Skip to content

Commit

Permalink
[osx/ios] since we build libboost_date_time, might as well make with …
Browse files Browse the repository at this point in the history 
…correct internal paths
  • Loading branch information
@davilla
davilla committed Jul 11, 2011
1 parent 6976ed2 commit bbdd73f
Showing 1 changed file with 3 additions and 0 deletions.
3 changes: 3 additions & 0 deletions tools/darwin/depends/boost/Makefile
View file
Expand Up @@ -25,8 +25,11 @@ $(TARBALLS_LOCATION)/$(ARCHIVE):
if test "$(DARWIN)" = "ios"; then \
mv $(PREFIX)/lib/libboost_thread.so.$(LIBVERSION) $(PREFIX)/lib/libboost_thread.dylib; \
rm $(PREFIX)/lib/libboost_thread.so; \
mv $(PREFIX)/lib/libboost_date_time.so.$(LIBVERSION) $(PREFIX)/lib/libboost_date_time.dylib; \
rm $(PREFIX)/lib/libboost_date_time.so; \
fi
install_name_tool -id $(PREFIX)/lib/libboost_thread.dylib $(PREFIX)/lib/libboost_thread.dylib
install_name_tool -id $(PREFIX)/lib/libboost_date_time.dylib $(PREFIX)/lib/libboost_date_time.dylib
touch .installed

clean:
Expand Down

6 comments on commit bbdd73f

@jimfcarroll
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Well ... this, and serialization are needed because of threads. Since I now have the threading so that we can plug in other implementations, and since I have a pthreads implementation done, I think we could drop these three boost packages if I did a native windows implementation (which should simply follow the same pattern as the pthreads implementation). What do you think?

@davilla
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

we don't link to libboost_date_time, nor does libboost_threads have refs to libboost_date_time. I just added it because I saw some ref to it and figured we might want it sooner or later.

^^ in other words, drop back to pthreads on osx/ios/linux and win native on windows ?

@jimfcarroll
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It's an option. It's what I was thinking when I wrote the above.

After talking to elupus I think perhaps boost ought to be the default implementation. But on platforms where that doesn't work there's options (currently native pthreads works). Eventually the separate implementations can be moved to the new std::thread stuff which will (I think) look a lot like boost anyway.

@jimfcarroll
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Turns out, even with the pthreads implementation, I need boost::date_time. Oh well.

@davilla
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

certainly not a link to libboost_date_time as xbmc does not static link to that lib. must be headers only.

@jimfcarroll
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

true. Which I guess makes it OK - as long as it works on all of our platforms.

Please sign in to comment.