Alternatives to gamecache a placebo?

Here it comes I need people to test stuff, this is just a poll to check out statistical results. Besides according to my tests I predict this shall make everybody think twice before using gamecache that much the next time. But we may as well discover that gamecache isn't that bad.

Look I need people to download the map at: http://www.wc3campaigns.net/showthread.php?t=78575

Then download this post's attachment : oldfrozenorb_puregamecache.w3x

Then simply test the frozen orb in both maps and opine if there was a difference lag-wise.

hmm well, i noticed no real difference

and I have my graphics in WC up all the way

ofcourse I have a pretty fast comp, atleast fast for WC standards

does the new version have the same bugs, or as many, as the regular tables method would? as far as freezing in game and crashes?

btw i saw a lot of getarray/ setarray stuff in the pure game cache one

It is not like the bugs you are talking about exist. But since it is a different method it shouldn't have it.

Recommended test: Add another archmage to the game and make 2 archmages cast Frozen Orb at the same time - That would make an effect in even the fastest computer

BTW, if you noticed no difference then it is a victory for arrays, cause the new frozenorb uses a 0.035 timer and the old one uses a 0.05 timer

Meh, I stopped using gamecache. I'm using it and the return bug only to destroy triggeractions, the rest, i'm using global arrays, which is much more safe in my opinion.

Ye, but sometimes its not convenient to sacrifice object orientated for speed. Most of the time actually its not convenient. Gonna test right away.

Edit: I dont know how you manage to save or test the latter one, cus it has some errors in it.

That's odd, maybe some issues with the optimizer, try just using testmap

I did, it booted me right to the main menu.

Holy lameness, I think I have deleted the gamecache version. The attached map won't work which makes you wonder how emijlr tested it. Does anyone have the r2 version of FrozenElementalLightningOrb.w3x ? So we could test?


vile: Global arrays are ok but they aren't dynamic, then they suck.

btw: Optimizer <= 3.7c seems to have had bugs with systems that used gamecache


INfraNe : you can keep the object oriented ness by abusing comments, this is the reason VJASS is going to be developped though, to keep the object orientedness but still use arrays

Vile: When you say gamecache isn't safe, what exactly are you refering to? If there are problems with it (not counting the speed) it's useful to know.

Vile has very valid points though.
Game cache is a piece of shit when you need to make spells for 15+ heavily triggered heros in an AoS map that already has dozens of systems running at the same time.

I myself have been hating the game cache as well.
I have had Blu and others check my work and it's all 100% working, but as soon as dozens of spells using cache are casted at once, it blows up.
I dont care what anyone says about game cache being awesome or whatnot, because it just is not reliable.
And for me, reliability is usability.

As of now, I've reverted myself to only use game cache on 1-2 spells per hero and do the rest global arrays as well.
It just works better.

i jsut checked again and I must have tested the same map twice lol, czu yea that other one doesnt open... ooppss

no wonder I was seeing array stuff in both of them

but yes, I agree with Vile, game cache has been screwing me up to, much more so after using the optimizer, so hopefully the new version fixed that

I know, but I was refering to what vile meant.

I've myself never had problems with gamecache so it has to be problems with things you are doing, not with gamecache.

Likewise, and I've been using it on a AoS with over 40 different heroes and their abilities as well as multiple systems without any problems whatsoever

how could there be a prob with what we are doing with it?

i use it the same as anyone else, and g et many missile freezes/ other bugs

ofcourse again that could have bee the optimizers fault

I'm sorry, Ive posted my problems here before and noone can solve them because they aren't problems with what I'm doing,
but weird screwups happening in the cache.

I don't post my opinion about something unless I've beaten it to a bloody pulp in my own tests, and let me assure you game cache is not reliable.
It's a logical fallacy to say "Because it works for me and not for you, you're doing something wrong".
In all reality, I think cache reacts differently to different maps and different levels of stress.
In most cases, it works wonders simply because it isn't stressed to the limits, whereas in cases of INSANELY, heavily jass'ed maps that don't use a single bit of gui and then have anywhere from 12-20 heavily game cache based spells running at the same time it can detonate because of the sheer stress on it.

You can run your millions of tests and claim what you like, but I've tested it in an actual application enough times and it has failed enough times in entirely random and inexplicable ways that I can assure you, game cache is not reliable.