Spring RTS Engine

I really don't like it that the same model is used 3x, really, it really, really, REALLY sucks, it just makes it feel a bit more fake

it can get pretty annoying telling them appart


A bigger issue with the spherebot is that its so detailed that it looks weird in normal playing distance, I think it only really works on the warrior atm.

Hoi wrote:I really don't like it that the same model is used 3x, really, it really, really, REALLY sucks, it just makes it feel a bit more fake

I think the opposite, personally. Imho, it's more realistic that, for similar roles, similar structures would be used. The problem for me is a gameplay one of unit distinction. Give SAM sphere black or silver shoulder pods or something and it might be a bit better, for example.

the black guns and white missilepods are pretty easy to spot in the thumb

Machiosabre wrote:it can get pretty annoying telling them appart


A bigger issue with the spherebot is that its so detailed that it looks weird in normal playing distance, I think it only really works on the warrior atm.

erm you aren't the first one who says something like that, but ... WTF are you blind!? the warrior is 2x as big! how can't you telling them appart?

obviously not the warrior, genius.

Now put an old jethro in that picture at the same zoom level and resolution and see if it's any easier?

Thing with that goofy shit is that the units are not really that well designed, the morty, pyro, ak and peewee are very similar to but they are obviously different to most ta vets. It is really fucking dumb that people argue that. You can only tell them apart because you learned to. We go around and around with the unit differentiation discussions and in the end it is really that you need to learn the project. Each time someone picks up a new project they bitch and moan about what is what, that is normal. All games are like that when you have a large unit count.

In many games the unit count is very small on a per faction basis. However, with spring projects, generally we have a large variety of units. This means that the learning curve for unit differentiation is much steeper. Add to the fact that you can have 100s of the units on the screen and variable zoom levels it makes unit differentiation harder.

Now some people argue that colors can help, when you have >100 units that many randomly colored units would look like a prismatic spew on the screen. The human eye is only able to process soo much at once and colors can add to the confusion because of the way our eyes focus on colors. Along with people having different perceptions because of colorblind and monitors/lighting fucking with the way they see colors.

In general I do not like the ta texturing style it is confused with lots of random panels and most of the units are very boxy. Sure something like the sumo stands out when compared to the ak. However, we are not talking about sumo vs ak.

ultimately no matter what the modelers do, if it is not DRAMATIC most of you will argue that it ruins what was done with the TA units. However, if all units are dramatic it becomes visually confusing. Players have to learn the visual patterns. Some are easier than others and it is important to try. Honestly, with this high a unit count it is hard to do at all and I sooner would rather poop on my desk than tell you that the players will accept ANYTHING new.

Distance recognition has always been my number 1 priority. I agree these models arent quite up to scratch on the distance recognition, but i tried to make them relatively distinct. The animation and size difference on the spherebot/spherewar is enough that you can really see it. The spherejeth uses the same walk anim as the bot though, and i really do think it needs something more to distinguish itself. The big white shoulders are pretty obvious (the other two spheresbots shoulders are teamcoloured), but while Smoth is right that its not hard to learn very quickly to tell them apart, i want it to be as absolutely easy as it can be. I want the minimum amount of your brain as is reasonable to be taken up telling units apart. Its that sort of split second delay in recognition or catching something out of the corner of your eye which is the difference between winning and losing in competitive gaming. You should be able to tell with the most cursory of unconscious glances whether a unit is AA or raider (big distinction!). Even regardless of the competitive thing, it also helps make the game easier to learn.

On using the same model over and over- id rather not. But when i made half a dozen chicken units in just in a couple of days by butchering up KDR's model, each using almost identical walk anims and textures, and yet each dead easy to tell apart, i thought 'Hey, perhaps i can apply this technique more broadly'. Sure id rather create similar, stylistically coherent but unique models. Its just that, even more than that, id rather be IP free with decent looking models and save myself a lot of work.

Smoth: Im afraid that texture was done by Wolf-In-Exile, he also edited the model a little and re-uv mapped it. Id like to get the source files off him some time (be good to have them, open source project and all) but you'll have to ask him yourself.

Here is the current Spherebot texture 1 and 2.
They were heavily edited by jK, a lot of quality was lost, but it improved distance recognition and coherancy with other models. Ideally i'd like to see a job done on them that doesnt lose so much of WIE's detail work but meh thats what you get for good distance recognition.

Here is WIE's origional textures 1 and 2, which if you want to edit them, you should use these instead of jK's since jK lost so much information (though any end result should, ideally, have as good distance recognition as jk's edits).

Look at arghs new circular widget, if something like that was in CA BUT, different shapes for different units, it would vastly increase the users ability to recognize units at a distance by increasing the shape variety factor of units without affecting their models.

AF wrote:Look at arghs new circular widget, if something like that was in CA BUT, different shapes for different units, it would vastly increase the users ability to recognize units at a distance by increasing the shape variety factor of units without affecting their models.

In that case, what are the models for? If I'm going to spend time focussing on the little circles and squares and whatnot, why bother making the models themselves?

They are not the only way of identifying a unit, otherwise we would be using your logic to make everything utterly pointless and we would have dots on a heightmap in grayscale.

1. The method I used would not be suitable for the purpose of identifying everything very specifically, for technical reasons.

2. I've done a few experiments, and using a lot of different types is visually confusing, imo.

I could see doing a "Core style" vs. an "Arm style", but you would not want a "Thud style" vs. "Hammer style"- it'd just get too confusing, and distract players away from what matters, i.e. the actual shapes of the units themselves. Just imo, of course, but you get to a point, with icons or other identifiers, where you're basically just raising the visual complexity, not simplifying things for players, which is bad practice.

Hoi wrote:

[16:51:37] <Springie> [drawesome]why did my comm just die?
[16:51:48] <Springie> [[LCC]jK]you scraped it

Yeah. Seriously the scrap button needs to have a left-click, right-click mechanism or soemthing. Over and over again people scrap things they don't want to, generally cons/coms. It sucks.

[are you sure you want to scrap (commander)/(these 15 units)?]
[Y/N]

Well there is a countdown with dramatic chat-spam then you start scraping something, and I think thats enough. I think that having a confirmation box or needing to give a special click ever time I clicked the scrap button would be anyoing...

All I ask is that jethro get a pair of arms. The spherebot family looks human, and armless jethro looks like an amputee. Otherwise awesome models.

And scrap is definitely something that needs work. Accidental scrap sucks, and waiting for scrap sucks, because you can't reclaim living units so now reclaiming your existing structures takes a click, a 5 second wait, then a click. That is like an eternity.

We'll be replacing it with negative-building in the next version of spring (when its possible).

IE you reclaim a friendly structure and it turns back into a nanoframe and goes back up the nanotube.

Hello everyone. Unfortunately, real life strikes again, and so I probably will not be able to do much devving in the near future. Also I can't seem to access caspring.org by browser or by SVN; I suspect my firewall is going psycho, but I've not been able to figure out how to fix it.

tbh you should rebuild the building frame sequence to look better using your fancy lua sfx, it'd make it possible to do before the next spring and make CA stand out more.