Compiling 0.4.2 on Mandrake 10.1

SpaceCowboy · Post by **SpaceCowboy** » Tue Oct 19, 2004 6:51 pm

H All,

I just started running Linux about three weeks ago, so I should start off by saying that I have a limited understanding of what's going on here.

Anyways, I downloaded and ran the 0.4.2 install script, which went well until it tried to actually start compiiling:

checking whether we are using the GNU C++ compiler... yes
checking whether g++ accepts -g... yes
checking how to run the C++ preprocessor... g++ -E
checking for gcc... gcc
checking whether we are using the GNU C compiler... yes
checking whether gcc accepts -g... yes
checking for gcc option to accept ANSI C... none needed
checking how to run the C preprocessor... gcc -E
configure: Using compiler: g++.
configure: error: cannot find install-sh or install.sh in . ./.. ./../..
Error running ./configure

Like I said, I only just started with Linux, so what's the deal with this?

Post by **hellcatv** » Wed Oct 20, 2004 5:53 am

hmm try going into the vegastrike directory that the install script downloaded and running

aclocal
autoheader
autoconf
automake -a
./configure

see if that helps you

SpaceCowboy · Post by **SpaceCowboy** » Wed Oct 20, 2004 6:22 pm

Yeah, that seemed to do it. Though now it complains:

Enabling user-specified flags:
checking for Win32 platform... no
checking for CygWin platform... no
checking for Mingw32 platform... no
checking for Linux platform... yes
checking for FreeBSD platform... no
checking for MacOSX platform... no
checking for python... configure: error: *** Python version 2.2 or later not found!

So I went and dug up a Python 2.3.4 rpm, only to have Mandrake inform me that it was already installed.

Guest · Post by **Guest** » Wed Oct 20, 2004 7:19 pm

SpaceCowboy wrote:...
checking for python... configure: error: *** Python version 2.2 or later not found!

So I went and dug up a Python 2.3.4 rpm, only to have Mandrake inform me that it was already installed.

I think you need the developer version of the python packages ("python-devel" or somehting like that) which includes the header (and sources?)

SpaceCowboy · Post by **SpaceCowboy** » Thu Oct 21, 2004 1:20 am

Hey guys, thanks for helping. Getting the python-devel did fix it.

Of course, now it's saying I don't have any OpenGLU library. I've downloaded and installed just about every Mesa library I could find, but I still get:

*** Hmm, you don't seem to have OpenGLU libraries installed in the standard
*** location (/usr/lib). I'll check in /usr/X11R6/lib, since
*** many distributions (incorrectly) put OpenGLU libs there.
checking for GLU library... no
checking for GLU library (with pthreads)... no
checking for MesaGLU library... no
checking for MesaGLU library (with pthreads)... no
checking for openglu32 library... no
checking for openglu32 library (with pthreads)... no
configure: error: Cannot find GLU library

Post by **hellcatv** » Thu Oct 21, 2004 4:24 am

move libGLU.so into /usr/lib instead of /usr/X11R6/lib

see if that fixes it

also make sure that /usr/include/GL/ has a copy of glu.h

SpaceCowboy · Post by **SpaceCowboy** » Fri Oct 22, 2004 4:41 am

So I just discovered that most of this stuff is covered in the FAQ. Sorry to be that guy who doesn't 'rtf'.

Anyways, I copied libGLU.so into the right directory and that satisfied configure.

Now though, I'm trying to run make, and get this error:

(a bunch of 'undefined reference to 'FireKeyboard'')
gfx/libgfx.a(cockpit.o)(.text+0x12061): In function `GameCockpit::SetCommAnimation(Animation*)':
: undefined reference to `VDU::SetCommAnimation(Animation*)'
gfx/libgfx.a(cockpit_xml.o)(.text+0x1158): In function `GameCockpit::beginElement(std::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, XMLSupport::AttributeList const&)':
: undefined reference to `parse_vdu_type(char const*)'
gfx/libgfx.a(cockpit_xml.o)(.text+0x1377): In function `GameCockpit::beginElement(std::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, XMLSupport::AttributeList const&)':
: undefined reference to `VDU::VDU(char const*, TextPlane*, unsigned short, short, short, float*, float*)'
collect2: ld returned 1 exit status
make[3]: *** [vegastrike] Error 1
make[3]: Leaving directory `/home/thomas/vegastrike/vegastrike/src'
make[2]: *** [all-recursive] Error 1
make[2]: Leaving directory `/home/thomas/vegastrike/vegastrike/src'
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory `/home/thomas/vegastrike/vegastrike'
make: *** [all] Error 2

I made sure that a current glext.h is in /usr/include/GL/ (which seems to really be /usr/X11R6/include/GL/), and whether or not I put --enable-boost-128 on ./configure doesn't matter.

Ari Rahikkala · Post by **Ari Rahikkala** » Fri Oct 22, 2004 7:05 am

Might be broken dependencies. Try make clean; make.

SpaceCowboy · Post by **SpaceCowboy** » Fri Oct 22, 2004 10:26 pm

Yep. Had to run make clean in the boost directories as well, but that fixed it.

Now I've been getting lots of GTk errors (EDIT: it can't find gtk-config or gtk.h). I understand that using GTk isn't necessary, but is there some GTk package that I'm overlooking (I got all the python-related GTk stuff I could see) that woudl fix this?

Ari Rahikkala · Post by **Ari Rahikkala** » Sat Oct 23, 2004 7:06 am

It should be something as simple as gtk-devel. gtk.h isn't used by Python, it's used by C.

SpaceCowboy · Post by **SpaceCowboy** » Mon Oct 25, 2004 5:27 am

Ah, Thanks Ari. Got it all compiled this afternoon.

But I think this little odyssey still has one more act to go. When I load vslauncher and try to begin a new game, the splash screen appears and the game begins to load, but seg faults most of the way through.

Specifically, it says 'Loading: generic.png' in the lower left corner of the splash screen, and then dumps me back to the console. The console has this to say:

Loading completed, now network init
Writing Save Game test2explore_universe_difficultyHi helper play 0
HereHi helper play 0
HereHi helper play 0
yzwFatal signal: Segmentation Fault (SDL Parachute Deployed)

I've been looking through the forums to see if anyone else has had this, but no luck. So any help would be appreciated.

Chunk · Post by **Chunk** » Mon Oct 25, 2004 5:43 am

do you have a joystick/pad plugged in? I get the same thing when I have my gamepad plugged in but the game runs fine when it's not (i've posted in the bugs forum about this too).

SpaceCowboy · Post by **SpaceCowboy** » Mon Oct 25, 2004 6:08 am

Nope, don't have anything plugged in and earlier in the start-up it recognizes that there isn't a joystick.

(Incidentally, how is it flying without a joystick? Do you use the keyboard?)

Post by **dandandaman** » Mon Oct 25, 2004 6:27 am

Try changing to the data dir before running it

Dan.a

Chunk · Post by **Chunk** » Mon Oct 25, 2004 6:47 am

SpaceCowboy wrote:Nope, don't have anything plugged in and earlier in the start-up it recognizes that there isn't a joystick.

(Incidentally, how is it flying without a joystick? Do you use the keyboard?)

yeah and the mouse in fights (warp mouse enabled). It sucks....

SpaceCowboy · Post by **SpaceCowboy** » Tue Oct 26, 2004 12:43 am

dandandaman wrote:Try changing to the data dir before running it

Dan.a

Good idea. That certainly helps. Now VS seems to load completely, but I only get a split-second of game (with sound) before it dumps me out again.

The console says:

yzw[hash.c:395] failed read
...<snip (more [hash.c:395] failed read) >...
[hash.c:395] failed read
Hereinit playerdat
done playerdat
launching bases for Crucible/Cephid_17
finding nonloaded quest
quest_introduction
finding quest
quest_introduction
nonpfact
resetting sigdist=20000.000000 detdist=40000.000000
dot -0.999996Processing News
xyzwxyzwxyzwCustom condition code, returning: False
Rejecting subnode with text=The mission robot.
Custom condition code, returning: 0
Rejecting subnode with text=The receiving merchant.
Checking 'IntroJP : 1'
Returning: 0
Rejecting subnode with text=knownjenek
Returning subnode with text=unknownjenek
Checking if can draw
['#\nimport quest_intro2\nresult = (quest_intro2.isAt("Crucible/Cephid_17/Dakar") or quest_intro2.isAt("Crucible/Enyo"))']
len 1
Custom condition code, returning: 0
uhoh
Fixer can draw? -- 0
Checking if can draw
['#\nimport quest_intro2\nresult = (quest_intro2.isAt("Crucible/Cephid_17/Dakar") or quest_intro2.isAt("Crucible/Enyo"))']
len 1
Custom condition code, returning: 0
uhoh
Checking if can draw
['#\nimport quest_intro\nresult = quest_intro.okayDrawJenek()']
len 1
Custom condition code, returning: True
okay
Fixer can draw? -- 1
Checking if can draw
['#\nimport quest_intro\nresult = quest_intro.okayDrawJenek()']
len 1
Custom condition code, returning: True
okay
xyzwxyzwFatal signal: Segmentation Fault (SDL Parachute Deployed)

SpaceCowboy · Post by **SpaceCowboy** » Tue Oct 26, 2004 6:53 pm

I ran backtrace full on it, and got out:

Code: Select all

Program received signal SIGSEGV, Segmentation fault.
0x401e35b9 in _tnl_playback_vertex_list () from /usr/X11R6/lib/libGL.so.1
(gdb) backtrace full
#0  0x401e35b9 in _tnl_playback_vertex_list () from /usr/X11R6/lib/libGL.so.1
No symbol table info available.
#1  0x401eb022 in _tnl_playback_vertex_list () from /usr/X11R6/lib/libGL.so.1
No symbol table info available.
#2  0x401eae4a in _tnl_playback_vertex_list () from /usr/X11R6/lib/libGL.so.1
No symbol table info available.
#3  0x401e28c5 in _tnl_playback_vertex_list () from /usr/X11R6/lib/libGL.so.1
No symbol table info available.
#4  0x401d7d9a in _tnl_run_pipeline () from /usr/X11R6/lib/libGL.so.1
No symbol table info available.
#5  0x401e2071 in _tnl_playback_vertex_list () from /usr/X11R6/lib/libGL.so.1
No symbol table info available.
#6  0x40123c86 in _mesa_compile_error () from /usr/X11R6/lib/libGL.so.1
No symbol table info available.
#7  0x40126442 in _mesa_CallList () from /usr/X11R6/lib/libGL.so.1
No symbol table info available.
#8  0x08721ebd in GFXCallList ()
---Type <return> to continue, or q <return> to quit---
No symbol table info available.
#9  0x083b453a in GFXVertexList::Draw ()
No symbol table info available.
#10 0x083b450e in GFXVertexList::Draw ()
No symbol table info available.
#11 0x083b44c1 in GFXVertexList::DrawOnce ()
No symbol table info available.
#12 0x0869b8bc in Mesh::DrawNow ()
No symbol table info available.
#13 0x085c00c3 in GameUnit<Unit>::DrawNow ()
No symbol table info available.
#14 0x08551e69 in BaseInterface::Room::BaseShip::Draw ()
No symbol table info available.
#15 0x08551f6b in BaseInterface::Room::Draw ()
No symbol table info available.
#16 0x08556f1e in BaseInterface::Draw ()
No symbol table info available.
---Type <return> to continue, or q <return> to quit---
#17 0x08552c7b in RefreshGUI ()
No symbol table info available.
#18 0x08552d0b in base_main_loop ()
No symbol table info available.
#19 0x0872a5e7 in winsys_process_events ()
No symbol table info available.
#20 0x08712ed1 in GFXLoop ()
No symbol table info available.
#21 0x083cce07 in GameUniverse::Loop ()
No symbol table info available.
#22 0x083deca6 in main ()
No symbol table info available.

Continuing in gdb, I get:

Code: Select all

Continuing.
Fatal signal: Segmentation Fault (SDL Parachute Deployed)

Program exited with code 0365.

Post by **hellcatv** » Wed Oct 27, 2004 9:13 am

sounds like a mesa bug for sure

try changing vegastrike.config to use compiled vertex arrays instead

SpaceCowboy · Post by **SpaceCowboy** » Fri Nov 05, 2004 5:15 am

Actually, I was poking around through the Linux Gamer's FAQ, and it recommended editing the /etc/ld.so.conf and adding in

/usr/local/lib
/usr/lib
/lib
/usr/X11R6/lib

Which makes VegaStrike work!

(Well, almost. But I started a topic in User Help for that one).