amd64x2, NVIDIA fc5 compiling blues

[wirser]

Well I'm back

short side of long tale
blew up old athlon1ghz nvidia fc3 - bought new asus amd64x2 mb and put fc5 on

back to vegastrike :^)
The depends of orge .devil and cegui caused me a bit of problem with proper linkings until I made them into rpms (my spec files don't yet have requirements) . Which solved the linking problem at compile time for these. 8^b
next I ran into problems with after running bootstrap , configure not finding libpython.so
I fixed this with editing .configure and changing "/usr/ltb" to "/usr/64" in the search path line
now I get
src/cmd/alphacurve.cpp:37: warning: converting to ‘int’ from ‘double’
src/cmd/alphacurve.cpp:41: warning: converting to ‘int’ from ‘double’
src/cmd/alphacurve.cpp:85: warning: converting to ‘int’ from ‘double’
src/cmd/alphacurve.cpp:95: warning: converting to ‘int’ from ‘double’
depbase=`echo src/cmd/asteroid_generic.o | sed 's|[^/]*$|.deps/&|;s|\.o$||'`; \
if g++ -DHAVE_CONFIG_H -I. -I. -I. -I./boost/1_33 -DHAVE_SDL=1 -DSDL_WINDOWING=1 -DHAVE_AL=1 -DHAVE_OGG -I/usr/include/python2.4 -DHAVE_PYTHON=1 -I./src -g -O2 -pipe -falign-loops=2 -falign-jumps=2 -falign-functions=2 -I/usr/include/SDL -D_REENTRANT -pthread -pipe -MT src/cmd/asteroid_generic.o -MD -MP -MF "$depbase.Tpo" -c -o src/cmd/asteroid_generic.o src/cmd/asteroid_generic.cpp; \
then mv -f "$depbase.Tpo" "$depbase.Po"; else rm -f "$depbase.Tpo"; exit 1; fi
./src/cmd/collection.h:62: error: extra qualification ‘UnitCollection::UnitIterator::’ on member ‘operator=’
./src/cmd/collection.h:97: error: extra qualification ‘UnitCollection::ConstIterator::’ on member ‘operator=’
./src/rendertext.h:26: error: extra qualification ‘RText::’ on member ‘text_width’
./src/rendertext.h:27: error: extra qualification ‘RText::’ on member ‘draw_text’
./src/rendertext.h:36: error: extra qualification ‘RText::’ on member ‘ConsoleKeyboardI’
./src/rendertext.h:38: error: extra qualification ‘RText::’ on member ‘conoutn’
./src/cmd/weapon_xml.h:35: error: extra qualification ‘weapon_info::’ on member ‘weapon_info’
./src/cmd/script/flightgroup.h:87: error: extra qualification ‘Flightgroup::’ on member ‘operator=’
make[1]: *** [src/cmd/asteroid_generic.o] Error 1
make[1]: Leaving directory `/home/none/build/vegastrike'
make: *** [all] Error 2

what do I doe to help eliminate these ?

[klauss]

Code: Select all

svn update
make

[wirser]

then mv -f "$depbase.Tpo" "$depbase.Po"; else rm -f "$depbase.Tpo"; exit 1; fi
depbase=`echo src/cmd/nebula_generic.o | sed 's|[^/]*$|.deps/&|;s|\.o$||'`; \
if g++ -DHAVE_CONFIG_H -I. -I. -I. -I./boost/1_33 -DHAVE_SDL=1 -DSDL_WINDOWING=1 -DHAVE_AL=1 -DHAVE_OGG -I/usr/include/python2.4 -DHAVE_PYTHON=1 -I./src -g -O2 -pipe -falign-loops=2 -falign-jumps=2 -falign-functions=2 -I/usr/include/SDL -D_REENTRANT -pthread -pipe -MT src/cmd/nebula_generic.o -MD -MP -MF "$depbase.Tpo" -c -o src/cmd/nebula_generic.o src/cmd/nebula_generic.cpp; \
then mv -f "$depbase.Tpo" "$depbase.Po"; else rm -f "$depbase.Tpo"; exit 1; fi
src/cmd/nebula_generic.h:7: error: extra qualification ‘Nebula::’ on member ‘beginElement’
make[1]: *** [src/cmd/nebula_generic.o] Error 1
make[1]: Leaving directory `/home/none/build/vegastrike'
make: *** [all] Error 2

[klauss]

Rinse and repeat?

[wirser]

Rinse and repeat?

repeat these
just for kicks I ran the following
make -k 2> vsmakeerr.txt
see attached vsmakeerr.txt
At revision 10260
(edit)svn rev
(edited prevvous rant)

[wirser]

and so on and so on and so on.... ....
see previous post

[klauss]

Ok - if this goes on, I would really appreciate someone with a compiler which exhibits this issue to submit a patch. Otherwise this could go on forever.

[wirser]

If I knew what to change for sure I cuold make a diif ang submit it

basic question
are all of the "extra qualifier "errors solved the same way?
if yes what is this methode?

[klauss]

You'd have to know C++.
But it's very easy, basically, there are functions like this:

Code: Select all

class SomeClass {
...
     SomeClass& SomeClass::operator+(blah)...
...
     int SomeClass::FunctionBlah() ...
...
}

In both cases, you have to delete the "SomeClass::" (not the other SomeClass..es):

Code: Select all

class SomeClass {
...
     SomeClass& operator+(blah)...
...
     int FunctionBlah() ...
...
}

(but only when they're inside a class { ... } block - usually, whatever line the error reference, is inside one of those blocks).

[wirser]

so as of svn 10246 compile only objconv/trisort.h had an extraqualifier so to fix it I would make the following change

[none@localhost objconv]$ diff trisort.h trisort.hold
41,42c41,42
< Index processfacevertex(char * chunk) const;
< Face processface( char * line)const;
---
> Index Mesh::processfacevertex(char * chunk) const;
> Face Mesh::processface( char * line)const;

my only problem in compile is (':oops:')
appearntly myself for not have freealut/devel installed
now I have vs compiled (yet uninstalled) data4.x and music at level described in topic:
http://vegastrike.sourceforge.net/forum ... php?t=6803
even tried the zipped sh file (I guess changing the name was too simple)
I have x-server problems after running also no new windows willl open until restarting x

what goes where?
how to start when it is put there please ?

Now about all of those warnings .....

[klauss]

About the X trouble, it's a bug we've been unable to track down. I suspect it's SDL, but since every other SDL app seems not to have that problem, it raises serious doubts. In any case, if you're not concerned about network security issues (which, BTW, are nonexistent unless you have your X listening on the network, which isn't the default last time I checked - it listens on an unix socket, which is local). Anyway... if you're not concerned about that, you do "xhost +local" prior to launching VS.

Ok, ok... before anyone says anything, yes, xhost +local could be troublesome if you provide remote terminals. But... who does? And plays VS there? Really...

[wirser]

well I finally have a 86-64 self compiled vegastrike what follows is how to repeat

Code: Select all

mkdir ~/build
cd ~/build

svn checkout https://svn.sourceforge.net/svnroot/vegastrike/trunk/vegastrike
svn checkout https://svn.sourceforge.net/svnroot/vegastrike/trunk/data4.x
svn checkout https://svn.sourceforge.net/svnroot/vegastrike/trunk/music
#svn update vegastrike data4.x music
cd vegastrike
./bootstrap
#edit ./configure to add "/usr/lib64" to python search paths
./configure
#make clean after first time
make
cd ~/build/data4.x
ln -s ../vegastrike/vegastrike
ln -s ../vegastrike/soundserver

#to run cd ~/build/data4.x
########EDIT#######
# xhost +localhost (+local doesn't work for me)  it should have been xhost +local:
#which does work ,the ":" makes all of the difference
# ./vegastrike 

BTW even using xhost my sound gets clobbered needing to restart x and may even need reboot to restore (that I'm still investigating)

[abolakomadic]

i don't like it.

[wirser]

i don't like it.

what is the "it" you dislike?

lots of things I don't like but at least I've filed bug reports on what I have been finding wrong have you?

[Halleck]

'tis a spambot, pay it no heed.

[wirser]

What a RIP

I just got the airlock party invititations back from the printer

[wirser]

as of svn 10260

0 errors
177 compiler warnings

In Chicago you get the "Denver boot" for less than that

[klauss]

I bet it's pointer-to-long-to-pointer conversions, or things like that.
If you either post, send, or whatever the log with the warnings, someone could fix them, if you won't

[wirser]

err I thought I did
must of deleted ithe old make output but didn't replace it with the latest output
on the repeat this reply, trying to be efficent and it backfired ... anyways

I bet it's pointer-to-long-to-pointer conversions, or things like that.

'bout right

anyways the previous compiler error output is both here and above
latest below only

[plueschinger]

Hello

If I tell silly things, sorry, but this looks very strang to me:

Code: Select all

In file included from src/cmd/collide/prapid.cpp:68:
src/cmd/collide/garray.h:142:1: warning: "CS_DECLARE_GROWING_ARRAY_REF" redefined
In file included from src/cmd/collide/prapid.cpp:63:
src/cmd/collide/cs_compat.h:80:1: warning: this is the location of the previous definition

Regards
Plueschinger

[klauss]

Thanks for the log.
I'll try to address them one by one when I get the time.
I get annoyed by warnings too - they clutter output, making it hard to look for real errors.
Some of them are hard to solve, though... like some of those float<->double warnings... it isn't the best idea to just put a cast to make the warning disappear. So I guess I'll take a look with a bit more time to think of a more appropriate solution.

[wirser]

well your putting a good dent in them

I understand that the latest compilers are making it tough to pass "unfinished" code too
It is good to see someone willing to work these problems out ,have the ime ...
If my feedback is helping you great
the charecter mis-conversions really looked to me to be able to solve game play problems in the right place but I not familar enough with the structures to be able to say where that would happen wink

so far during battle my fps goes to single digits is my biggest problem
every system give alot of factions pngs not found msgs varies with system
and the comms used to be better in 0.4.3
it is amussing to have the robot bartender praising the laws on banning robot bartenders

but todays national combustion day, national instabilty day,national loud oise day, national main street inspection day, national dreams realized day

[wirser]

It's broke and the attached proves it

a bit off topic but are there any objections to me deleting all of my post older than say one month ?

here are some things that either will put a dent in the amount of warnings
or are easy fixes

src/cmd/collection.cpp
src/cmd/alphacurve.cpp
src/cmd/csv.cpp
src/endianness.cpp
src/networking/softvolume.h
src/cmd/alphacurve.cpp
warning: no newline at end of file

src/cmd/csv.h:32: warning: overflow in implicit constant conversion
33 times

src/cmd/collection.h:54: warning: inline function ‘void UnitCollection::UnitIterator::GetNextValidUnit()’ used but never defined
41 times

[klauss]

a bit off topic but are there any objections to me deleting all of my post older than say one month ?

Why would you do that?
In any case, it's generally not advised to delete old posts (if you can - I got the impression you couldn't delete old posts unless you were a mod), mainly because it makes the conversations there hard to follow (you know... the telephone syndrome).

[wirser]

a bit off topic but are there any objections to me deleting all of my post older than say one month ?

Why would you do that?
In any case, it's generally not advised to delete old posts (if you can - I got the impression you couldn't delete old posts unless you were a mod), mainly because it makes the conversations there hard to follow (you know... the telephone syndrome).

It's more of a precaution against using up my alloted user/dl space. I really haven't been active until lately and I'm may get involved with debugging the xauthoriy problem that myself and other are having. Depending on how long it will take to solve I cound generate a lot of debugging msgs

As far as breaking conversations, what about making an archive of the forums on a regular bassis? Now this project ihas grown to it's present size the forums have gottem a bit cluttered. I've got post a couple of years old that most could care less about.

So I say "Why not"