Loading...
  OR  Zero-K Name:    Password:   

Spring 95.0 released, will be enforced on lobby Thursday, 7. Nov

108 posts, 4750 views
Post comment
Filter:    Player:  
Page of 6 (108 records)
sort


11 years ago


http://springrts.com/phpbb/viewtopic.php?f=2&t=31118&sid=18c3c7aaa3ef796dacbe57e345335917

If you haven't bothed to test yet, do so now! You have 2 days to report any game-breaking bugs.
+6 / -0
We need a v95 autohost so we can test it...

And need Benchmarks vs old engines too to test performance...

http://zero-k.info/Forum/Thread/4906

https://docs.google.com/spreadsheet/ccc?key=0Ar854YTTzQWtdE5WcWUwMzM1aTN0LXd4SG9QODlia3c#gid=0
+0 / -0
Make sure you have LogFlush=0 if you want to try bleeding edge or 95.0... engine default is 1 and it causes uber low FPS
+0 / -0
11 years ago
Finally! Lets try to avoid another 94.1.
+0 / -0


11 years ago
It might be a bit late to "prevent another 94" if it's already rolling out...
+2 / -0


11 years ago
Well it's most certainly not as bad as 94 or 94.1, the worst bugs have been squished, but performance is definitely still worse then 91.0
+0 / -0


11 years ago
whats the upside?
+0 / -0
Pathing, less bomber related bugs, static builds for linux. There is a much faster version of the UI framework in development which requires 95.0.
+1 / -0
11 years ago
Yes, flying spaghetti monster heard my prayers!

91 was pain for sooo long.
+0 / -0


11 years ago
whats the 95.0 pathing like?
+0 / -0
11 years ago
Bugs, bugs, bugs...

New pathfinder sucks.

http://zero-k.info/Battles/Detail/208682
Athena - broken (floating like shit, and not resurrecting)
Stucking units - commanders and other units stucking and didn't move.

http://zero-k.info/Battles/Detail/208677
Ravager - didnt move on the target way...
+0 / -0
Upload the replays and give the coordinates of each of those problems. Map coordinates can be seen by holding down space and mousing over the map. Don't forget time.

I happen to already have the replay for the first game.
  • There is already a workaround for the bouncing commander bug.
  • DinoCzar spent the entire game whining about the pathfinding because he had Vehicles which "went the long way around" when he told them to go into craters on MoonQ20X. The pathfinding was actually very impressive, they managed to find the slightly less steep parts of the crater walls, dive around on the rim and them drive down into the crater itself using a very small area of less steep wall.
  • The Athena thing looks quite easy to workaround. The engine always has these small bugs which we have the patch with lua. The difference is that the previous engine mostly has these flaws patched.

I can't parse "Ravager - didnt move on the target way...".
+1 / -0


11 years ago
quote:
There is a much faster version of the UI framework in development which requires 95.0.


Faster Chili?

That sounds like a huge, huge win for Zero-K.
+2 / -0
FIrankFFC
11 years ago
omg it lags so much
+3 / -0
11 years ago
Killer is right i went through 2 hours of troble shooting just to find out nothing is wroung with my Pc. my problem is when i would start a game it will all load fine until i get the intalizing pathcache that wen u haft to wait for 2-5 mins then when u finally enter a game ur kicked for inactivity. and befor the 2-5min wait u would haer the "Commad Consle Activated" but it still stays the same for the next 5 mins. Then when it finally shows u the game u cant do anything but catch up
+0 / -0


11 years ago
out of interest, what computer specs are you guys running?
+0 / -0

11 years ago
Also,
quote:
Make sure you have LogFlush=0 if you want to try bleeding edge or 95.0... engine default is 1 and it causes uber low FPS

Granted, worst engine default ever.
+1 / -0


11 years ago
what does logflush do, and can ZK lobby make that off by default?
+0 / -0
11 years ago
http://springrts.com/wiki/Springsettings.cfg#LogFlush
"Instantly write to the logfile, use only for debugging as it will cause a slowdown"
Instant writing vs buffered (delayed) writting of infolog.
LogFlush=0 can be annoying for testing: for example you see a lua wupget fail, alt-tab to open infolog.txt but the errors have not been written (yet).
On crash the infolog can sometimes be incomplete.
http://springrts.com/mantis/view.php?id=2796
With LogFlush=1 infolog is written instantly but in case of textspam (chat or error messages or messages that are not shown in ingame chat) it is slower.

tl;dr
=0 for playing
=1 for testing, esp if crashing


quote:
i get the intalizing pathcache that wen u haft to wait for 2-5 mins
It should only take that long the very first time.
If it happens often:
"Switching between engine versions causes long startup times"
http://springrts.com/phpbb/viewtopic.php?f=11&t=29070
As I understand the reason is that you have multiple spring versions installed and they use different folders for maps/games/rapid-pool. Then spring will rebuild its cache on start.
By making sure all spring versions use the same folder for maps/games/rapid-pool, and not having any files anywhere else, all installed springs now start instantly.
+0 / -0
11 years ago
So why not just delete the enitre game then reinstall it.
+0 / -0
Page of 6 (108 records)