next »

[trabitboy]

Hi,
can't get the 0.4 demo of your gorgeous entry to work under XP:

Code:

C:\Documents and Settings\tlacourt\Desktop\hardaether-0.0.4(2)\hardaether-0.0.4>hardaether.exe
Traceback (most recent call last):
  File "hardaether.py", line 17, in <module>
  File "ogre\renderer\OGRE\__init__.pyo", line 10, in <module>
  File "ogre\renderer\OGRE\_ogre_.pyo", line 12, in <module>
  File "ogre\renderer\OGRE\_ogre_.pyo", line 10, in __load
ImportError: DLL load failed: This application has failed to start because the application configuration is incorrect. Reins
talling the application may fix this problem.

C:\Documents and Settings\tlacourt\Desktop\hardaether-0.0.4(2)\hardaether-0.0.4>PAUSE
Press any key to continue . . .

do we need to install the python package to make it work?

[nihilocrat]

You shouldn't have to, but I guess you could. Every machine I have built / run it on has Python on it.

There's a bug about the issue (http://bugs.python.org/issue4566), but it's for 2.6.1, and I embed 2.5.4 and I'm sort of restricted to it because PythonOgre only has binaries for 2.5, and I am sure as hell not going to build it from source.

I will try cloning my virtual WinXP machine at home and uninstalling Python on it to see if I can figure out the bug and put a fix into the next version.

[trabitboy]

thanks,
environment crash is always painful.
will test on a second xp machine. 

[nihilocrat]

UNION SPACE NAVY MEMORANDUM USN-XO-309-24

TITLE: P.R.M. ZhL-61 "Flamberge" Cruiser: Advance Report
VERSION: 0.0.1 rev 1

CLASSIFIED LEVEL 8 XO
IF YOU ARE NOT AT LEAST LEVEL 8 XO CLEARANCE,
PLEASE DESTROY THIS MEMORANDUM AND PROCEED TO THE CLOSEST DEPARTMENT OF HOMESPACE SECURITY OFFICE FOR IMMEDIATE IMPLOSION TESTING
==========

Spy telescopes have confirmed the People's Republic of Mars is now mass producting of the Zhukhov-Li 61 (ZhL-61), Union designation "Flamberge", Cruiser-class spacecraft.

Shot at 2309-03-26

[Chris Z]

FIRE ZEE MEEZOS!

[Mikademus]

Launch all 'Zig'!

[nihilocrat]

Shot at 2309-03-26

Notes about this shot:
* The buttons have labels. Some of them make sense. Others are just "SHIP" or I dunno WTF.
* There is now a working mouse cursor, which you can switch from "flight" mode (an angle-of-attack icon, controls attitude) and "select" mode (normal old mouse pointer, used to click buttons)
* A skybox that I can settle with is now present
* You can now target missiles. This is a core feature of gameplay, because you need to intercept the ones coming towards you.

However, there are bugs, and a fair deal of missing stuff. I think I'll manage to get it all done this weekend, though, given I don't have any distractions.

[Chris Z]

Code:

Mesh: Loading geosphere_1km.mesh.
Texture: test.png: Loading 1 faces(PF_A8R8G8B8,256x256x1) with 5 generated mipma
ps from Image. Internal format is PF_A8R8G8B8,256x256x1.
Mesh: Loading arbiter.mesh.
added body 250.0 250.0 -250.0 -250.0
Mesh: Loading ferret.mesh.
added body 250.0 250.0 -250.0 -250.0
Mesh: Loading firebat.mesh.
added body 250.0 250.0 -250.0 -250.0
Mesh: Loading cockpit.mesh.

C:\temp\hardaether-0.0.4>PAUSE

I hear the music kick in for a split second and then it crashes while the screen is still black.  Maybe theres a problem during/after loading cockpit.mesh?

[Will Vale]

Sorry it's taken me so long to reply and try again, I've been mad busy with GDC stuff.

I don't recall hearing the computer voice when I played before. With the latest version (with folders hopefully copied as per the instructions) I get a crash on startup under Vista.

Latest screens look cool though!

Cheers,

Will

[nihilocrat]

Wow, this is really pissing me off  . I can run every build just fine on Vista, my virtual XP box, at work, etc.. Ogre.log should contain any OGRE-specific errors / messages, it might give more information on what's wrong. If it just suddenly quits with no sign of an error, the only case where I've seen that is where a FrameListener returns False, but only the Controller's listener does that when you hit Esc.

The computer voice wasn't working due to some path-related stuff that I uncovered, which of course worked on my development box because everything was in the right place. I've since fixed it.

This, and Python memory management, are really annoying. If you want to delete an object, you have to hunt down every reference and delete it. This is causing the ships to stick around when you kill them, and missiles to keep truckin' on if you intercept them. More technical issues than I really need, seeing as how I still have to actively develop a few key portions (button response, main menu, mission begin / player death transitions).

Due to time constraints, I'm throwing out visible laser beams. If you're lucky, I will have some sort of flash particle effect when you fire or see an enemy fire. Conveniently, this is the way lasers look in reality. I'm noticing the more visual cues you take away, the more you need to find ways of replacing them, or else you feel disconnected from the game world and your actions in it. Thusly, the cockpit will start to blip and bloop a lot more.

Anyways, I will publish a 0.0.5 build today sometime, simply to see if people can run it. I would have preferred to not release any more builds, such that the final build would seem much more improved in comparison, but oh well.

[Mikademus]

Due to time constraints, I'm throwing out visible laser beams. If you're lucky, I will have some sort of flash particle effect when you fire or see an enemy fire. Conveniently, this is the way lasers look in reality.

Sometimes In a perfect vacuum an efficient (well-aligned) laser beam would be invisible for a few thousand to a few hundred thousand kilometers. However, if there's any medium (particles), or the beam is travelling through a string magnetic field, there will be a trail from (1) refraction (scattering of photons from the medium or light flashed from particles annihilated on-route), (2) Chernokov radiation depending on the density of the medium, and (3) synchrotron radiation from angle shearing in magnetic fields. All of which is you should immediately ignore and finish the core of the game instead

Nice drawing of a realistic laser impact, there are two lasers firing from the ship on the right (the two tiny blue dots on the hull edge and right below the shadow region start); notice the flash point and the angled vapour geysers on the target :

[nihilocrat]

New techdemo. I added in the 'manifest' workaround to the py2exe problems, so hopefully it works for everyone.

Thanks for pointing out where I'm wrong. Like I said, I know hardly any physics, and I definitely don't know what the hell "Chernokov radiation" is. I think after completion, I will go back and refine the game and pass it along hard sci-fi buffs for their opinion of it and its presentation of reality. It would be really cool if it became a sort of "reference" hard sci-fi game. I-War and others get really close but aren't quite hard enough, probably because they expected people to actually buy the game .

[Mikademus]

It would be really cool if it became a sort of "reference" hard sci-fi game. I-War and others get really close but aren't quite hard enough, probably because they expected people to actually buy the game .

Yeah, bloody wimps! 

[jute]

It would be really cool if it became a sort of "reference" hard sci-fi game. I-War and others get really close but aren't quite hard enough, probably because they expected people to actually buy the game .

It would be nice if this game ushered in an era of too-complex-to-market hard-sf indie games.

[nihilocrat]

Now I'm realizing how silly that sounds... I would be happy if people just liked it enough to play it more than once.

Please chime in whether you can run the latest version (0.0.5) or not, if you've got the time and are in a situation to try it out.

[Pishtaco]

It isn't working for me. First I got something that looked like a missing directx file, but updating my directx got me past that. Now I get a lot of text that looks like stuff loading, and then "hardaether.exe has encountered a problem and needs to close." The last page or so is:

Code:

Info: Freetype returned null for character 162 in font Frontier
Info: Freetype returned null for character 164 in font Frontier
Info: Freetype returned null for character 165 in font Frontier
Texture: FrontierTexture: Loading 1 faces(PF_BYTE_LA,256x256x1) with  hardware g
enerated mipmaps from Image. Internal format is PF_BYTE_LA,256x256x1.
Bad overlay attribute line: 'arbiter/Cockpit' for overlay arbiter/subtitleOverla
y
Bad overlay attribute line: '{' for overlay arbiter/subtitleOverlay
Bad overlay attribute line: 'entity cockpit.mesh(cockpit)' for overlay arbiter/s
ubtitleOverlay
Bad overlay attribute line: '{' for overlay arbiter/subtitleOverlay
Bad overlay attribute line: 'position 0 0 0' for overlay arbiter/subtitleOverlay

Bad overlay attribute line: 'rotation 0 0 0 0' for overlay arbiter/subtitleOverl
ay
Parsing script Compositor.overlay
Parsing script DP3.overlay
Parsing script Example-CubeMapping.overlay
Parsing script Example-DynTex.overlay
Parsing script Example-Water.overlay
Parsing script gui.overlay
Bad element attribute line: 'horiz_align center' for element MouseCursor in over
lay arbiter/CursorContainer
Bad element attribute line: 'vertical_align center' for element MouseCursor in o
verlay arbiter/CursorContainer
Texture: cursor_select.png: Loading 1 faces(PF_A8R8G8B8,16x16x1) with 4 hardware
 generated mipmaps from Image. Internal format is PF_A8R8G8B8,16x16x1.
Parsing script PythonOgreDebugPanel.overlay
Texture: popanel.png: Loading 1 faces(PF_A8R8G8B8,638x90x1) with 5 hardware gene
rated mipmaps from Image. Internal format is PF_A8R8G8B8,1024x128x1.

C:\Documents and Settings\Neil\Desktop\hardaether-0.0.5>PAUSE
Press any key to continue . . .

[Galaxy613]

It's working fine for me, I just need to find the manual.

[nihilocrat]

I just need to find the manual.

What do you mean... "manual"? :D

No, seriously, WASD to turn, X to thrust forward, Z to thrust back, 1 to target ships, 2 to switch weapons, 3 to target missiles. I probably forgot to put the manual in the zip... my build process isn't exactly clean and orderly if you can't tell.

I'll have another build out soon with another attempt at a fix for all the other people. I think it involves bombarding the main directory with OGRE-related DLLs. Also, this new build will have fully-interactive buttons. I am too lazy right now to make the buttons light up when pressed, so instead they just make noise (which I think is enough response for a decent UX anyways). Still needing to wrestle with ship/missile deletion; hopefully I can get interception to properly work, meaning the game is :gasp: gameplay-complete in a gimpy way.

[Fuzz]

It keeps on telling me that it doesn't have necessary OGRE-related DLLs, so that'll probably help it.

[nihilocrat]

I think it's the case where, for some reason, it doesn't include them. What's confusing to me, and what makes this very non-obvious, is that I can run the builds in any random directory, not just a subdirectory of PythonOgre, and the game seems to work for people with Vista out-of-the-box.

Anyways, I'll push out a build in a bit. I'm still working on managing the death of missiles and ships, so it might crash if you die or you shoot down a missile or something.

update: Haven't made much progress today, because the game finally has core gameplay finished! I have mainly just been playing one-on-one engagements and refining tiny things here or there. I am surprised by how fast the ships and weapons travel given their current accelerations of 1G ans 2G respectively, it's making me think I've got the math wrong somewhere.

On the other hand, there are a handful of new and old bugs that I'm having to fix, and I'm a little concerned that it becomes easy for missiles to completely miss their target (overshoot and have way too much velocity to hope of ). I wanted this to be possible, but not to the point where engagements at high speeds were impossible because the missiles would never hit. I'm not intentionally putting inaccuracy into the missile tracking algorithm, it is just a gimp one because I'm bad at math, which happens to work with my other games / demos because the speeds are relatively low. I think I heard once that a proper target-prediction algorithm involves some sort of differential calculus, and I haven't found a straight answer anywhere with Google.

Next items on the menu: a main menu, debriefing screen (when you die / kill everything else), and a never-ending "gauntlet" mode which will be the sole mission available for the compo version.

After that, I hope to have a compo-final version. If the deadline gets extended, I will try and add a 3D icon to the target screen so you can actually see an image of the craft targetted (like what TIE Fighter did), which should make engagements a lot easier because you know exactly what you're targetting without having to consult my cheezy code-names.

After that, a slightly more difficult feature, would be a much-needed compass/attitude indicator overlay to replace/compliment the STATUS screen. Right now it's pretty terse, just a series of numbers, and you have to sort of imagine which way you are going. I could even add indicators to help you point yourself at your chosen target, but that make the game too easy ! The information is already there, it's just hard to use.


Also, it should be present in the 0.0.5 build, and it will be present in the 0.0.6 build (hopefully people can run this one!), but you can modify the game in minor ways through the files media/data.py and media/missions.py. Data.py is pretty straightforward, and missions.py has a convenience function that lets you create enemy formations in a single function call, so hopefully it's not hard to create your own missions.