Reducing the featured map-pool

quote: [18:07] Licho ok guess whats the size of steam package with all featured maps [18:10] Skasi all? [18:10] Skasi 10gb? [18:10] Licho just featured which are cycled on springies [18:10] Licho 7gb [18:10] Licho pretty close [18:11] Skasi without any maps it's 200mb or so? [18:11] Licho yeah but we are using steam so we will let push it 7gb to all [18:11] Licho so that we dont have to wait for any loads [18:11] Skasi does steam have download levels like "Minimum", "Full", etc.? [18:12] Skasi some games have that so people can start playing sooner [18:12] Licho we could do that but waste of effort imo

Instead of reducing the number of featured maps I suggest creating different levels of downloads and let new players choose how long to wait before jumping into the game. Let players download as little as necessary, but also as much as they want. Possibly have it download less popular featured maps and potentially other non-critical files in the background. I assume a platform such as Steam would have settings for those kinds of things.

Personally I feel like there's already too few maps in the map rotation; games played on public hosts very often come down to the same 10-20 maps because of awful artificial coded limitations. (Titan Duel is the new DSD)

quote: Heartbreaker v2 (dosn't play well)

Shush! Heartbreaker is an awesome map! Its only downside is the lack of deco/textures, but that could also be good to keep it clean and readable.

but but but but on steam it says:

quote: Enjoy endless variety of tactics and strategies with over 100 units and 3000 hand-crafted maps.

Also I said this in dozen other threads before, the way featured/supported system gets used makes no sense.
Better is:
Supported = all that are good enough to play.
Featured = maps that have special attention. (For example "map of the week" or "winter maps" themed mappool before christmas or whatever)

For steam if you still expect the "zomg epic wave of players" it makes no sense to have 7 GB of maps because players will have little understanding that game with so old/bad graphics takes so long to download. At start the important thing is that they play something at all. So have 3 maps, have autohosts use only those. When situation slowly goes to normal you can introduce more maps.

quote: Possibly have it download less popular featured maps and potentially other non-critical files in the background

During playing? That will lag out some players and for little gain.

https://github.com/ZeroK-RTS/Zero-K-Infrastructure/wiki/Initial-map-pool-proposal

I think i'm going to randomly kill maps i don't like from the featured pool when i get grumpy at something.

I kinda agree with the OP.
There are many maps that are technically very good, look great, and are balanced but just don't play well, and because of that are unpopular. Especially maps that are too big and laggy for actual teamgames. I agree with most of the list, but there are some exceptions so if we decide so trim some maps, there would definitely have to be more discussion.

Agreed that Industrial Revolution is definitely one of the maps I wouldn't touch.

quote: It's also worth considering filesize of the individual maps. Filesizes vary wildly depending on how aware the mapper was of Spring's tile-engine... although paying attention to the tiling engine is one more step in an already arduous process of making a map, and programs like WorldBuilder and L3DT don't make the maps in a such a way that they can be optimized for tiling.

You are somewhat incorrect. Map size depends mostly on the size of compression (that is like you said, tiling). The more compressed it is, the more tiled it gets but because of that also artifacts start to appear which is obviously not wanted. (look: Red Comet).
In addition nowdays it is common for maps to include normal, detail, specular and emmision maps to acheive superior looks. These usually take quite a bit of space.

This is a good example of lack of knowledge amongst admins who have rights to influence mappools' contents while not entirely knowing the processes behind creating maps. Perhaps then such function should be restricted to trusted people who know ins and outs of mapmaking such as Anarchid, Klon and perhaps Sprung (I would add Saktoth here too but he has not been around ZK for a long while).

I was under the impression that "featured" as a category was going to simply be "the maps used for automated matchmaking" (once that is finally done/public). If that's the case, there shouldn't be more than 25 or so across all game types, and no more than 12 in any one game type.

Shadowfury333

quote: I was under the impression that "featured" as a category was going to simply be "the maps used for automated matchmaking" (once that is finally done/public). If that's the case, there shouldn't be more than 25 or so across all game types, and no more than 12 in any one game type.

This depends on whether ZK is supposed to be about developing strategies unique to each map or being able to adapt to all kinds of maps. I'm not even close to being a good player, but for me it seems like most good players are able to adapt their strategies/cheese to different maps. Compare that to a game like counter-strike where not knowing a map will get you rekt(tm). Limiting the map pool for such a gameplay will also lead to map specific build queues, expansion timings, terraform spots & co. I don't think that zk maps are refined enough to make such things interesting and would thus suggest including all fair and playable maps for each category.

If it's just about making it easier for new players to get in a game, such a restricted map pool makes sense. But as Skasi said, it will drive many seasoned players away from matchmaking. Still, many players might not want to play on a completely random map, which is why some filtering options would definitely be helpful. It could be like the existing map search, allowing players to select maps with tags like "non-sea, funny, hilly, tiny". Featured would then be the default tag that allows newcomers to get in a game without downloading anything.

[2up]knorke I just spent 10 minutes watching this gif. Didn't expect the left machine's battery to run out so quickly.

quote: both have elements that are not supported in the engine version that ZK operates on atm

101.0 is going to be released before ZK will be released on steam, and while I don't decide that, I have reasons to believe it will be picked up by ZK soon-ish. (fixes nasty bugs like the LOS offset for instance)