: conflicting types for 'getline' /usr/include/stdio.h:651: note: previous declaration of 'getline' was here [...] And why this not happend before? because you eithere upgraded your system or reinstalled a new version and it gets a new glibc. OOo 3.1.1 does not build with it because of the above. Workaround: http://cgit.freedesktop.org/ooo-build/ooo-build/tree/patches/dev300/soltools-mkdepend-getline.diff?h=ooo-build-3-1-1 (afaik upstream too already, but of course not for 3.1.1 because it was dfound after the release...) Grüße/Regards, Rene - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Recompiling OOo 3.1.1 Ubuntu (karmic) package source
Hi, On Tue, Nov 10, 2009 at 05:42:45PM +0800, Spider Larkin wrote: If I try to skip the cleaning and unpacking: $ rm debian/stampdir/build That just removes the debian stampfile. ooo-build has one, too. $ debuild -uc -us -b -Tbuild ...debuild sets up its environment... ...cds into ooo-build... ...calls /usr/bin/make... ...which checks every subdirectory of ooo-build except ooo-build/build, and decides that nothing needs to be done! How do I get make to notice that ww8atr.cxx has changed and recompile the parts that need recompiling? All of the documentation I have seen only tells me how to build from clean sources. It seems to me that it didn't even try to recompile. rm ooo-build/stamp/build (or even rm ooo-build/config.status) should do it. I can see that making a patch, tossing it into the patches directory and rebuilding from scratch will get my code into the build, but at the cost of 1.5 hours every time I change a line of code. You even cna put it there and debuild -b -nc after rm debian/stampdir/build. ooo-build should figure out that something in apply changed... Grüße/Regards, Rene - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Planned outage: EIS not available on 2009-10-23 07:30h - 08:00h MEST
Hi, On Tue, Oct 20, 2009 at 04:18:29PM +0200, Frank Mau wrote: we scheduled an outage for half an hour for the external EIS Friday, 23th October, 07:30 a.m. - 08:00 a.m. MEST (means UTC +2) No updates will be recognized during the outage so please stay tuned for further updates, thank you. And why are you hiding this important (to non-CEST people) message in a unrelated thread by replying to an unrelated mail and thus every serious MUA with threading will sort it below that thread? You should have composed a new mail... Grüße/Regards, Rene - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Where to download libagg source code package?
Hi, On Fri, Oct 09, 2009 at 12:23:14AM -0300, Ariel Constenla-Haile wrote: On Thursday 08 October 2009, 23:30:22, Peng Yu wrote: Hi, I know that openoffice website provide the source of agg. But where I can download its official source code release. Anti-grain geometry has its own download page at its site http://www.antigrain.com/download/index.html Please note you can't use the uptodate agg version for license reasons (well, you can if you want to make your OOo copy GPLed...) Grüße/Regards, Rene - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Dropping tcsh support for Mac (all?) / changing configure's default
Hi, On Fri, Sep 25, 2009 at 01:42:25PM +0200, Christian Lohmaier wrote: because of 105003, own bad experience with tcsh (random build freezes, slow performance), encouraged by the kick4nt cws I'd like to go on and drop tcsh support at least for Mac OS X. dropping tcsh support means: changing configure's default to bash and issue a warning (or error) when tcsh is selected. Yeah. The default login shell for Mac OS X is bash, so it definitely is available and it is proven to work without problems (Pavel builds with bash for years, so do I on all systems) So now to the reason why the d...@ooo list is adressed as well: Is somebody opposing chaning the default to bash for all systems? No. But a suggestion. But what about defaulting to the current login shell? People who might want tcsh already might use it as login shell Grüße/Regards, Rene - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Dropping tcsh support for Mac (all?) / changing configure's default
Hi, On Fri, Sep 25, 2009 at 02:01:34PM +0200, Stephan Bergmann wrote: On 09/25/09 13:49, Rene Engelhard wrote: But what about defaulting to the current login shell? People who might want tcsh already might use it as login shell I would vote for getting rid of tcsh-support in the build environment completely, sooner or later. Making the shell used within the build default to bash is a move in that direction, while making it depend on $SHELL is not. True. But using $SHELL has the benefit that we can still get rid of the configure check at all and if we switched to bash we can just silently do it because no one will expect anything else because he's specifying --with-use-shell= (which IMHO was a bad idea to b egin withg, but at that time we had various shells, including 4nt...) Grüße/Regards, Rene - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Re: [tools-tinderbox] Moving to bost 1.3?
Hi, On Wed, Aug 19, 2009 at 08:36:37PM +0200, Rene Engelhard wrote: command - where's the no net access from here onwards-line drawn? After the download of the sources. When the actual build begins, there is no net access allowed. And what it downloads is just the source package (tar.gz/.diff/.dsc). After that, nothing is allowed. See http://buildd.debian.org/openoffice.org for example. (was unclear in my last message) Grüße/Regards, Rene - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Re: [porting-dev] Re: [tools-dev] Moving to bost 1.3?
Hi, On Wed, Aug 19, 2009 at 08:30:39AM +0200, Frank Schoenheit - Sun Germany - ham02 - Hamburg wrote: [ Seems so, do you now strictly need 1.39? No, I think 1.34 would do, too. As said, a previous incarnation of the CWS compiles fine with 1.34, but again, I didn't do extensive runtime checks with this version. Well, that I understood. Sorry for me not expressing better. The code as-is now will fail to build/work with system-boost 1.39, I assume? Then we probably should adapt the system-boost check in configure; or we #ifdef it as necessary ] Ah, okay, will put this onto my TODO list (unless you volunteer, of course :) ). That was the point of this question to find out what is needed :) Grüße/Regards, Rene - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Moving to bost 1.3?
Hi, On Wed, Aug 19, 2009 at 08:56:23AM +0200, Regina Henschel wrote: The new Math libraries 'Special Functions' and 'Statistical Distribution' are not included in Boost 1.34, but first in Boost 1.35. The accuracy of those functions is ongoing work in Calc and doing it ourself has been a lot of work. Please talk with Eike about the need for that libraries. Though that also would result in linkage against the actual lib while we so far only used boost headers. Nothing bad per se, but it should be kept in mind (and the boost module needs to be changed to actually build boost :-)) Grüße/Regards, Rene - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Re: [porting-dev] Re: [tools-dev] Moving to bost 1.3?
Hi,
On Wed, Aug 19, 2009 at 10:31:59AM +0200, Frank Schoenheit, Sun Microsystems
Germany wrote:
Hi Rene,
The code as-is now will fail to build/work with system-boost 1.39,
I assume?
Most probably. Didn't explicitly try that, though.
OK, just tried, it fails e.g. in connectivity because of the headers...
Why did you change the headers? My boost 1.39 perfectly has a spirit/core.hpp...
#elif defined(__GNUC__) || defined(__HP_aCC) || defined(__SUNPRO_CC) ||
defined(__IBMCPP__)
# warning This header is deprecated. Please use:
boost/spirit/include/classic_core.hpp
ah, I see
So we either need something like this (when did they change that?):
--- configure.in(Revision 275137)
+++ configure.in(Arbeitskopie)
@@ -3936,6 +3936,17 @@
[AC_MSG_ERROR(boost/shared_ptr.hpp not found. install boost)], [])
AC_CHECK_HEADER(boost/spirit/core.hpp, [],
[AC_MSG_ERROR(boost/spirit/core.hpp not found. install boost)], [])
+ AC_MSG_CHECKING([boost version])
+ AC_TRY_RUN([
+#include boost/version.hpp
+
+int main(int argc, char **argv) {
+ if(BOOST_VERSION 103800)
+return 1;
+ else
+return 0;
+}
+ ], [AC_MSG_RESULT(OK)], [AC_MSG_ERROR([not suitable, we need boost =
1.38])])
AC_LANG_POP([C++])
else
AC_MSG_RESULT([internal])
or check for the headers in configure and use #ifdefs. (either -DNEW_HEADERS or
whatever or just #if BOOST_VERSION = 103800). Or we revert that change
until they finally remove the old header :-)
Grüße/Regards,
Rene
-
To unsubscribe, e-mail: dev-unsubscr...@openoffice.org
For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Re: [porting-dev] Re: [tools-dev] Moving to bost 1.3?
Hi, On Wed, Aug 19, 2009 at 03:12:11PM +0200, Frank Schoenheit, Sun Microsystems Germany wrote: So we either need something like this (when did they change that?): [configure patch] I'd prefer that. Making the source code dependent on different boost versions (even if it's only about different headers to include) is too prone to breakage, IMO. The problem is for people who might want to use system-boost for their older distros not having boost = 1.38. But it might be possible to use internal boost for them... BTW, if the problem only is that header (build still running), we probably shouldn't do my original patch but just patch the AC_CHECK_HEADER directly... Like this: --- configure.in(Revision 275137) +++ configure.in(Arbeitskopie) @@ -3934,8 +3934,8 @@ AC_LANG_PUSH([C++]) AC_CHECK_HEADER(boost/shared_ptr.hpp, [], [AC_MSG_ERROR(boost/shared_ptr.hpp not found. install boost)], []) - AC_CHECK_HEADER(boost/spirit/core.hpp, [], - [AC_MSG_ERROR(boost/spirit/core.hpp not found. install boost)], []) + AC_CHECK_HEADER(boost/spirit/include/classic_core.hpp, [], + [AC_MSG_ERROR(boost/spirit/include/classic_core.hpp not found. install boost = 1.38)], []) AC_LANG_POP([C++]) else AC_MSG_RESULT([internal]) Grüße/Regards, Rene - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Re: [tools-tinderbox] Moving to bost 1.3?
Hi, On Wed, Aug 19, 2009 at 04:53:28PM +0200, Jens-Heiner Rechtien wrote: Or invent a nice solution that does auto-downloads, and switch a few other huge external libs to that (like icu). ;) That would be a problem for some builders unless you don't download it when the file is there... For example, for Debian requiring any form of net access on a build is a no-go (and for some libs we have to use the internal versions, and be it sometime, in emergency) And please also think on people bulding somewhere without net access.. Grüße/Regards, Rene - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Re: [tools-tinderbox] Moving to bost 1.3?
Hi, On Wed, Aug 19, 2009 at 08:18:28PM +0200, Frank Schoenheit, Sun Microsystems Germany wrote: For example, for Debian requiring any form of net access on a build is a no-go (and for some libs we have to use the internal versions, and be it sometime, in emergency) What's the definition of build here? I.e., between getting the sources (which most probably requires net access), and issueing the actual build depends. not if you have them locally. The rule also applies for local builds. But yes, the build daemons of course need to get the sources from the mirror they use :) command - where's the no net access from here onwards-line drawn? After the download of the sources. When the actual build begins, there is no net access allowed. Grüße/Regards, Rene - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Re: [tools-tinderbox] Moving to bost 1.3?
Hi, On Wed, Aug 19, 2009 at 09:11:53PM +0200, Frank Schoenheit, Sun Microsystems Germany wrote: After the download of the sources. Then let's define download the pre-requisites as a sub task of download the sources :) Doesn't work. I need to upload a self-contained thing. THAT gets downloaded from incoming.debian.org or a Debian mirror. Nothing externally. Thus I need to include the downloaded stuff *before* the upload in the source package. Grüße/Regards, Rene - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Re: [porting-dev] Re: [tools-dev] Moving to bost 1.3?
Hi, On Wed, Aug 19, 2009 at 08:20:50PM +0200, Frank Schoenheit, Sun Microsystems Germany wrote: BTW, if the problem only is that header (build still running), no, there are much more of those. No. (At least on Linux) cws boost134 build with system-boost *1.34* only needs that header #include fix to work here. (Oh, and btw, in my first patch with the version check you need to adapt the check anyway in the same way, or just remove it because it would be redundant. Otherwise it'll break when the boost people will remove that deprecated header) Grüße/Regards, Rene - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
[dev] Re: [tools-dev] Moving to bost 1.3?
[ dropping tinder...@tools, as they don't use system-boost anyway ] Hi, On Tue, Aug 18, 2009 at 10:56:34PM +0200, Frank Schönheit - Sun Microsystems Germany wrote: We invite everybody porting OOo to another platform to give feedback to this project. As rumor has it, boost 1.39 creates problems when used on some platforms (either at compile- or runtime), so if your platform is know to be one of those, or if you just want to be sure - please give the CWS a try. Did you need code changes for boost 1.39 to work at either build or runtime? [ Seems so, do you now strictly need 1.39? Then we probably should adapt the system-boost check in configure; or we #ifdef it as necessary ] Grüße/Regards, Rene - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] libxml2 dependency on zlib
Hi, On Mon, Aug 10, 2009 at 04:26:44PM +0200, Stephan Bergmann wrote: On 07/20/09 12:46, Rene Engelhard wrote: The easiest fix is just to use system-zlib. At least for UNX. Any (UNIX) system will jave it No, the easiest fix still appears to just drop zlib support from our libxml2. (For example, this also nicely fixes the open issue that the Solaris URE package erroneously lacked a dependency on the corresponding Solaris package providing the system zlib.) And a libxml2 in the ure package linked against libz.so.1 would not cause a dependency on whatever package that is in for you? Sounds broken to me. Otherwise it just should pick it up and you're fine; and again zlib is there on all platforms anyway, and stuff we already use externally in any case are linked against zlib anyway to start with. Regards, Rene - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Mercurial Pilot Feedback, Results
Thorsten Ziehm wrote: Not difficult if you ignore problems ;-( I don't think, that Heiner ignore problem reports. He was very open for comments for this pilot and did a lot to evaluate problems. As you can read he has listed positive and also negative feedback. Also your feedback is listed with the result he had taken from the discussion. No, it's listed with what he told me in the discussion. Which was frankfly speaking not a discussion at all but he just pointing out what he pointed out in this summary. Again, I don't believe the arguments at all a) our mercurial package *has* .sos and is still slow - the native libs argument doesn't hold. He didn't even tell me what I should check when I asked him. At least Debians mecurial package is not at fault. b) he said the low mem/slow cpu/slow(er) hd combination was at fault That does not explain why it's equally slow on a (quite) up-to-date machine. So please be constructive with your feedback and be friendly. I am. Regards, Rene - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Mercurial Pilot Feedback, Results
Jens-Heiner Rechtien wrote: Scalability: - overall perceived good performance, some were even quite enthusiastic about it (SVN users are easy to please ...). - there were three mentions of sub-par performance which all have been investigated shortly: - unexpected slow clone times on a very old machine with slow disks and little memory: this machine was probably simply underpowered for use with Mercurial, which is somewhat memory intensive due to the implementation in python. Also there was a misunderstanding about when hg uses hard links as an optimization. If that was true, please tell me why it also was that slow on my MacBook 2Ghz, 1GB RAM with (quite) fast disk, not an old like my Mac mini G4 with 512M RAM? But I told you that at that time, too - unexpected slow update of the working tree: caused by using the pure python replacements instead of the hg native shared libs. This should be avoided by any project of the size of OOo. You didn't read what I wrote. How please is http://packages.debian.org/lenny/mercurial not using native libs? (See the libc6 dependency and the .sos). Conclusion: === The purpose of the pilot was to find out if there are any important aspects which render Mercurial unusable as SCM for OOo. We found that there are none. This doesn't mean that Mercurial couldn't use some Not difficult if you ignore problems ;-( Regards, Rene - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Mercurial Pilot Feedback, Results
[ forgot something ] Jens-Heiner Rechtien wrote: intensive due to the implementation in python. Also there was a misunderstanding about when hg uses hard links as an optimization. There wasn't actually, I just followed your pilot documentation 1:1. If that sucked you can't blame it on the person using that documentation. Regards, Rene - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Re: [l10n-dev] Localisation moved into own module
Hi, Thorsten Ziehm wrote: Ivo and others are working on it to realize it in the near future ;-) http://eis.services.openoffice.org/EIS2/cws.ShowCWS?Path=DEV300%2Fl10nframework01 How is that helpful? #build the l10n tools cd transex3 build --all deliver cd xmlhelp build --all deliver cd rsc build --all deliver OK. Obvious. # build only l10n targets , skip HID generation cd instsetoo_native build --all L10N_framework=1 NO_HIDS=1 Why -all? Just for completeness and because you need some stuff for packaging? Or do we really need *all* the other modoules? This effectively needing the whole source tree anyway? What we need to have and what was Eikes question aboiut was to just need: cd l10n build --all deliver do some packaging steps if that works out, fine. If not, that issue isn't what we talk about. Regards, Rene - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Re: [l10n-dev] Localisation moved into own module
Hi, Thorsten Ziehm wrote: the goal to to build and deliver language packs without any dependencies. This CWS is the first step. There are still some steps open. But as I heard by Ivo, when this CWs is ready you can you can build the localization without any obj's / libs / whatever. But the Except transex3, rsc etc. but that its unavoidable, I agree :) [1] source code is still needed in this step. So it doesn't help, as you need a full copy of the stuff anyway and thus also could just build it with the rest of the office And our FTP people won't allow a second copy of the OOo source in the archive just to build the translations out of it... So there will be a benefit. ;-) Though only a small one... :) Regards, Rene [1] packagers probably can work around it, though - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] finding binaries et al from OO3.0.1 build
Hi, lux-integ wrote: No EPM: do no packaging at this stage ^^^ Reading helps... ;) rmdir /tmp/7113 ## However I am unable to find the binaries etc to install them. Of course, you disabled the packaging alltogether. (You used --disable-epm I bet) Regards, Rene - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] help with openofice-3.0.1-64-bitCLFS-linux
Hi, lux-integ wrote: Test Complete ^^^ No environment file will be generated ^^^ There were no warnings however there appears to be no Makefile nor bootstrap.sh Of course not. Re-check your configure flags again, and please learn to read. iDo you you have --enable-check-only in the configure line by chance? AC_ARG_ENABLE(check-only, [ --enable-check-only Use this option option if you just want to check your environment. This option stops the generation of an env.set Usage: --enable-check-only=yes Regards, Rene [1] btw, it's not openoffice - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] help with openofice-3.0.1-64-bitCLFS-linux
Hi, T. J. Frazier wrote: We thought it useful to say something like: **--enable-check-only=yes** No environment file will be generated Is it not. It is redundant. People who used --enable-check-only should have read the description it at which time it would have been clear. Regards, Rene - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Please fix these security flaws that could be considered bugs ...
Hi, Nathanael Kennedy wrote: There are serious problems, even at a security level, with OpenOffice... There is no EASY OR STRAIGHTFORWARD WAY to DELETE the entire list of Recent Documents. Documents I no longer care about, I of course delete. DOCUMENTS REMAIN ON THE RECENT DOCUMENTS LIST LONG AFTER THEY'VE BEEN DELETED (FOR MONTHS, SOMETIMES!) This is a serious annoyance could even be considered a serious security flaw on a computer used by several individuals. Nonsense. For this, you have separate accounts for the people and each person has its own local configuration - no problem when sharing the pc. No security thing at all, an annoyance maybe, yes, but if yu can't set up your computer *properly* for multi-user Regards, Rene - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] Please fix these security flaws that could be considered bugs ...
Hi, Regina Henschel wrote: I do not share your view. In my school we have got PCs for pupils, where not each pupil has got an account, and therefore only one OOo user exists. Same in our teacher library, one PC user for all teachers. There And you don't know what teachers are there so you can give them all one account for him/herself? No argument. For the pupils, this might be if it's really a public PC, but then again you also know what pupils you have at the school and when you don't allow guests there you can add them as users, too. exist situations where a PC is quasi public and has only one PC user and one OOo user. A easy way to erase personal information in such Obviously a bad configuration. (And yes, we had the same at school in our library when I was at school) Regards, Rene - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
Re: [dev] buildbot builds vs standard builds
Hi, Thorsten Ziehm wrote: I do not see the need to bring the build bots near to the build environment here in Hamburg. The request for build bots was (as I know) to have builds in different environments to find build issues in these different environment. When these environment will be nearly the same, then we miss to find these build breakers. buildbots != tinderbox. Regards, Rene - To unsubscribe, e-mail: dev-unsubscr...@openoffice.org For additional commands, e-mail: dev-h...@openoffice.org
