Hi Pavel,

First, I want to explain "why" I have some strange "commits" : because we're several to build on Max OSX, it is important to propose *first* all builders the complete donated code, and choose *after* which part has to be used or deleted.I simply refuse to decide alone in this case.

Of course, even if all this code will be integrated, OOo will always looks like OOo. ;-)


Pavel Janík a écrit :

   > berkeleydb (libdb_java not built)

this change is:

.IF "$(OS)"=="MACOSX"
OUT2LIB+=$(BUILD_DIR)$/.libs$/libdb*.jnilib
.ENDIF

but we use different code in 2.0:

.IF "$(OS)"=="MACOSX"
OUT2LIB+=$(BUILD_DIR)$/.libs$/libdb_java*jnilib
.ENDIF  # "$(OS)"=="MACOSX"

Can you please use the same code?

Yes, you're right, this is more clean to use the same for both 1.1.x and 2.0.


> desktop (mozwrapper and nswrapper) ,
module desktop contains mozwrapper also in 2.0. Should it be adapted there
too?

I think so. I just wasn't present last week. This will be done too as soon as possible.


   > dlcompat (fixed for 10.2 build)

module dlcompat's build.lst contains this change:

Index: prj/build.lst
===================================================================
RCS file: /cvs/external/dlcompat/prj/build.lst,v
retrieving revision 1.1
retrieving revision 1.1.10.1
diff -u -r1.1 -r1.1.10.1
--- prj/build.lst       22 Mar 2004 15:39:02 -0000      1.1
+++ prj/build.lst       3 Jul 2005 15:01:19 -0000       1.1.10.1
@@ -1,3 +1,3 @@
-dc     dlcompat        :       soltools NULL
+dc     dlcompat        :       soltools dlcompat NULL
 dc     dlcompat                usr1    -       all     dc_mkout NULL
 dc     dlcompat      nmake     -       all     dc_dlcompat NULL

It makes dlcompat depending on itself which is nonsense. What is the
purpose of this fix? Reopening #i51507#.

P. Luby's commit is not complete here. Some other part, probably located in external, is still missing.
IMHO : I propose to delete this change


   > sc (fixes for Hindu and Kannada)

Hmm, he change here is:

+
+# Indic resources are too big for the resource compiler
+RES_HINDI :=
+RES_KANNADA :=

in makefile.mk. This is probably trying to fix #i33228#. See this issues

Yes. I had no issue about this, believing it was Neo or something like that only. Thank you

for more details and the reasoning for not fixing this and the workaround
that already is in the source. We do not need to pollute code with another
workaround.


Ok for delete this too.


   > shell (the mailer can open documents with exotic filenames)

The changes here are adding

... && !defined USE_JAVA

to #if defined MACOSX. Why? Look at rev. 1.2.24.1.14.2 of cmdmailmsg.cxx
where pluby did this:

-#if defined MACOSX && ! defined USE_JAVA
+#if 0 // defined MACOSX

Why you reverted it back to

-#if 0 // defined MACOSX
+#if defined MACOSX && !defined USE_JAVA

in revision 1.2.24.1.14.3?

Maybe I've made a mistake ?  I'll have a closer look.


   > svtools (?)

Yes, there is no change in module svtools.

I have found nothing. Maybe someone can complete ?

Thank you Pavel for all your comments, this is very helpfull for me.


Regards,
eric bachard

--
<[EMAIL PROTECTED]>
Francophone OpenOffice.org Community developer ( Linux PPC and Mac OS X / X11)
See : <http://fr.openoffice.org>


---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to