Spring 94.0/1

[klapmongool]

So this does not auto update through lobby to latest version?

[zerver]

94.11 is an unofficial release so definitely not. I think the lobbies provide major releases only (95.0 --> 96.0 etc) but the bugfix releases have to be manually downloaded.

[fruehling]

Under which circumstances should I decide for APATH or not-APATH versions?

MT (APATH) is much faster. The other one is simply 94.1 + bugfixes. I recommend you get both versions.

If you play single player setup the lobby to start spring-mt to get the performance boost.

If you play online the correct version will be started automatically, so you must join an MT host to enjoy the added performance.

Aye, thanks -- I didn't see the no-APATH-version was not an MT version.

[klapmongool]

94.11 is an unofficial release so definitely not. I think the lobbies provide major releases only (95.0 --> 96.0 etc) but the bugfix releases have to be manually downloaded.

So you can play with this version with people who didnt download it?

[zerver]

On a normal host, yes. MT or not does not matter there.

On an MT host, all users must first download MT.

[abma]

zerver: please create a dedicated thread about your release somewhere else. It is really confusing and annoying to have one thread about bugs/questions about your "fork" and about spring 94.1.

[Google_Frog]

Can we have a syncing bugfix release with LogFlush defaulting to 0 and fixed Ceg spikes?

[Kloot]

ZK is still happy and content using 91.0, so why should we bother making a 94.2 and why do you even care?

Unless your position is that those two minor issues are the only reason ZK has not upgraded yet, which would be ridiculous.

[Silentwings]

A mod-side hotfix for LogFlush could be done using SetConfigInt.

[Google_Frog]

Why do you think we are happy with 91.0? We cannot update while doing so would break the game for a portion of our current playerbase. It's the trouble with using features of fragile compatibility (often ATI related) and the arguments in lua tend to shuffle around. If we switched to a new engine every time it was released we would have no playerbase. The serious issues with this release can be fixed modside but it doesn't happen instantly.

There is even a list here http://code.google.com/p/zero-k/wiki/NewEngineTodo

Most important fixes that I can think of just now:

• Aircraft attack order brokenness.
• Pathfollowing is much better in 94.1. Units bounce on terrain instead of slowing down. I don't care how unphysical this is, it makes the pathing good.
• Apparently sim is much faster.
• More callins.

LogFlush is a serious issue. Even if we can reset springsettings with ZKL there will be many players without the reset. SetConfigInt could fix this.

Cegs are not vital and 94.1 is good enough that I would switch if they were the only problem (I don't speak for everyone but I haven't asked them either).

That said I would not want to do without these Cegs for the next year.

[raaar]

Why was CEG broken in the first place? I Made lots of effects based on glowy spikes, and now they don't show up..

There's also the ship transport bug

EDIT: added the CEG problem to mantis

[Kloot]

Why do you think we are happy with 91.0?

http://zero-k.info/Forum/Thread/4713

There are a lot of new bugs in 94.1.

91.0 works for everyone and 94.1 doesn't seem to have many improvements.

Now you are praising the "important fixes" and even asking for a 94.2 (which would still be full of those same "new bugs" as a syncing release), so which is it?

[Jools]

The biggest drawback with spring 94.0/94.1 seems to be that it's very heavy on the cpu. People can't play bigger maps, some people must connect using the 'safe mode', and some people can't play at all.

Is this due to the new pathfinder or is it because of the new multithread thing?

[Silentwings]

The biggest drawback with spring 94.0/94.1 seems to be that it's very heavy on the cpu.

It's true that 94+ is slightly heavier on the cpu than 91.0, but I think its a small effect and I've not seen any big change of the types of maps large games are being played on, or to the number of people who's computers can handle it. Overall, big improvement over 91/2/3 imho.

[raaar]

spring 91 had some nasty problems with units colliding with factories and displacing transports or making them get stuck.

[zerver]

It's true that 94+ is slightly heavier on the cpu than 91.0, but I think its a small effect

I wrote earlier that 94.0 is faster but I have changed my mind. In synthetic tests it may appear faster but in a real game 94.0 is quite a bit slower than 91.0.

Is this due to the new pathfinder or is it because of the new multithread thing?

Changes to the pathfinder + the pathfinder is being used more heavily. I didn't touch this code so I leave it to others to say whether it is worth it or not in terms of improved unit behavior.

[Silentwings]

I leave it to others to say whether it is worth it or not in terms of improved unit behavior.

My experience is that the overwhelming consensus is it's worth it.

[danil_kalina]

Sping 94.1, Mac OS X: Spring-Multithreaded GML Crash
http://pastebin.com/ar4m6Gvy

[danil_kalina]

Spring 94.1, Mac OS X: rendering artifacts

Square Bordes

http://pbrd.co/ZG6RL3

[Google_Frog]

Why do you think we are happy with 91.0?

http://zero-k.info/Forum/Thread/4713

There are a lot of new bugs in 94.1.

91.0 works for everyone and 94.1 doesn't seem to have many improvements.

Now you are praising the "important fixes" and even asking for a 94.2 (which would still be full of those same "new bugs" as a syncing release), so which is it?

Clearly something changed my opinion so it is the most recent one. Do I have to make a snap decision whether to like something and then stick to it while ignoring future evidence?