The stars that give you some idea of how fast and what direction you are turning don't work anymore. It seems that this commit broke it:

commit 3fe4e66

separate entity camera transform from translation so model to world to camera
translation can happen ouside of the rotation matrix.  having translation in
the transform matrix was causing rotation aliasing at the far ends of the universe.

A couple of nice-to-have's in the future for the Main screens

Magnification:
Be able to magnify the center of the screen to see what's going on a little farther out. More of a Star Trek-ism more than anything.

Sensors:
for the Captain to plan out his next course, having the sensors on main screen could be an option. This could be a subset of what science sees.

The Help Screen for the Main Screen states that "W" toggles weapons view. Currently it controls the pitch of the ship (WASD controls). This should probably be removed.

In snis_client.c

static char *help_text[] = {

        /* Main screen help text */
        "MAIN SCREEN\n\n"
        "  CONTROLS\n\n"
        "  * USE ARROW KEYS TO TURN SHIP\n\n"
        "  * W KEY TOGGLES BETWEEN WEAPONS VIEW\n"
        "    AND MAIN VIEW\n\n"

On respawn, ships may randomly choose to join a fleet. But there's no way for a ship to leave a fleet (other than dying and respawning). As a consquence, fleets tend to grow indefinitely in size, and since joining a fleet forces the faction of the member to be the same as the faction of the leader, what happens, is all factions but one go extinct as a massive fleet crushes all before it.

Needs balancing.

Rogue leaderless fleets occasionally wander off into the hinterlands.

There are lots of places in snis_client.c where we use pseudo random numbers to generate object properties. For instance, the spin on asteroids, choosing of asteroid models, the trajectories of sparks, the configuration of nebulas, etc. If the clients do not happen to use the same seed and same random number algorithm, or if multiple threads in the client use the single random number generated (glibc's rand(), at the bottom of it all), then the different clients can display different things. For example, the same asteroid might use one model on one client, and another on a different client, and the asteroid might spin differently, the nebulas might appear different from one client to the next, etc. Actually, it so happens that the nebulas are generated early enough that it seems the clients do make the same nebulas (have not tried on 32 and 64 bit clients to see if those are the same), but relying on this behavior seems wrong.

While in a nebula, weapons and helm screens are scrambled. Enemies will still head toward you and attack, apparently with their systems unaffected.
Make enemies not be able to attack with such ease or at least misfire a few shots.
Make it so you can hide in nebulas, in-case you are being attacked. Also hide enemies in nebulas.

The lasers look cool, but laser targeting is no fun -- less fun than the old system with manual aiming of the old lasers (that were just photon torpedoes that were weaker and faster). Make it funner.

Had some reports of game servers not showing up in the lobby.

It may be related to running the game inside VMs, but not sure.

After recently checking out source code I am now getting an error when running make. /lib/x86_64-linux-gnu/libpng12.so.0 does exist. This has worked on previous revisions.

LINK snis_server
/usr/bin/ld: snis_graph.o: undefined reference to symbol 'png_set_sig_bytes@@PNG12_0'
/usr/bin/ld: note: 'png_set_sig_bytes@@PNG12_0' is defined in DSO /lib/x86_64-linux-gnu/libpng12.so.0 so try adding it to the linker command line
/lib/x86_64-linux-gnu/libpng12.so.0: could not read symbols: Invalid operation
collect2: error: ld returned 1 exit status
make: *** [snis_client] Error 1

I'm planning on building some custom keyboards for the game and I'd like to be able to control everything in game from the keyboard without having to use a mouse. (These will get double use with Artemis SBS). A config file would be nice to be able to remap keys for future upgrades to code and to my custom keyboards.

I doupt this game was thought for Windows but I wonder if someone is going to make it compatible.
Please tell me if there's a need for an error log, it's huge.

I built SNIS on Mageia Cauldron (development version of the upcoming Mageia
5). It was not that straight-forward since I had to package OpenSCAD, and
to do so also package CGAL and OpenCSG, but either way it's good that
OpenSCAD is now available in the distro :-)

I get a segfault when running snis_client with my system locale
(fr_FR.UTF-8): http://pastebin.com/Y84xFhE3
When running SNIS with an English locale (LC_ALL=C ./bin/snis_client), it
seems to work as expected.

My guess would be that the issue comes from the decimal delimiter, which is
a dot in English but a comma in French (and other locales too, such as
German or Spanish). I've already experienced quite a few games crashing
because of such locale-related issues (mainly indie games on Steam though).

This would explains errors such as:

Error reading 'orientation' from file './share/snis/materials/nebula0.mat'

If the system awaits commas and finds only dots, it doesn't understand the
numbers. I'm not sure how you could best fix this on the programming side
though, I don't have much developer skills.

A user was testing out snis on a computer with i3 window manager and two 1920x1080 monitors, which lead to ui placement issues. h ignored them because of the unusual environment, but then a community member in Steam Linux User Group also had issues.

This other guy was kind enough to take a couple screenshots:

He said "line 225 and 226 of snis_font.c .. I added in * 0.7" and that made it so the fonts did not overlap anymore.

It might suitable, for now, to have a way to set the resolution. Perhaps an option --resolution=XxY

Ran into an issue this weekend with a 6 computer game. For some reason the ship was spinning in circles. I tried closing out of the game on the Nav client computer as well as the Main screen client, but, it continued.

I've seen this issue off and on previously but it usually get corrected after a commit, or so I thought.

This was on 32bit Debian Stable but have seen it on Ubuntu off and on.

I tried to compile the latest git on ArchLinux x86_64.
There's a few #include issues regarding "gtk/gtkgl.h" but my programming knowledge is insufficient to fix the problem and propose a patch.

Minor bug with using a cubemap ( NO "-s" or "-S" ) in gaseous-giganticus.c

at the end will be a link to a zip file with examples

a cube map is normally h =3 w=4 proportion

reading in the width is FINE however the tile height that is not correct

the tile height is being inputed as 3/4 of the 1024 rows then outputing a 1024 image ( of the 768 rows read)

so far i have NOT been able to find the loop that is causing this read error

included in the zip is a standard test image i use simple cylindrical 10 degree lines
a cube map of it
and a "work around " remapped cube map that DOSE work
but that is a work around only

for a correct cube map --- not outputting correctly
./gaseous-giganticus -c 1 --noise-scale 0.1 --velocity-factor 1 -i ./lines10Cube.png --bands 1

a "fudged cubemap that will output correctly "lines10Cube1.png"
./gaseous-giganticus -c 1 --noise-scale 0.1 --velocity-factor 1 -i ./lines10Cube1.png --bands 1

" Test.zip"
https://drive.google.com/file/d/0B6ZYAd08tZL-UnREOGtRMy04UGM/view?usp=sharing

When trying to Dock with stations, the shields must be lowered.
After lowering the shields to what at least looks like zero (both power and shield strength), sometimes it takes 2 requests from Comms to be able to dock. It's hard to tell if it's the shields still having some power left or just a Comms bug/issue.

A Shields Up/Down button would make it easier to turn on & off the power and to tell the current status of the shields. This would help coincide with the request of the Captain for Shields Up or Down and make it easier for the engineer.

One seems to work ok, more than one and none work.

I think the game assets should be completely decoupled from the codebase. Right now it's a bit of a downer to see all the openscad files being re-compiled every time I do amake clean.

Ideally, SNIS should be able to switch between sets of game assets (models, textures, shaders, materials, sounds, commodities, factions) easily. The current way of overriding the asset directory looks good enough for me.

I suggest following the example of Artemis with VesselData.xml - create some kind of plain text file listing all available models&textures&materials, parse it, dinamically load assets from there, and have the exhaust ports in plain text rather than .h.

For what I've seen, snis_client.c needs some .h files from the models to know where the exhaust ports are. I guess this should be one major bit of code to clean up and a good place to start.

Bonus points given for:

• A graphical interface to choose available assets whenever a server is started.
• Make the game server (or even the lobby) the only piece that can choose the assets, and have all clients sync the assets from the server they're connected to.
• Provide more than one set of assets.
• Making separate Debian packages (e.g. snis-core, snis-assets-default and snis-assets-startrek).
• Splitting the Git repo structure into code and assets.

Missing file share/warning-hull-breach-imminent.ogg

Error when running snis_client:
Decoding audio data...stat('share/warning-hull-breach-imminent.ogg') failed.

As far the game runs with out issue, but haven't tested to the point of audio triggering.

There is code in gaseous giganticus for computing the curl of the noise gradient like this:

        if (m1 < m2 * left_right_fudge) /* rotate left or right if uphill or downhill */
            quat_init_axis_v(&rotation, &axis, M_PI / 2.0);
    else
            quat_init_axis_v(&rotation, &axis, 3.0 * M_PI / 2.0);

However, left_right_fudge is 0.995. This was originally put in when trying to debug discontinuities in the velocity field which later turned out to be due to something else (see commit 8896774). However, the fudge was left in. The idea behind it is not important as that idea was not correct. So, we should be able to remove this fudge factor. Doing so breaks the program (use the -h option to see areas particles have vacated). It turns out, the value 0.995 causes the if statement to always evaluate to true.

So, the projected noise gradient is always turned right, never left. Why does that work? It (as I understand it) should be the case that we have to turn "uphill" noise gradients one way and "downhill" noise gradients the opposite way. Yet, this is not the way we are doing it, and "fixing" it (by setting fudge to 1.0, or removing the fudge factor altogether) breaks the program (introduces discontinuities and voids (sinks?) in the velocity field.)

So something's weird and wrong, but the mistakes appear to be cancelling each other, (or very nearly so).

Add new button to Comms to toggle repeating Red Alert sound on Main Screen client.

When comms hits the red alert button the main screen computer continuously plays an ogg of a red alert sound. Comms can then hit the button again to turn off.
Eventually I'd like to have it control some lights and/or other screens.

I figure for the sound use a mix of http://www.freesound.org/people/zimbot/sounds/178032/
and
http://www.freesound.org/people/Ionicsmusic/sounds/137948/

I attempted to add the feature myself, but had no luck playing a file.

Hello. Ive been trying to build SNIS on my EEEPC-900HD. I am running Linux Mint 17 Xfce 32-bit and have installed all of the required dependencies without issue. Unfortunately the build fails. Here is the output:

CGAL Cache insert: multmatrix([[3,0,0,0],[0,3,0,0],[0,0,3,0 (606236 bytes)
CGAL Cache insert: multmatrix([[1,0,0,0],[0,1,0,0],[0,0,1,- (606236 bytes)
CGAL Cache insert: group(){multmatrix([[1,0,0,0],[0,1,0,0], (606236 bytes)
CGAL Cache insert: multmatrix([[1,0,0,0],[0,1,0,0],[0,0,1,4 (219516 bytes)
ICE default IO error handler doing an exit(), pid = 16591, errno = 32
make: *** [share/snis/models/wormhole.stl] Error 1
make: *** Waiting for unfinished jobs....
CGAL Cache insert: union(){group(){difference(){multmatrix( (663244 bytes)
CGAL Cache insert: group(){union(){group(){difference(){mul (663244 bytes)
  OPENSCAD share/snis/models/starbase4.scad

Any suggestions? Thanks!

I was testing the latest commits last night and I was being shot by an enemy ship on the other side of a planet. Both beams and torpedoes were coming though. The enemy came out from behind the planet before I could see if I was able to hit it before my ship was destroyed.

I recently setup a machine with Lubuntu Linux 13.04 64 bit (fresh install). I can compile and run the snis_client fine and connect to another ssgl_server and snis_server (on a Raspberry Pi). I am not able to connect to snis_server on localhost. I can connect the lobby server on localhost. Just no servers show up on the list. It is running and I see an open TCP connection listening for the server. I also tried to connect to 127.0.0.1 and the computer name.

I have been able to make and run on another Lubuntu Linux 13.04 32 bit machine with no problems.

Just to make sure I compiled against commit 4d3c0b4 from 5 days ago to make sure it wasn't something from the new Makefile changes. It has the same problem.

There were no errors on compilation. I'm guessing this has something to do with the version of Linux being 64 bit.

Guessing there's some info missing on the new model from commit 16f7874

  COMPILE graph_dev_opengl.c
  COMPILE snis_graph.c
snis_client.c:4765:55: fatal error: share/snis/models/vanquisher.scad_params.h: No such file or directory
  #include "share/snis/models/vanquisher.scad_params.h"
                                                       ^
compilation terminated.
make: *** [snis_client.o] Error 1

I am on Ubuntu 13.04 (32bit) and when i run ./snis_client I get:

kapcom01@ProBook:~/space-nerds-in-space$ ./snis_client 
ALSA lib pcm.c:2217:(snd_pcm_open_noupdate) Unknown PCM cards.pcm.rear
ALSA lib pcm.c:2217:(snd_pcm_open_noupdate) Unknown PCM cards.pcm.center_lfe
ALSA lib pcm.c:2217:(snd_pcm_open_noupdate) Unknown PCM cards.pcm.side
bt_audio_service_open: connect() failed: Connection refused (111)
bt_audio_service_open: connect() failed: Connection refused (111)
bt_audio_service_open: connect() failed: Connection refused (111)
bt_audio_service_open: connect() failed: Connection refused (111)
ALSA lib pcm_dmix.c:957:(snd_pcm_dmix_open) The dmix plugin supports only playback stream
Portaudio reports 11 sound devices.
Portaudio says the default device is: 10
Decoding audio data...ALSA lib pcm.c:7339:(snd_pcm_recover) underrun occurred
Done.
failed reading normal at line 2
failed reading normal at line 2
failed reading normal at line 2
reading 'asteroid.stl'
failed reading normal at line 2
Σφάλμα κατάτμησης (segmentation fault)

But if first do export LANG=en_US.UTF-8 and run again, then it starts fine.

In the middle of a firefight, especially during a single player game and switching screens, accidentally pressing ESC closes out of the game. This could cause the loss of the ship, even though you can reconnect. This takes time.
Having the user confirm that he/she wants to disconnect will reduce this issue.

If you have a torpedo loaded, and your ship is destroyed, when you respawn the ship has a torpedo still loaded.

Have all torpedoes unloaded upon destruction of ship. Also reset controls to default for Phaser Wave Length and any other changed settings.

Since Space Stations should be in known places on the map in the universe for friendlies, they should always be marked on each of the stations (Sci, Weapons, Helm).

Another way to possibly handle this, is to have Comms send out a broadcast to all stations and have the stations give their positions upon this request. The screens could then update with the information.
Range could be reduced for this to not identify farther out stations.

In order for me to compile with the new Lua additions on Ubuntu 12.10 I had to install liblua5.2-dev to resolve dependency for lua.h. Apparently I already had it installed on my 13.04 machine.
This info needs to be added to the Makefile to make the info complete.
sudo apt-get install liblua5.2-dev lua5.2

Recent commit 2eeca4e to make NPC ships warp if travelling distance is too great effectively speeds up the universe, and exposes latent problems with NPC ai.

E.g. they will warp way off the map sometimes. Sometimes they seem to lose their minds and just sit around doing nothing (not in 'H'angout state.) Fleets lose their leader, and nobody steps up, so you end up with leaderless fleets wiggling around.

So... NPC ai needs work.

If you put a space in the Ship Name in the network setup screen, enter your password and connect to lobby - upon clicking on the server and Connect to Server, the screen goes blank and does not continue to show the ship screens. ESC will close the game out properly. I see no error when this happens. Tested on Ubuntu Linux Desktop and Raspberry Pi.

I want to make a packaging script to make snis available to everyone who uses the Arch User Repository, but it is very difficult because the makefile for this project does not support the DESTDIR or INSTALL variables. Making packages easier to produce could greatly increase the popularity of this project, so I think it would be beneficial for everyone if these variables were added.

This is minor.

The build dependencies listed on the webpage are good, but one could be added for Debian Wheezy: libsdl-image1.2-dev

When you are entering Lobby Server, Game Server, Ship Name & Password, allow the TAB key to move focus to the next item on the screen. Similar to how most GUI interfaces work in a windowed environment. This makes the entry of the information quicker without having to go between Keyboard and Mouse to connect to the servers.

Currently (at least as far as I can tell) there's not a way to shoot enemy ships below the crew's ship.

Way to possibly fix it:
Add another turret on the bottom of the ship, and allow Weapons to toggle between the 2 turrets. There could be an additional weapons screen for bottom turret (and one for the top if there is enough players). This would also cause a struggle for resources occasionally. In my opinion having to keep up with this wouldn't be a bad idea.

Just recording this here so these things don't fall off the map and to make it clear that these are known issues.

Some rendering modes don't work anymore after commit bbd7409

• flat shading with wireframe outline
• wireframe with black_tris
• bright line wireframe
• dotted lines in a liine mesh
• nav trident isn't colored by depth
• Derelict objects show up as gray silhouettes (probably due
  to incorrect normals, mesh_derelict() needs attention, and new
  code for computing vertex normals and coplanar triangles for meshes
  outside the context of reading meshes from stl files needs to be
  written.
• Skybox is broken

This commit appears to have broken lighting of planets e16ee2d

However, I am not seeing why that should be the case.

After Docking with a station, not all the sliders are reset to zero. The bar itself is reset but you have to move the slider to have it correctly set it's value.

Also, if there are engineering screens on 2 separate clients the sliders do not match. The bar adjusts fine. Not sure if other stations have this problem.

Before running make on Ubuntu 13.04 I have to remove the -Werror section of the Makefile to be able to compile. I've been doing this for a while now. Honestly, I don't have a big problem with it, but other users might have the problem. I can send the output of make if need be.

Scott Benesh reports:

So here's the commit that broke my apple builds:

053ced4 is the first bad commit
commit 053ced4
Author: Stephen M. Cameron [email protected]
Date: Fri Jul 4 08:35:27 2014 -0500

do not hard code ambient light and allow to be overridden by shader include

And here are some of the error messages I'm getting when I try running

Initialized GLEW
OpenGL 3.0 available
sRGB framebuffer supported
sRGB texture supported
Vertex shader compile error in files:
0: "string_vertex_header"
1: "single-color-lit-per-vertex.vert"
Error:
ERROR: 0:2: '' : #version must occur before any other statement in the program

Vertex shader compile error in files:
0: "string_vertex_header"
1: "wireframe_transparent.vert"
Error:
ERROR: 0:2: '' : #version must occur before any other statement in the program

Vertex shader compile error in files:
0: "string_vertex_header"
1: "wireframe-transparent-sphere-clip.vert"
Error:
ERROR: 0:2: '' : #version must occur before any other statement in the program

These files already have a #version 120 as the first line.

So I did something like this in graph_dev_opengl.c

define AMBIENT_LIGHT_DEFINE "#define AMBIENT 0.05\n"

define UNIVERSAL_SHADER_DEFINES "#version 120\n#define AMBIENT 0.05\n"

and removed the #version 120 line from teh .vert and .frag files which seems to make it work again.

Not sure if there's a better way to do this. Or why Linux doesn't complain.

Devices like the Raspberry Pi, Android phones and devices, Beagleboards and other embedded systems do not support full OpenGL. Please add support for OpenGLES for those and other systems.

Sorry but when I run make after extracting the zipped version of this I get the following error..

Package vorbisfile was not found in the pkg-config search path.
Perhaps you should add the directory containing vorbisfile.pc' to the PKG_CONFIG_PATH environment variable No package 'vorbisfile' found Package portaudio-2.0 was not found in the pkg-config search path. Perhaps you should add the directory containingportaudio-2.0.pc'
to the PKG_CONFIG_PATH environment variable
No package 'portaudio-2.0' found
ogg_to_pcm.c:31:31: fatal error: vorbis/vorbisfile.h: No such file or directory
compilation terminated.
make: *** [ogg_to_pcm.o] Error 1

We attempted a game via the internet, using a VPS to prevent the routing issues currently present in SNIS when not on a lan. After some testing, it seemed that multiple ships didn't have a problem joining snis_server, but if a new crew member joined there was a segfault. The server works fine on a local network.

Here is some gdb output:

(gdb) r 192.227.164.140 0 Arch VPS
Starting program: /usr/bin/snis_server 192.227.164.140 0 Arch VPS
warning: Could not load shared library symbols for linux-vdso.so.1.
Do you need "set solib-search-path" or "set sysroot"?
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/usr/lib/libthread_db.so.1".
Lua setup done.
[New Thread 0x7ffff6ecd700 (LWP 3757)]
[New Thread 0x7ffff64cc700 (LWP 3758)]
found!
[New Thread 0x7ffff58bf700 (LWP 3767)]
[New Thread 0x7ffff4ebe700 (LWP 3768)]
[New Thread 0x7fffe7fff700 (LWP 3776)]
[New Thread 0x7fffe75fe700 (LWP 3777)]

Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0x7ffff6ecd700 (LWP 3757)]
0x00007ffff728ceb8 in __lll_unlock_elision () from /usr/lib/libpthread.so.0
(gdb) bt
#0  0x00007ffff728ceb8 in __lll_unlock_elision () from /usr/lib/libpthread.so.0
#1  0x000000000042ff5d in ?? ()
#2  0x000000000043035e in ?? ()
#3  0x00007ffff7283124 in start_thread () from /usr/lib/libpthread.so.0
#4  0x00007ffff6fb74bd in clone () from /usr/lib/libc.so.6

After locking on to an object such as a destroyed ship and turning on tractor beam, you can not turn it off. The towed object keeps bouncing back and forth.

The slider for Maneuvering is missing until it is clicked on. This has happened with the newest update(s). Not sure if it is a resolution issue, but I am running at 800x600.

Game engine should be abstracted to allow fully configurable bridge stations.

Put 3D models in a sub-folder to clean up main directory and allow future mods.

Move *.stl, *.scad (and *.ogg) files into a sub-directories so it will be easier for Mods to be created

Something simular to this:
/home/user/space-nerds-in-space/snis

Sub-directories for the files themselves
/home/user/space-nerds-in-space/snis/audio/ for .ogg
*/home/user/space-nerds-in-space/snis/models/ for *.scad & *.stl

This would allow directories for Mods such as the below to be created to house their models and audio

/home/user/space-nerds-in-space/startrek
/home/user/space-nerds-in-space/starwars
/home/user/space-nerds-in-space/babylon5
/home/user/space-nerds-in-space/bsg