Problem compiling latest CVS-release

[Frix]

Since I was curious about the improvement in the logviewer, I picked up the latest build from the CVS.
However, it won't compile on my Xubuntu 8.10 when it comes to '3d_vetable.c'

Code: Select all

3d_vetable.c: In function ‘ve3d_draw_runtime_indicator’:
3d_vetable.c:1180: error: incompatible type for argument 2 of ‘get_fixed_pos’
3d_vetable.c:1180: error: too many arguments to function ‘get_fixed_pos’
3d_vetable.c:1181: error: incompatible type for argument 2 of ‘get_fixed_pos’
3d_vetable.c:1181: error: too many arguments to function ‘get_fixed_pos’
3d_vetable.c: In function ‘ve3d_key_press_event’:
3d_vetable.c:1454: warning: dereferencing type-punned pointer will break strict-aliasing rules
3d_vetable.c:1789: warning: dereferencing type-punned pointer will break strict-aliasing rules
3d_vetable.c:1805: warning: dereferencing type-punned pointer will break strict-aliasing rules
3d_vetable.c: At top level:
3d_vetable.c:2415: error: conflicting types for ‘get_fixed_pos’
../include/3d_vetable.h:217: error: previous declaration of ‘get_fixed_pos’ was here
make[2]: *** [3d_vetable.o] Error 1
make[2]: Leaving directory `/home/stefr95/cvs/MegaTunix/src'
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory `/home/stefr95/cvs/MegaTunix'
make: *** [all] Error 2

The stable release compiles ok, and runs just fine (except for the Logviewer...)

[pinzgauer]

I can get it to compile on ubuntu 8.04 with a scratch checkout from cvs, but get segmentation faults and other errors depending on which controller I select. Feels to me like something has changed in the realtime map, profile, or other files again.

I'm going to work backwards datewise in the CVS until I can see what changed. The dev .19 branch was working (mostly) in early january. We know Dave is changing many things, so may need to let it settle.

Have fun,

Alan

[Frix]

No problem, I can wait... I was just curious to see what the logviewer looked like.
Of course, 5 minutes after I did my post here, I went over to the msefi.com forums
just to find an identical report. (http://www.msefi.com/viewtopic.php?f=76&t=34455)
Oh well...

[djandruczyk]

No problem, I can wait... I was just curious to see what the logviewer looked like.
Of course, 5 minutes after I did my post here, I went over to the msefi.com forums
just to find an identical report. (http://www.msefi.com/viewtopic.php?f=76&t=34455)
Oh well...

If you want to test CVS, you need to use the correct branch. PRE_0_9_19-branch. Right now it's very much in flux. I've broken the API for the interrogation profiles and a bunch of other things, so custom profiles will not load, and there's still a handful of instability until I get my latest set of changes completed.. Most of this is due to MS2 and the pile of "implicit rules" that it broke compared to MS-I firmwares, forcing me to make the changes.

The logviewer hasn't changed much, and is sorely in need of a rewrite, as it's pretty kludgy, that I haven't gotten around to it yet. I hope to model it based upon some of the better points from megalogviewer (which is an excellent app in most respects)