Eve of the Apocalypse: Twilight

It's just a vague idea, but couldn't you split EotA into multiple maps ?

Explanation : develop a map file (Say, eota1.w3x) with only 1 eota "map" in it at most to prevent slowing it too much, with all the triggers and shit in it. Then you make another map file (eota2.w3x) that's got all the objects and triggers from eota1.w3x. Then you just copy a few map-specific triggers you've put in another map to save time.

Wouldn't that work ?

No.
Too much work updating and too many maps floating around.

Also, unless you really know what you're talking about.. Well.. I'll be as obvious as I can with this: There are a lot of factors to performance on a map, and some factors might not be as big as you think as others. Other factors, unless you know your stuff, you might not even realize exist.

I'm not saying Yak is an idiot when I say this, but I really would not doubt it if there were factors unknown to him. Basically, it might not even make a difference in how well the map performs.

You'd have to make updates to each map individually, meaning new versions would take X times as long. And in general it'd be a total mess.

Near as I can tell, only arena-specific assets are loaded in the game. So splitting it up would not improve performance, though it would reduce the file size for downloading -- but you would want to have all the maps, anyways, so it would actually take more hard drive space* and it would be harder to determine a noob dler from a decent player who just doesn't have that specific map.

*hard drives dole out space for files in chunks that are multiples of a specific size to make it faster to access, so more files = more empty space just to please the OS
Not enough wasted space to make any real impact, but since it's the only appreciable difference between the current way and the proposed way, I'll use it as an argument against.

Kibiyama wrote:Near as I can tell, only arena-specific assets are loaded in the game. So splitting it up would not improve performance..

And do you know exactly how JASS works?

This is what I mean. On what level is your understanding to be able to confidently say what will and will not increase the maps performance?

I'm not about to make a claim that I can't fully back, because honestly, I don't know jack about what hits on performance the most, I can only assume.


Edit - Yes, this does pertain to map splitting. More triggers DO have a chance at causing more lag, I don't fully understand by how much though. I don't know enough to make an educated comment more than "it's possible".

Something wrote:

Kibiyama wrote:Near as I can tell, only arena-specific assets are loaded in the game. So splitting it up would not improve performance..

And do you know exactly how JASS works?

Oh, and you're a master programmer yourself huh?

Nope, I'm only stating that unless you really know what you're talking about, all you can do is make an assumption.

And I'm not making an assumption by stating such.

Just incase you didn't notice, I had already stated that I don't know enough to make an educated statement of fact that it will or will not be such, I can only offer my thoughts based on what I know.

It would create hell for Yak. He would have to redo a lot of triggers, probably redo the terrain for all of the maps, and just have to do a lot of shit that would make it take more time.

I really would like to know what the gain from this would be. I recognize that it would slow down development, but if it actually did significantly improve load time, then it would be worth weighing the pros and cons.

Of course, if Yak has said that he won't do it then that's that, but I would like to hear his appraisal of whether or not it would significantly quicken load time.

wrong season of love, you can copy paste terrain and doodads.
And even so, all maps except stormwail will only take 2-3 hours terraining.
Terraining is one of the easiest things, the trigger will take more time, checking every ability for the right raw code.

Hmm well when i think about it, it wont take to much time. Since he can copy trigger terrain doodads and object data.
Only changing the raw code and a mass amount of variable will take most of the time.
And ofcourse Upgrading a version...

I don't know JASS.

However I do a fair bit of C++ game coding.

I've said this multiple times, but nobody ever seems to beleive me.

The large terrain size adds at most 3-5 seconds to the load time (probably less), no processing time in game, and only a small amount of memory (around 1mb I would guess).

The only reason Iwould go with a multi map approach (which i'm inclined to try if Iever get started on eota2) is for making the editor easier to work with.

DarnYak

DarnYak wrote:I've said this multiple times, but nobody ever seems to beleive me.

You'll have to add it to an FAQ, then.

Something wrote:Nope, I'm only stating that unless you really know what you're talking about, all you can do is make an assumption.

And I'm not making an assumption by stating such.

Just incase you didn't notice, I had already stated that I don't know enough to make an educated statement of fact that it will or will not be such, I can only offer my thoughts based on what I know.

What we typically call this in the world of business is putting your nose where it don't belong.

I don't ask the HR manager about the company's internal controls over financial reporting. Why? Because they don't know enough about it to make an educated statement.

All they can say is put their stupidity towards it... which is basically what you've been doing.