Spring RTS Engine

simple rebut: trepan said it could be done but it is expensive.

I wrote a quick gadget that gets the map dimensions, and polls a point of the map every 32 pixels each direction (64 is nearly acceptable, but it fails more on some maps with small patches like Islands at war) and places a flag if there is no other flag within 256 pixels. It works decently and doesn't seem to contribute any major load time, but for some reason it's bombing out the rest of our luarules, so I'm going to have to keep prodding at it.

The most obvious flaw is that automation is never perfect. Different maps have different metal output ratios (what if map X has 10 metal patches that give 30 metal and map Y has 1000 that give 0.3?). The cutoff isn't so simple if you want to support all maps, particularly those that use the 'natural' style like small divide. You could probably derive a formula based on the metal tag in the smd and stuff, or do all the map reading and do a comparison later and determine a cutoff on the fly, then place the flags from an array of coordinates that match the cutoff and the range limitations, or you could do any number of other things but it'd never work right for every map. Personally, though, I don't give a flying fuck about metal maps

I think if we could have access to the same routines or whatever that the AIs use to place metal patches, it'd be much more feasible, but I imagine there'd still be similar problems.

The best solution, I think, would be to use the automatic placing sequence as a last-ditch fallback routine, not as the primary means of playing the game. The script could check for a metal coordinate profile of some sort which the mapper would create in a similar way to how they lay out start points. It'd basically be asking the mappers to write down all the strategically interesting positions on the map--they wouldn't even necessarily need to correlate to the metal map. If that file is missing, then the script would just wing it and pre-place stuff.

Anyway, I do think it's worth looking into further. I think the weakest part of SWS .. sorry, SW:IW now is that you have to build the flags to begin with and then capture them the rest of the time. It's particularly bad when two people build flags on the same largish metal spot right next to each other. Taking the user out of that part is definitely the best action, IMO.

Gnome wrote:The best solution, I think, would be to use the automatic placing sequence as a last-ditch fallback routine, not as the primary means of playing the game. The script could check for a metal coordinate profile of some sort which the mapper would create in a similar way to how they lay out start points. It'd basically be asking the mappers to write down all the strategically interesting positions on the map--they wouldn't even necessarily need to correlate to the metal map. If that file is missing, then the script would just wing it and pre-place stuff.

+1

I suggested something like this in a different context and people didn't like the idea. It would also allow the SW:IW team or the community to release these text files or whatever for any popular map for which the original creator fails to create one.

As discussed in IRC before South Africa's useless internet kicked me off, I think that would be a good solution, as it'd allow us to create custom profiles for the most popular maps, or maps that fall down with the auto-spawner, but still let us provide working solutions for difficult maps such as Small Divide and other island maps.

Would these profiles need to be packaged with the map smd's, or would we be able to simply make the profiles ourselves for people to download later, independent of the maps?

AF - I wouldn't consider it a failure of the system if four flags were spawned in the centre of small divide; it is a pretty important area, and something we'd want the game to focus on. However, with the above system, we could always provide a custom profile for it if it was clear it wasn't working, which is why it seems to be a robust system worth pursuing.

Zsinj: You only said the last half of a conversation somewhat unrelated to what your post is about on IRC

I don't think the profiles would be necessary in the map archive itself (though it would obviously make packaging easier) but the mod would need to include them or include another archive with them. Either way it'd just be a few hundred KB of text maximum so the only thing required for that kind of effort is time. Hopefully mappers creating maps in the future would generate their own for maximum compatibility.

Just need to bugfix it first

New Art Preview now that our new site and forums, http://www.imperialwinter.com , are up and running.

Rebel Air Transport



Model by Guessmyname, UV by Maelstrom, texture by me.

As stated in that thread, this unit's current name is only a place-holder, but we're all ears for suggestions!

Actually, Maelstrom had to wrap it because the model kept crashing me when I tried to do any UV projections.

Oh; well, it was quite a good map either way. But so many itty-bitty bits!

godamn, those shots make me want to make a new map with that skybox.

I wouldn't stop you

Imperial TIE Crawler

bloody hell those air transports look incredible

Thanks Ivory - it was an excellent model by GMN.

I personally prefer the Imperial Versions

Warlord Zsinj wrote:Imperial TIE Crawler

Is it just me or is that an utterly ridiculous design? No offense to the recreators, of course

Perhaps, I like it, but the 'crawler' track bits stick out, however looking at the images in the thread I must say I like it a lot.

It is an actual Star Wars design; expanded universe, of course. We did alter it a little to suit our gameplay needs (we gave it side mounted guns like the old WWI tanks)

It first appeared in a comic:

And then was filled out later

You can buy official model kits and lego kits of it.

I personally am not a big fan of it, it's a reasonably silly design. However, it does have the Star Wars flavour, having a TIE in the design, as well as just generally being a strange design you'd just end up seeing in the Star Wars universe.

When I first started texturing it, I was seriously considering dropping it from SW:IW and replacing it with something else. But now that the texture is finished, I guess it's earned it's place. It is star wars flavoured, and reasonably well known, and didn't turn out all that bad.

The texture is really well done. And yes, the TIE-C design is retarded. Oh well, no big.

The Empire doesn't waste time designing a whole new tank when a retrofit that can use existing manufacturing infrastructure instead I suppose... it would help if your model looked more like a tie-fighter... independently though it looks very Cool and StarWarsy so I will accept it.

Wasn't the TIE Crawler in EaW?

SinbadEV wrote:The Empire doesn't waste time designing a whole new tank when a retrofit that can use existing manufacturing infrastructure instead I suppose.

Wow... That was was a really bad sentence.