Spring 0.78.1.1 is out

[lurker]

Yes, that *is* how the new versioning system works.
0.78.1 - synced
.1 / whatever - unsyned

[Warlord Zsinj]

Hmm, my usual way of testing seems to no longer work.

Usually I'll jump into spring.exe, load up the 'commanders' option, and then jump into /spectator then /godmode, which allows me to easily spawn and test units etc.

In this new version it seems to want to load an opposing player ('waiting for players to join, hit ctrl+enter to start...'), whereas before it simply waited for you to hit enter to start the game. Then it ends the game when no other player is loaded.

Was this intentional? If so, how come? What is the usual way for other people to quickly test their mods?

[imbaczek]

move the "game over" window off screen and proceed as usual ;p

[FLOZi]

[quote="imbaczek"]move the "game over" window off screen and proceed as usual ;p[/quote]

That would work if you got the game over window, you don't.

[lurker]

I do...

[FLOZi]

Spring just craps out on me. :( Some error about no arm commander :rolleyes:

[lurker]

Different bug.

[imbaczek]

ah, get the test build for that.

[thesleepless]

CPU usage seems to have increased hugely since 0.77
mainly in sim time, getting over 40% in sim time on a small game
quite noticeable framerate drop. with settings on lowest.

this never happened before, any idea what could be causing this?

linux build, archlinux, amd athlon xp 3000+, 2 GB RAM, geforce 6200

[lurker]

What fps? Spring now counts two frames per second in the sim time.

[jK]

zwzsg:
"The color of the bit shots switch between yellow(???) to blue(what should be) depending on the parity of the number of time I alt-tab. Meanwhile pointer shot trail do the opposite."

Seems to be an Ati bug. To fix I would need you (or somebody else who has this issue) to run this http://home.arcor.de/jkei/spring_debug.exe
Run the game in fullscreen, AltTab once, close it and then open the infolog.txt and tell me the number in a line like this:
[ 45] internal format: 8058
[ 45] internal format: 8058

[lurker]

alt+tab away once or alt+tab away and back once?

[zwzsg]

[ 0] Finalizing...
[ 0] Spring 0.78+.0.0
[ 0] Build date/time: Jan 13 2009 02:46:18
[ 0] Player You connected with number 0 (client version 0.78+.0.0)
[ 0] Player You finished loading and is now ingame
[ 0] GameID: c76e6f4948eb02bd306548bdf9c3c988
[ 383] internal format: 8058 <- this one is after the very first alt tab
[ 1330] internal format: 8058
[ 1611] internal format: 8058
[ 1827] internal format: 8058
[ 2051] internal format: 8058
[ 2268] internal format: 8058
[ 2684] User exited
[ 2684] Path cache hits 14 40%
[ 2684] Path cache hits 26 19%
[ 2684] Statistics for local connection:
Received: 5826 bytes
Sent: 27243 bytes


But I alt tabbed a little more than once...

[jK]

k not what i expected, need some more info then:

http://home.arcor.de/jkei/spring_debug2.exe

same procedure. This time the wanted lines in the infolog.txt start with "[ 0] pixelstore: "

edit: this time you should alttab once back (need download and upload states)

[zwzsg]

here:

[jK]

k thx.

result:
It is a bug in the driver and it is not possible to detect it at any state value :/
Also it seems the bug is very old (2005) and still not fixed. see here: http://www.rage3d.com/board/showthread.php?t=33829986

So I need to introduce a new atihacks config to workaround it mehh :<

[zwzsg]

Thanks. How come I had never seen it in all the previous Spring version?

[jK]

It is a bugfix for the fbo corruption issue (garbled particleatlas) -> it's new in 78

[Pxtl]

Saw it before on my onboard intel chip, but fiddling with settings made it go away. May be a different bug, but it looked the same.

[vatoslocos]

Nice! playing the new version for some days now, but forgot to say thanks!
so, thanks!