Debian 9.3 5.1 and svn compile issues
-
- Bounty Hunter
- Posts: 201
- Joined: Wed Dec 01, 2004 8:06 pm
- Location: not where i want to be
Debian 9.3 5.1 and svn compile issues
Well after taking a wrong turn into an unstable jump point, I finally made it back to civilization. (more on this later)
After fially getting back online in something I can compile I I tried compiling both 0.51r1 and svn and failed
I did make progress using -ansi cflag and cxxflags
boost ,ogre, ffmpeg gave me problems on both versions , I could only get svn to compile after disabling ffmpeg via ccmake this is not an op-tion on 0.51
btw I disabled boost and ogre as well
I am now checkingout data from svn, and have not yet taken time to investigate further.
What I do know is 1 ffmpeg changed some function labels
2 boost and ogre both made changes to there apis as well
alot of warning about using "auto_pointer" which appears to be getting phased out with implemening of ISO standards
I am also setting a 64 bit box was on 32 bit before
more to follow
After fially getting back online in something I can compile I I tried compiling both 0.51r1 and svn and failed
I did make progress using -ansi cflag and cxxflags
boost ,ogre, ffmpeg gave me problems on both versions , I could only get svn to compile after disabling ffmpeg via ccmake this is not an op-tion on 0.51
btw I disabled boost and ogre as well
I am now checkingout data from svn, and have not yet taken time to investigate further.
What I do know is 1 ffmpeg changed some function labels
2 boost and ogre both made changes to there apis as well
alot of warning about using "auto_pointer" which appears to be getting phased out with implemening of ISO standards
I am also setting a 64 bit box was on 32 bit before
more to follow
if current_situation=fact and if current_situation=faith
then current_forcast = excitement
then current_forcast = excitement
-
- The Shepherd
- Posts: 5841
- Joined: Fri May 13, 2005 8:37 pm
- Location: Ottawa
- Contact:
Re: Debian 9.3 5.1 and svn compile issues
Glad to see your post a bit of a blast from the past and yeah quite a few of the supporting libraries have changed their API's there is no need for ffmpeg as it is not used anywhere in the code.Ogre is only used for mesher our cmd-line utility for converting .obj format meshes to our internal binary mesh format or Ogre3D's .mesh format so here you have two options update mesher's code to the new API or drop the .mesh conversion functionality completely that is you intend to add new models to your local copy The boost library is a bit more problematic as it's main utility was to decrease execution time of a lot of the kinetic equations as we use 4 order Runge Kutta method the solve those differentials.Someone else asked these questions a while back I would have a look here in Compiling Problems for the relevant threads.
Enjoy the Choice
Enjoy the Choice
my box::HP Envy i5-6400 @2Q70GHzx4 8 Gb ram/1 Tb(Win10 64)/3 Tb Mint 19.2/GTX745 4Gb acer S243HL K222HQL
Q8200/Asus P5QDLX/8 Gb ram/WD 2Tb 2-500 G HD/GF GT640 2Gb Mint 17.3 64 bit Win 10 32 bit acer and Lenovo ideapad 320-15ARB Win 10/Mint 19.2
Q8200/Asus P5QDLX/8 Gb ram/WD 2Tb 2-500 G HD/GF GT640 2Gb Mint 17.3 64 bit Win 10 32 bit acer and Lenovo ideapad 320-15ARB Win 10/Mint 19.2
-
- Bounty Hunter
- Posts: 201
- Joined: Wed Dec 01, 2004 8:06 pm
- Location: not where i want to be
Re: Debian 9.3 5.1 and svn compile issues
Yeah Yeah BOOST This BOOST That
I suppose it mean s that some of the of boost calls need to be changed ,even with -ansi I cant get 64 bit to compile.I got it to compile on 32 bit ver cmake has some options in th 32 bit vers but seem to be absent on 64 bit. I may of did something on one that disabled boost on then 32 but not 64
second thing I don't know " 4 order Runge Kutta method" either
but I do know how to make patches and basic grep
I have a tar'd svn and 5.1r1
And if I really screw up I may be enshrined upon the Luddite wall of shame
( and no it's not a good thing)
this looks like a job for both doxygen and valgrind
I am wondering how much assembly is used
How about some goals
1 get boost ver 1.62 to compile (for me)
2 get boost code to compilw without warnings
3 work on older boost versions
In closing I still getting set hereIt will take me some time to get up to speed
good news is I will have some time in not too long I can look at it
Of course no guaranties
I suppose it mean s that some of the of boost calls need to be changed ,even with -ansi I cant get 64 bit to compile.I got it to compile on 32 bit ver cmake has some options in th 32 bit vers but seem to be absent on 64 bit. I may of did something on one that disabled boost on then 32 but not 64
second thing I don't know " 4 order Runge Kutta method" either
but I do know how to make patches and basic grep
I have a tar'd svn and 5.1r1
And if I really screw up I may be enshrined upon the Luddite wall of shame
( and no it's not a good thing)
this looks like a job for both doxygen and valgrind
I am wondering how much assembly is used
How about some goals
1 get boost ver 1.62 to compile (for me)
2 get boost code to compilw without warnings
3 work on older boost versions
In closing I still getting set hereIt will take me some time to get up to speed
good news is I will have some time in not too long I can look at it
Of course no guaranties
if current_situation=fact and if current_situation=faith
then current_forcast = excitement
then current_forcast = excitement
-
- Bounty Hunter
- Posts: 201
- Joined: Wed Dec 01, 2004 8:06 pm
- Location: not where i want to be
Re: Debian 9.3 5.1 and svn compile issues
well I got my soures wrong
0.5.1r1 has boost compile issues
svn has iso related messages in file vegastrike/src/gfx/aligned.h
I put these in 0.5.1 trying get furteher and previously used them in svn build "-ansi -fpermissive " almost got to compile
After removing,
now I got svn to compile and link
I dont know if I met some undisclosed dependencies or what ( I did alot of things trying to get it to work) .Sooner or later I will try it with virgin svn copy to see if I can do it again
It even starts up
0.5.1r1 has boost compile issues
svn has iso related messages in file vegastrike/src/gfx/aligned.h
I put these in 0.5.1 trying get furteher and previously used them in svn build "-ansi -fpermissive " almost got to compile
After removing,
now I got svn to compile and link
I dont know if I met some undisclosed dependencies or what ( I did alot of things trying to get it to work) .Sooner or later I will try it with virgin svn copy to see if I can do it again
It even starts up
if current_situation=fact and if current_situation=faith
then current_forcast = excitement
then current_forcast = excitement
-
- Bounty Hunter
- Posts: 201
- Joined: Wed Dec 01, 2004 8:06 pm
- Location: not where i want to be
Re: Debian 9.3 5.1 and svn compile issues
Progress is made
Even repeatable
Once all dependencies are met
all changes were made using ccmake
python in beginning of ccmake, python version is found correctly
but further down python ver must be entered ie 2.7 ( it matters)
use system boost must be on.
I also removed the vegastrike/boost contents as well to be certain
disable ffmpeg and disable ogre must be on
compiles and almost runs
problem with planets and ships disappearing with new(er) 3d card radeon r5 220
orig Intell shows objects but is sloww as hell (intel gma 3100)
I am still working on getting graphics up to par
because of it being svn ,vssetup is having an identity crisis not knowning it's version
if any cmake generated files are useful let me know
Even repeatable
Once all dependencies are met
all changes were made using ccmake
python in beginning of ccmake, python version is found correctly
but further down python ver must be entered ie 2.7 ( it matters)
use system boost must be on.
I also removed the vegastrike/boost contents as well to be certain
disable ffmpeg and disable ogre must be on
compiles and almost runs
problem with planets and ships disappearing with new(er) 3d card radeon r5 220
orig Intell shows objects but is sloww as hell (intel gma 3100)
I am still working on getting graphics up to par
because of it being svn ,vssetup is having an identity crisis not knowning it's version
if any cmake generated files are useful let me know
if current_situation=fact and if current_situation=faith
then current_forcast = excitement
then current_forcast = excitement
-
- The Shepherd
- Posts: 5841
- Joined: Fri May 13, 2005 8:37 pm
- Location: Ottawa
- Contact:
Re: Debian 9.3 5.1 and svn compile issues
Those disappearing ships on your newer GPU are most likely from depreciated OpenGL calls lucky you don't have OSX system Apple has announced that they dropping support of OpenGL completely
Enjoy the Choice
Enjoy the Choice
my box::HP Envy i5-6400 @2Q70GHzx4 8 Gb ram/1 Tb(Win10 64)/3 Tb Mint 19.2/GTX745 4Gb acer S243HL K222HQL
Q8200/Asus P5QDLX/8 Gb ram/WD 2Tb 2-500 G HD/GF GT640 2Gb Mint 17.3 64 bit Win 10 32 bit acer and Lenovo ideapad 320-15ARB Win 10/Mint 19.2
Q8200/Asus P5QDLX/8 Gb ram/WD 2Tb 2-500 G HD/GF GT640 2Gb Mint 17.3 64 bit Win 10 32 bit acer and Lenovo ideapad 320-15ARB Win 10/Mint 19.2
-
- Bounty Hunter
- Posts: 201
- Joined: Wed Dec 01, 2004 8:06 pm
- Location: not where i want to be
Re: Debian 9.3 5.1 and svn compile issues
The problem is that I remember this problem being corrected with vssetup (long ago in a virtual galaxy far ...).
It would appear any information has however redacted
It would appear any information has however redacted
if current_situation=fact and if current_situation=faith
then current_forcast = excitement
then current_forcast = excitement
-
- The Shepherd
- Posts: 5841
- Joined: Fri May 13, 2005 8:37 pm
- Location: Ottawa
- Contact:
Re: Debian 9.3 5.1 and svn compile issues
All that vssetup does is edit the config file's XML tags which was created on the fist run in the .vegastrike folder don't know if it fallows the current convention of putting that folder under .config so just check the hidden files BTW there is info on config file editing on the wkii.
Enjoy the Choice
Enjoy the Choice
my box::HP Envy i5-6400 @2Q70GHzx4 8 Gb ram/1 Tb(Win10 64)/3 Tb Mint 19.2/GTX745 4Gb acer S243HL K222HQL
Q8200/Asus P5QDLX/8 Gb ram/WD 2Tb 2-500 G HD/GF GT640 2Gb Mint 17.3 64 bit Win 10 32 bit acer and Lenovo ideapad 320-15ARB Win 10/Mint 19.2
Q8200/Asus P5QDLX/8 Gb ram/WD 2Tb 2-500 G HD/GF GT640 2Gb Mint 17.3 64 bit Win 10 32 bit acer and Lenovo ideapad 320-15ARB Win 10/Mint 19.2
-
- Bounty Hunter
- Posts: 201
- Joined: Wed Dec 01, 2004 8:06 pm
- Location: not where i want to be
Re: Debian 9.3 5.1 and svn compile issues
running from .../vs/build (dir made at top of vegastrike svn source to compile in) (( for the kiddies)) . I get the following repeated errors
looki9ng aroung I found r600 located under
Code: Select all
EE ../../../../../src/gallium/drivers/r600/r600_shader.c:3663 r600_shader_from_tgsi - GPR limit exceeded - shader requires 125 registers
EE ../../../../../src/gallium/drivers/r600/r600_shader.c:183 r600_pipe_shader_create - translation from TGSI failed !
EE ../../../../../src/gallium/drivers/r600/r600_state_common.c:799 r600_shader_select - Failed to build shader variant (type=0) -12
I suspect some of your "depreciated OpenGL calls" connect here. being I haven't even compiled debug versions I don't know.locate r600
/usr/include/libdrm/r600_pci_ids.h
/usr/lib/x86_64-linux-gnu/dri/r600_dri.so
/usr/lib/x86_64-linux-gnu/dri/r600_drv_video.so
/usr/lib/x86_64-linux-gnu/vdpau/libvdpau_r600.so
/usr/lib/x86_64-linux-gnu/vdpau/libvdpau_r600.so.1
if current_situation=fact and if current_situation=faith
then current_forcast = excitement
then current_forcast = excitement
-
- Bounty Hunter
- Posts: 201
- Joined: Wed Dec 01, 2004 8:06 pm
- Location: not where i want to be
Re: Debian 9.3 5.1 and svn compile issues
looks like somebody should study
http://www.opengl-tutorial.org/beginners-tutorials
( smashing oneself in the luddite tradition of slamming a sharp stone to the side of the head ) aka Thudding which loosely translates to mean thinking combined with luddite., seems to be more fun.
search for depreciated opengl calls in vs
and after many trials have something that once again works
Is this what needs to be done?
http://www.opengl-tutorial.org/beginners-tutorials
( smashing oneself in the luddite tradition of slamming a sharp stone to the side of the head ) aka Thudding which loosely translates to mean thinking combined with luddite., seems to be more fun.
search for depreciated opengl calls in vs
and after many trials have something that once again works
Is this what needs to be done?
if current_situation=fact and if current_situation=faith
then current_forcast = excitement
then current_forcast = excitement
-
- Bounty Hunter
- Posts: 201
- Joined: Wed Dec 01, 2004 8:06 pm
- Location: not where i want to be
Re: Debian 9.3 5.1 and svn compile issues
I lesson 2 I discovered a problem with shaders not ready
There is a rather long proceedure ot init them
It is nice to find the same problem elsewhere
There is a rather long proceedure ot init them
It is nice to find the same problem elsewhere
if current_situation=fact and if current_situation=faith
then current_forcast = excitement
then current_forcast = excitement
-
- Bounty Hunter
- Posts: 201
- Joined: Wed Dec 01, 2004 8:06 pm
- Location: not where i want to be
Re: Debian 9.3 5.1 and svn compile issues
Well Ive got a copy of makes output (1mb)
Ive also ran doxygen and valgrind --leak-check=yes
leakcheck found lots of leaks
the gcc compiler puts out lots of warnings about " warning: ‘template<class> class std::auto_ptr’ is deprecated [-Wdeprecated-declarations]"
at least a both when trying to find gl related messages. supposedly it is a fairly easy fix. though if backward compatibility is desired a bit more coding is needed
opengl calls will need shaders initialization to be gone over as well
That is all I have for now
Ive also ran doxygen and valgrind --leak-check=yes
leakcheck found lots of leaks
the gcc compiler puts out lots of warnings about " warning: ‘template<class> class std::auto_ptr’ is deprecated [-Wdeprecated-declarations]"
at least a both when trying to find gl related messages. supposedly it is a fairly easy fix. though if backward compatibility is desired a bit more coding is needed
opengl calls will need shaders initialization to be gone over as well
That is all I have for now
if current_situation=fact and if current_situation=faith
then current_forcast = excitement
then current_forcast = excitement
-
- Bounty Hunter
- Posts: 201
- Joined: Wed Dec 01, 2004 8:06 pm
- Location: not where i want to be
Re: Debian 9.3 5.1 and svn compile issues
I found out running uninstalled the "./vegastrik " directory for 0.5.1rc1 is not compatables with svn
so I have two copiers of "./vegastrike" under different names and copy them over for which version I am running
I figured out gettings shaders to work Use vssetup, i used "simpilest shader" to get my radeon to display
then there are the bugs
now if i upgrade the reactor then I can't jump
If I try to pause my system goes to sleep
I don't yet know if I caused any problems while getting rid of compiler warnings ( it's my copy to abuse if I choose)
so no diff for svn yet
so I have two copiers of "./vegastrike" under different names and copy them over for which version I am running
I figured out gettings shaders to work Use vssetup, i used "simpilest shader" to get my radeon to display
then there are the bugs
now if i upgrade the reactor then I can't jump
If I try to pause my system goes to sleep
I don't yet know if I caused any problems while getting rid of compiler warnings ( it's my copy to abuse if I choose)
so no diff for svn yet
if current_situation=fact and if current_situation=faith
then current_forcast = excitement
then current_forcast = excitement
-
- Bounty Hunter
- Posts: 201
- Joined: Wed Dec 01, 2004 8:06 pm
- Location: not where i want to be
Re: Debian 9.3 5.1 and svn compile issues
Both 0.5.1rc1 and svn have crashing issues
jump point error due to operator ( needed bigger spec capacitor) after upgrading other ship systems , error in info tap will show error about needing bigger caps (FTL caps)
I suspect crashes may be due to boost_python needing python 3
most crashes locked up system requiring cycling power, one time vegastrike quit to system And I saw python error mmsg about someting being called already.
Bottom line it can be played but do not be running any other program you may want/need. also save when possible.
I am not even thinking about network yet
jump point error due to operator ( needed bigger spec capacitor) after upgrading other ship systems , error in info tap will show error about needing bigger caps (FTL caps)
I suspect crashes may be due to boost_python needing python 3
most crashes locked up system requiring cycling power, one time vegastrike quit to system And I saw python error mmsg about someting being called already.
Bottom line it can be played but do not be running any other program you may want/need. also save when possible.
I am not even thinking about network yet
if current_situation=fact and if current_situation=faith
then current_forcast = excitement
then current_forcast = excitement
-
- Bounty Hunter
- Posts: 201
- Joined: Wed Dec 01, 2004 8:06 pm
- Location: not where i want to be
Re: Debian 9.3 5.1 and svn compile issues
More work today
after having more experience compiling/gunning now @ branches
Previous warnings apply but for who ever want to try the svn here is how I got compiling sucess
!) in a terminal make a empty directory and change into it
2) using " svn checkout svn://svn.code.sf.net/p/vegastrike/code/trunk/data vs_svn/data" get the data filed
3) using "svn checkout svn://svn.code.sf.net/p/vegastrike/code/trunk/vegastrike vs_svn/vs_src" to get the source code
Not the above example I choose "vs_svn" and "vs_src" but you can use other names ( just remember what you choose
4"cd into source directory
5 mkdir "build" (again you can use other names) and enter in it
6) issue "ccmake .." ( debian pkg is cmake-curses-gui)
first time you will see an empty tree if you hit "c" for configure it will establish needed files and issue any warnings about missing dependancies
youe nee to hit "e" then you should see various options, satisfy what you missing from "c" command
NOTE I have been able to compile with the following missing/notfound messages
chances are you need to repeat this until "g" appears when it does issue "g" if all is gooc ccmake will exit
9) Finally issue "make" if all is well you should suceed in conpiling vegasrtike
10) at this point you have two choices to " make install" or run what you have
11) if you have a "./vegastrike" directory in your home folder move it elsewhere or just delete it
12)back in your build dir copy "vegastrike, ./setup/vssetup into "data" folder
12 optionaly copy "./objconv/asteroidgen, ./objconv/mesh_tool, ./objconv/replace, ./objconv/trisort" into data dir as well
13) enter the data dir and first run "./vssetup" ( I suggest using simplest shader" make changes for best result try simplest settings first you can change them any time with vssetup
14) take a deep breath and issue "./vegastrike"
***WARNING***
please read previous post about system crash/hanging this applies to all three " 0.5.1rc1. py3, and svn
If anyone wants to set build type to debug to try to help troubles shoot ( I am starting to do so) run "gdb ./vegastrike" then"t" to run have at it
after having more experience compiling/gunning now @ branches
Previous warnings apply but for who ever want to try the svn here is how I got compiling sucess
!) in a terminal make a empty directory and change into it
2) using " svn checkout svn://svn.code.sf.net/p/vegastrike/code/trunk/data vs_svn/data" get the data filed
3) using "svn checkout svn://svn.code.sf.net/p/vegastrike/code/trunk/vegastrike vs_svn/vs_src" to get the source code
Not the above example I choose "vs_svn" and "vs_src" but you can use other names ( just remember what you choose
4"cd into source directory
5 mkdir "build" (again you can use other names) and enter in it
6) issue "ccmake .." ( debian pkg is cmake-curses-gui)
first time you will see an empty tree if you hit "c" for configure it will establish needed files and issue any warnings about missing dependancies
youe nee to hit "e" then you should see various options, satisfy what you missing from "c" command
NOTE I have been able to compile with the following missing/notfound messages
- Boost_DIR Boost_DIR NOT-FOUND
GTK2_GTKGL_INCLUDE_DIR GTK2_GTKGL_INCLUDE_DIR-NOTFOUND
GTK2_GTKGL_LIBRARY GTK2_GTKGL_LIBRARY-NOTFOUND
OPENGL_xmesa_INCLUDE_DIR OPENGL_xmesa_INCLUDE_DIR-NOTFOUND
PYTHON_LIBRARY_DEBUG PYTHON_LIBRARY_DEBUG-NOTFOUND
- DISABLE_FFMPEG on
DISABLE_OGRE on
USE_SYSTEM_BOOST on
VS_PYTHON_VERSION 2.7 (was 2, it does make a difference)
chances are you need to repeat this until "g" appears when it does issue "g" if all is gooc ccmake will exit
9) Finally issue "make" if all is well you should suceed in conpiling vegasrtike
10) at this point you have two choices to " make install" or run what you have
11) if you have a "./vegastrike" directory in your home folder move it elsewhere or just delete it
12)back in your build dir copy "vegastrike, ./setup/vssetup into "data" folder
12 optionaly copy "./objconv/asteroidgen, ./objconv/mesh_tool, ./objconv/replace, ./objconv/trisort" into data dir as well
13) enter the data dir and first run "./vssetup" ( I suggest using simplest shader" make changes for best result try simplest settings first you can change them any time with vssetup
14) take a deep breath and issue "./vegastrike"
***WARNING***
please read previous post about system crash/hanging this applies to all three " 0.5.1rc1. py3, and svn
If anyone wants to set build type to debug to try to help troubles shoot ( I am starting to do so) run "gdb ./vegastrike" then"t" to run have at it
if current_situation=fact and if current_situation=faith
then current_forcast = excitement
then current_forcast = excitement