Re: svn commit: r1827215 - in /openoffice/trunk/main: configure.ac solenv/gbuild/platform/macosx.mk

2018-03-19 Thread Don Lewis 
On 19 Mar, Jim Jagielski wrote:
> The later doesn't help w/ the bridges stuff :/
> 
> I'll try to figure out where the gcc3 is leaking in and plug that.

It is probably leaking in via the environment.  Did you rerun the script
to set the environment after re-running autoconf and configure?  In any
case, look for gcc3 in the environment.



-
To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org
For additional commands, e-mail: dev-h...@openoffice.apache.org

Re: svn commit: r1827215 - in /openoffice/trunk/main: configure.ac solenv/gbuild/platform/macosx.mk

2018-03-19 Thread Jim Jagielski 
The later doesn't help w/ the bridges stuff :/

I'll try to figure out where the gcc3 is leaking in and plug that.

> On Mar 19, 2018, at 8:10 PM, Don Lewis  wrote:
> 
> Either go back to the earlier code and change the equality test to a
> case statement that matches darwin*, or change the other instances of
> s5abi to gcc3.
> 
> On 19 Mar, Jim Jagielski wrote:
>> OK, so what do I do? How to fix this? Any ideas?
>> 
>>> On Mar 19, 2018, at 7:58 PM, Don Lewis  wrote:
>>> 
>>> On 19 Mar, Jim Jagielski wrote:
 Yeah, I see that now... Before my commits, the macOS builds
 broke mid-stream, due to the rdbmaker stuff, et.al... as could
 be seen, there was places where macOS was using/expecting
 gcc3 and others where it was s5abi. 4.1.5 seemed to use
 s5abi universally, but so the issue was/is gcc3 slipping in
 somewhere.
 
 We do have bridges/source/cpp_uno/gcc3_macosx_intel
>>> 
>>> That's a 32-bit build.
>>> 
>> 
>> 
>> -
>> To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org
>> For additional commands, e-mail: dev-h...@openoffice.apache.org
>> 
> 
> 
> -
> To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org
> For additional commands, e-mail: dev-h...@openoffice.apache.org
> 


-
To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org
For additional commands, e-mail: dev-h...@openoffice.apache.org

Re: svn commit: r1827215 - in /openoffice/trunk/main: configure.ac solenv/gbuild/platform/macosx.mk

2018-03-19 Thread Don Lewis 
Either go back to the earlier code and change the equality test to a
case statement that matches darwin*, or change the other instances of
s5abi to gcc3.

On 19 Mar, Jim Jagielski wrote:
> OK, so what do I do? How to fix this? Any ideas?
> 
>> On Mar 19, 2018, at 7:58 PM, Don Lewis  wrote:
>> 
>> On 19 Mar, Jim Jagielski wrote:
>>> Yeah, I see that now... Before my commits, the macOS builds
>>> broke mid-stream, due to the rdbmaker stuff, et.al... as could
>>> be seen, there was places where macOS was using/expecting
>>> gcc3 and others where it was s5abi. 4.1.5 seemed to use
>>> s5abi universally, but so the issue was/is gcc3 slipping in
>>> somewhere.
>>> 
>>> We do have bridges/source/cpp_uno/gcc3_macosx_intel
>> 
>> That's a 32-bit build.
>> 
> 
> 
> -
> To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org
> For additional commands, e-mail: dev-h...@openoffice.apache.org
> 


-
To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org
For additional commands, e-mail: dev-h...@openoffice.apache.org

Re: svn commit: r1827215 - in /openoffice/trunk/main: configure.ac solenv/gbuild/platform/macosx.mk

2018-03-19 Thread Jim Jagielski 
OK, so what do I do? How to fix this? Any ideas?

> On Mar 19, 2018, at 7:58 PM, Don Lewis  wrote:
> 
> On 19 Mar, Jim Jagielski wrote:
>> Yeah, I see that now... Before my commits, the macOS builds
>> broke mid-stream, due to the rdbmaker stuff, et.al... as could
>> be seen, there was places where macOS was using/expecting
>> gcc3 and others where it was s5abi. 4.1.5 seemed to use
>> s5abi universally, but so the issue was/is gcc3 slipping in
>> somewhere.
>> 
>> We do have bridges/source/cpp_uno/gcc3_macosx_intel
> 
> That's a 32-bit build.
> 


-
To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org
For additional commands, e-mail: dev-h...@openoffice.apache.org

Re: svn commit: r1827215 - in /openoffice/trunk/main: configure.ac solenv/gbuild/platform/macosx.mk

2018-03-19 Thread Don Lewis 
On 19 Mar, Jim Jagielski wrote:
> Yeah, I see that now... Before my commits, the macOS builds
> broke mid-stream, due to the rdbmaker stuff, et.al... as could
> be seen, there was places where macOS was using/expecting
> gcc3 and others where it was s5abi. 4.1.5 seemed to use
> s5abi universally, but so the issue was/is gcc3 slipping in
> somewhere.
> 
> We do have bridges/source/cpp_uno/gcc3_macosx_intel

That's a 32-bit build.


-
To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org
For additional commands, e-mail: dev-h...@openoffice.apache.org

Re: svn commit: r1827215 - in /openoffice/trunk/main: configure.ac solenv/gbuild/platform/macosx.mk

2018-03-19 Thread Jim Jagielski 
Yeah, I see that now... Before my commits, the macOS builds
broke mid-stream, due to the rdbmaker stuff, et.al... as could
be seen, there was places where macOS was using/expecting
gcc3 and others where it was s5abi. 4.1.5 seemed to use
s5abi universally, but so the issue was/is gcc3 slipping in
somewhere.

We do have bridges/source/cpp_uno/gcc3_macosx_intel


> 
> Until my commit in r1826296, Mac was using s5abi.  There are still some
> instances of s5abi lurking about in odk/setsdkenv_unix.sh.in and
> odk/settings/settings.mk.  The bridges code for the Mac is under
> bridges/source/cpp_uno/s5abi_macosx_x86-64, but is enabled by $(OS)$(CPU)
> and not $(COMID) or $(COMNAME).
> 


-
To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org
For additional commands, e-mail: dev-h...@openoffice.apache.org

Re: svn commit: r1827215 - in /openoffice/trunk/main: configure.ac solenv/gbuild/platform/macosx.mk

2018-03-19 Thread Don Lewis 
On 19 Mar, Jim Jagielski wrote:
> 
> 
>> On Mar 19, 2018, at 6:47 PM, Don Lewis  wrote:
>> 
>> I don't think this is correct.  As I recall the bridges code for Mac
>> uses s5abi.
>> 
> 
> As long as we remain consistent, we should be OK... Defaulting
> to gcc3 seems to work better.

Until my commit in r1826296, Mac was using s5abi.  There are still some
instances of s5abi lurking about in odk/setsdkenv_unix.sh.in and
odk/settings/settings.mk.  The bridges code for the Mac is under
bridges/source/cpp_uno/s5abi_macosx_x86-64, but is enabled by $(OS)$(CPU)
and not $(COMID) or $(COMNAME).


-
To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org
For additional commands, e-mail: dev-h...@openoffice.apache.org

Re: extension problem

2018-03-19 Thread Pedro Lino 

> On March 19, 2018 at 8:32 PM Dave Brondsema  wrote:
> 
> 
> On 3/16/18 4:37 PM, Pedro Lino wrote:
> > But Comments on extensions are still broken and return
> > 
> > "Your comment is waiting for moderator approval.
> > 
> > Error message
> > 
> > Unable to send e-mail. Contact the site administrator if the problem 
> > persists."
> > 
> 
> Ok, emailing has been fixed now too.

Thanks! Now I got "Your comment is waiting for moderator approval."
Hopefully some moderator will approve it :)

-
To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org
For additional commands, e-mail: dev-h...@openoffice.apache.org

Re: svn commit: r1827215 - in /openoffice/trunk/main: configure.ac solenv/gbuild/platform/macosx.mk

2018-03-19 Thread Jim Jagielski 


> On Mar 19, 2018, at 6:47 PM, Don Lewis  wrote:
> 
> I don't think this is correct.  As I recall the bridges code for Mac
> uses s5abi.
> 

As long as we remain consistent, we should be OK... Defaulting
to gcc3 seems to work better.

-
To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org
For additional commands, e-mail: dev-h...@openoffice.apache.org

Re: Openoffice and unsupported gstreamer 0.10 branch (for openoffice libavmediagst.so library)

2018-03-19 Thread Jim Jagielski 


> On Mar 19, 2018, at 4:44 PM, Rory O'Farrell  wrote:
> 
> On Mon, 19 Mar 2018 16:34:44 -0400
> Jim Jagielski  wrote:
> 
>> Does it make sense to have both gstreamer 0.10 and 1.0 as exclusive options.
>> That is, enabling gstreamer during the build will determine what
>> version is supported and then use that?
>> 
>> ie, sort of merge how AOO415 and AOO420 currently do things related
>> to gstreamer?
> 
> I think all recent versions (6-7 years?) of linux use gstreamer 1.0.
> 

CentOS6 does not, and that is, or has been, our official
community build server.


-
To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org
For additional commands, e-mail: dev-h...@openoffice.apache.org

Re: svn commit: r1827215 - in /openoffice/trunk/main: configure.ac solenv/gbuild/platform/macosx.mk

2018-03-19 Thread Don Lewis 
On 19 Mar, j...@apache.org wrote:
> Author: jim
> Date: Mon Mar 19 16:53:04 2018
> New Revision: 1827215
> 
> URL: http://svn.apache.org/viewvc?rev=1827215=rev
> Log:
> Use gcc3 for COMID for macOS/OSX
> 
> Modified:
> openoffice/trunk/main/configure.ac
> openoffice/trunk/main/solenv/gbuild/platform/macosx.mk
> 
> Modified: openoffice/trunk/main/configure.ac
> URL: 
> http://svn.apache.org/viewvc/openoffice/trunk/main/configure.ac?rev=1827215=1827214=1827215=diff
> ==
> --- openoffice/trunk/main/configure.ac (original)
> +++ openoffice/trunk/main/configure.ac Mon Mar 19 16:53:04 2018
> @@ -2237,13 +2237,8 @@ case $COM in
>   CCNUMVER=`echo ${CCVER} | $AWK -v num=true -f 
> ${_solenv}/bin/getcompver.awk`;
>   # The version in tg_compv.mk for darwin does not look correct
>   # CCNUMVER = `echo ${CCVER} | sed -e"s/\.//";
> - if test "$build_os" == "darwin"; then
> - COMID="s5abi";
> - COMNAME="s5abi";
> - else
> - COMID="gcc3";
> - COMNAME="gcc3";
> - fi;;
> + COMID="gcc3";
> + COMNAME="gcc3";
>   MSC)
>   CCVER=`${CC} 2>&1 | $AWK -f ${_solenv}/bin/getcompver.awk`;
>   CCNUMVER=`${CC} 2>&1 | $AWK -v num=true -f 
> ${_solenv}/bin/getcompver.awk`;
> 
> Modified: openoffice/trunk/main/solenv/gbuild/platform/macosx.mk
> URL: 
> http://svn.apache.org/viewvc/openoffice/trunk/main/solenv/gbuild/platform/macosx.mk?rev=1827215=1827214=1827215=diff
> ==
> --- openoffice/trunk/main/solenv/gbuild/platform/macosx.mk (original)
> +++ openoffice/trunk/main/solenv/gbuild/platform/macosx.mk Mon Mar 19 
> 16:53:04 2018
> @@ -22,8 +22,6 @@
>  
>  
>  GUI := UNX
> -COM := S5ABI
> -COMID := s5abi
>  #COM := S5ABI
>  #COMID := s5abi
>  
> 
> 

I don't think this is correct.  As I recall the bridges code for Mac
uses s5abi.


-
To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org
For additional commands, e-mail: dev-h...@openoffice.apache.org

Re: rdbmaker issues

2018-03-19 Thread Don Lewis 
On 19 Mar, Jim Jagielski wrote:
> The migration of rdbmaker to build is causing issues w/ macOS:
> It's looking for libuno_cppuhelpers5abi.dylib but what is being
> created is libuno_cppuhelpersgcc3.dylib
> 
> No idea where the names are being generated or managed, since
> the makefile uses
> 
> 
> $(eval $(call gb_Executable_add_linked_libs,rdbmaker,\
>   cppu \
>   cppuhelper \
>   reg \
>   sal \
>   salhelper \
>   stl \
>   $(gb_STDLIBS) \
> ))

This could be due to my changes to get away from tg_compv.mk.
configure.ac contains:

if test "$build_os" == "darwin"; then
COMID="s5abi";
COMNAME="s5abi";
else
COMID="gcc3";
COMNAME="gcc3";
fi;;

Other places that look at $build_os allow for trailing stuff, so this
may need to be changed to some sort of wildcard match rather than strict
equality.


-
To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org
For additional commands, e-mail: dev-h...@openoffice.apache.org

Re: Openoffice and unsupported gstreamer 0.10 branch (for openoffice libavmediagst.so library)

2018-03-19 Thread Kay Schenk 



On 03/19/2018 01:44 PM, Rory O'Farrell wrote:

On Mon, 19 Mar 2018 16:34:44 -0400
Jim Jagielski  wrote:


Does it make sense to have both gstreamer 0.10 and 1.0 as exclusive options.
That is, enabling gstreamer during the build will determine what
version is supported and then use that?

ie, sort of merge how AOO415 and AOO420 currently do things related
to gstreamer?

I think all recent versions (6-7 years?) of linux use gstreamer 1.0.

As near as I can determine, neither CentOS 6 or CentOS 7 provide 
gstreamer 1.0. The pack namings for CentOS7 were different than for 
CentOS 6, but, I think they're both gstreamer 0.10. I couldn't find 
gsteamer 1.0 in any of the accepted "extra" repos for either of these 
either. Maybe someone else can investigate. I, personally, don't plan on 
using a different Linux.


--
--
MzK

"Ring out the false, ring in the true."
 -- poem "In Memoriam", Alfred Lord Tennyson


-
To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org
For additional commands, e-mail: dev-h...@openoffice.apache.org

Re: Openoffice and unsupported gstreamer 0.10 branch (for openoffice libavmediagst.so library)

2018-03-19 Thread Rory O'Farrell 
On Mon, 19 Mar 2018 16:34:44 -0400
Jim Jagielski  wrote:

> Does it make sense to have both gstreamer 0.10 and 1.0 as exclusive options.
> That is, enabling gstreamer during the build will determine what
> version is supported and then use that?
> 
> ie, sort of merge how AOO415 and AOO420 currently do things related
> to gstreamer?

I think all recent versions (6-7 years?) of linux use gstreamer 1.0.

-- 
Rory O'Farrell 

-
To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org
For additional commands, e-mail: dev-h...@openoffice.apache.org

Re: Openoffice and unsupported gstreamer 0.10 branch (for openoffice libavmediagst.so library)

2018-03-19 Thread Jim Jagielski 
Does it make sense to have both gstreamer 0.10 and 1.0 as exclusive options.
That is, enabling gstreamer during the build will determine what
version is supported and then use that?

ie, sort of merge how AOO415 and AOO420 currently do things related
to gstreamer?

-
To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org
For additional commands, e-mail: dev-h...@openoffice.apache.org

Re: extension problem

2018-03-19 Thread Dave Brondsema 
On 3/16/18 4:37 PM, Pedro Lino wrote:
> 
>> On March 16, 2018 at 8:12 PM Andrea Pescetti  wrote:
>>
>>
>> Dave Brondsema wrote:
>>> On 3/15/18 6:57 PM, Andrea Pescetti wrote:
 For example, https://t.co/sZTK7wwASJ and https://t.co/HV0HdhJLGR (listed 
 there)
 are still broken. And others from that list are broken too. But I guess 
 it's
 just a matter of moving them to the right place now.
>>>
>>> The other extensions are fixed now too.
>>
>> Thanks, it works for me now: I can download recent releases of other 
>> extensions too.
> 
> But Comments on extensions are still broken and return
> 
> "Your comment is waiting for moderator approval.
> 
> Error message
> 
> Unable to send e-mail. Contact the site administrator if the problem 
> persists."
> 

Ok, emailing has been fixed now too.


-- 
Dave Brondsema : d...@brondsema.net
http://www.brondsema.net : personal
http://www.splike.com : programming
  <><

-
To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org
For additional commands, e-mail: dev-h...@openoffice.apache.org

macOS (Re: Status on 4.2.0? )

2018-03-19 Thread Jim Jagielski 
The macOS builds are ready. I'll wait until my Linux-64bit builds are
done, via Ubuntu 14.04, before I upload them. Then I'll start on the
Linux 32 bit ones, also Ubuntu.



-
To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org
For additional commands, e-mail: dev-h...@openoffice.apache.org

rdbmaker issues

2018-03-19 Thread Jim Jagielski 
The migration of rdbmaker to build is causing issues w/ macOS:
It's looking for libuno_cppuhelpers5abi.dylib but what is being
created is libuno_cppuhelpersgcc3.dylib

No idea where the names are being generated or managed, since
the makefile uses


$(eval $(call gb_Executable_add_linked_libs,rdbmaker,\
cppu \
cppuhelper \
reg \
sal \
salhelper \
stl \
$(gb_STDLIBS) \
))


-
To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org
For additional commands, e-mail: dev-h...@openoffice.apache.org

Re: Issue w/ lucene build/patch in trunk

2018-03-19 Thread Jim Jagielski 
Fixed in r1827202... there were some EOL gremlins in the patch-file which
caused patch to barf

> On Mar 14, 2018, at 10:11 AM, Jim Jagielski  wrote:
> 
> I am getting the below error:
> 
> =
> Building module lucene
> =
> 
> Entering /Users/jim/src/asf/AOO420/main/lucene
> 
> missing header for unified diff at line 3 of patch
> patching file lucene-2.9.4/common-build.xml
> Reversed (or previously applied) patch detected!  Assume -R? [n] y
> patching file lucene-2.9.4/common-build.xml
> Reversed (or previously applied) patch detected!  Assume -R? [n] y
> patching file lucene-2.9.4/contrib/analyzers/common/build.xml
> Reversed (or previously applied) patch detected!  Assume -R? [n] y
> (Stripping trailing CRs from patch.)
> patching file lucene-2.9.4/contrib/analyzers/smartcn/build.xml
> Hunk #1 FAILED at 23.
> 1 out of 1 hunk FAILED -- saving rejects to file 
> lucene-2.9.4/contrib/analyzers/smartcn/build.xml.rej
> dmake:  Error code 1, while making 
> './unxmaccx.pro/misc/build/so_patched_so_lucene'
> 
> 
> The rejected patch is:
> 
> % cat 
> lucene/unxmaccx.pro/misc/build/lucene-2.9.4/contrib/analyzers/smartcn/build.xml.rej
>  
> ***
> *** 23,30 
>  Smart Chinese Analyzer
>
> 
> -   
> -   
>   
> location="../../../build/contrib/analyzers/smartcn" />
> location="../../../dist/contrib/analyzers/smartcn" />
> --- 23,30 
>  Smart Chinese Analyzer
>
> 
> +   
> +   
>   
> location="../../../build/contrib/analyzers/smartcn" />
> location="../../../dist/contrib/analyzers/smartcn" />
> 
> 
> -
> To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org
> For additional commands, e-mail: dev-h...@openoffice.apache.org
> 


-
To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org
For additional commands, e-mail: dev-h...@openoffice.apache.org