Re: [gpsd-dev] [PATCH] fix DESTDIR handling with udev rules install again

[Fulup Le Foll]

Eric,

I do not have a login to commit patch directly on gpsd git. All my patches where send to the list and applied by someone else. Which does not mean they did not embed stupid things.

I send two patches,

• First one was applied  [gps2udp creation]
  
• Second one was lost [see attached version rebuilt with current git version]
  

1st patch was send 30th of April [gps2udp.c utility creation]

• add two files: gps2udp.c and gpsd2udp.xml
• modify SConstruct to include compilation rules for gps2udp
  

2nd patch send 25th of July [minor gps2udp correction] This patch was not lost somewhere in the process.

Conclusion:

• my patches should not impact udev/DESTDIR, if they do it is a mistake
  
• my second patch here attached should be committed in order to fix gps2udp man page and command line parsing.
  

Fulup

Le 23/11/2013 06:07, Eric S. Raymond a écrit :

Mike Frysinger <address@hidden>:

Commit 102e29d16c3fd9b414eeac53c794302f902ae8da reverted the DESTDIR
logic when upgrading the udev path handling.

I fixed that with commit a41cd69c87e312198834a2e6232134176c9352e5.

It's been reverted yet again in an unrelated commit
57e9bdeae2f00664616ee1d9a936a5cfe27bfd30.  Why do people keep doing this?

That one commit by Fulup Ar Foll seems to have randomly stepped on the
build recipe in *two different* places. Bernd Zeimetz was having one
of his usual post-release shit-fits about scons on #gpsd about three
hours because 57e9 broke linkage of gpsmon when the DBUS option is on.
(I was off at Friday night gaming and missed the excitement.)
 
I've copied Fulup; maybe he can explain why these apparently unnecessary
and completely unrelated changes were in there with 57e9.

This is partly my fault for not having noticed the build changes in that
patch.  Of course, nobody else noticed them either, and that was
six months ago in April...

0001-Updated-documentation-and-fixes-gps2udp-command-line.patch
Description: Text Data