next »

[Kyzrati]

Just yesterday someone on IndieDB made essentially the same comment as you have about none of the tiles being right for Cogmind... I'm beginning to agree.

These are further from what I originally imagined, which were not true pixel art sprites but instead something more akin to "recognizable symbolic icons," exactly what you've been saying. Perhaps we could say that the whole point of the tileset is to make the game more accessible to those who don't "get" something symbolic and are used to "normal pixel art," but if it doesn't fit the game as a whole, I think that might be the wrong direction to take. I'm considering contacting a few of the more experienced artists and seeing if they can manage something like that instead. Kinda difficult now that so many have applied to do something at least somewhat close to traditional game tiles :/

I think this is what attracted me to koiwai's quick sample, because being familiar with what the game needs, I immediately saw that and thought "stylized test tube for Researcher bot!" (perhaps stylized even further)

Now I'm really going to have a hard time figuring out what to do with this. Maybe I should spend more time creating some sprite samples of my own, and see about getting one of the more experienced artists to expand on the concept and make it into a full set.

About ruining the mood, I think that's why for a long time I've been very reluctant to even consider allowing a tileset at all. ASCII works so well for the immersion/mood. The tiles need to match that. Basically, it looks like we just need "an ASCII that is not ASCII," i.e. something simple and symbolic but not your everyday letters that some people find too familiar to convey the idea that they represent something unfamiliar from the game world itself.

[Kyzrati]

While we're on this topic, what existing sets do you/others think might come close to achieving a more Cogmind look?

How about F? It's already very symbolic (interestingly it's based on my original concept).

K has been pointed out by several as a good option, and it is fairly symbolic, though I don't much like the style myself. For one it needs to be brighter, plus there aren't many movable objects demonstrated there.

M is also pretty simple and symbolic (also based or mixed with my own concepts), though it uses too much shading.

With less shading (and/or just brighter in general) something close to P might work well.

I wonder if it might be too difficult to create something slightly abstract that both makes objects sufficiently different across the multiple tiles and "makes sense" visually.

For some reason I still like A a lot, even in this new context, but I guess it's too detailed to fit the bill at its regular size. Doubled at 24x24 it looks good.

Now I'm going to go crazy over this.

[happymonster]

Basically, it looks like we just need "an ASCII that is not ASCII," i.e. something simple and symbolic

I would do it like this:


Edit: A few tweaks..

[JobLeonard]

If it's driving you nuts, I'd say let it rest for a bit, and let the people who contacted you in on the discussion. Make it clear what it is you are struggling with - it's only fair towards them, really. Let them share their thoughts on the matter with you.

[Kyzrati]

If it's driving you nuts, I'd say let it rest for a bit, and let the people who contacted you in on the discussion. Make it clear what it is you are struggling with - it's only fair towards them, really. Let them share their thoughts on the matter with you.

Yep, that's why I spent a few days coding already and not pushing this except to reply to others' posts. Just letting it rest to see how it feels in a while after everything's sunk in. That and slowly evolve my viewpoint and gauge my opinion of it all in the process.

No matter what, a week ago I told the artists I'd be getting back to them this weekend, and I plan to do that. I'll be 1) pointing them to all the places the concepts are being discussed (though some have already found at least some of the ongoing discussions) and 2) tell them what the next steps will be (specifically, it's going to take longer than I thought, and depend on the outcome of further testing and mockups).

Basically, it looks like we just need "an ASCII that is not ASCII," i.e. something simple and symbolic

I would do it like this:


Edit: A few tweaks..

Thanks for that happymonster. That actually looks pretty good and is akin to what I was planning on trying as a mockup myself soon. Something very simple and symbolic.

Some concerns/considerations I was having about that style (even before you posted):

-Will something like that work at larger font sizes and still look good? Is it simply enlarged with no extra detail? Obviously more tests will show.

-More importantly, is there enough variation possible with a truly simply ASCII-like style at small sizes? ASCII can do it because we can already recognize and distinguish glyphs that are familiar to us--that's part of the reason it works, yet also a reason to not use ASCII itself, according to some players.

-To effectively create a symbolic set it takes a pretty close understanding of everything the game has to offer, and for that reason I'm the best equipped to do it at this point... I may make my own representative mockup that sufficiently and realistically replicates a common game situation, and then substitute the more promising concepts and some other symbolic approaches for a direct comparison.

[TheWing]

Really solid and well designed ascii-stuff!

I'm only wondering if this can be played in an actual terminal.. would probably limit the amount of eyecandy but I'd like to be able to play this on my secondary computer which has no graphical interfaces :D

[Kyzrati]

Really solid and well designed ascii-stuff!

I'm only wondering if this can be played in an actual terminal.. would probably limit the amount of eyecandy but I'd like to be able to play this on my secondary computer which has no graphical interfaces :D

Thanks! No true terminal support, no :/. I do get this question occasionally, though . I've imagined seeing how far I could take an actual terminal interface (as an experiment), but I don't have the time for something like that these days. The only thing I can say about the simplicity of running Cogmind is it doesn't care about your video card--software rendering only.

Your secondary computer has no graphical interfaces? What is it? If you're looking for a cool modern roguelike that works in a terminal, I highly recommend The Ground Gives Way.

[TheWing]

Thanks! No true terminal support, no :/. I do get this question occasionally, though . I've imagined seeing how far I could take an actual terminal interface (as an experiment), but I don't have the time for something like that these days. The only thing I can say about the simplicity of running Cogmind is it doesn't care about your video card--software rendering only.

Not a problem, I do know it would add quite some stuff to worry about.. and take time to develop keeping that in mind too. AND the worst part, you'd have to say goodbye to all those fancy particlethings!! (or at least their colours and smooth animations)

Your secondary computer has no graphical interfaces? What is it? If you're looking for a cool modern roguelike that works in a terminal, I highly recommend The Ground Gives Way.

Yeah it's an asus eeePC 901, but the battery's so worn out that I decided to take X off its linux :D I use it mainly for taking notes and IRC, so there isn't much more I'd want to do with it. Except some games like these!

I did install The Ground Gives Way a while ago, but have yet to dive into it.

[happymonster]

Thanks for that happymonster. That actually looks pretty good and is akin to what I was planning on trying as a mockup myself soon. Something very simple and symbolic.

Some concerns/considerations I was having about that style (even before you posted):

-Will something like that work at larger font sizes and still look good? Is it simply enlarged with no extra detail? Obviously more tests will show.

-More importantly, is there enough variation possible with a truly simply ASCII-like style at small sizes? ASCII can do it because we can already recognize and distinguish glyphs that are familiar to us--that's part of the reason it works, yet also a reason to not use ASCII itself, according to some players.

-To effectively create a symbolic set it takes a pretty close understanding of everything the game has to offer, and for that reason I'm the best equipped to do it at this point... I may make my own representative mockup that sufficiently and realistically replicates a common game situation, and then substitute the more promising concepts and some other symbolic approaches for a direct comparison.

I see it as not adding extra detail as you would for a sprite, but refining the shapes of the glyphs at the higher resolutions. For example at small sizes there are not a lot of different ways to do a upper class 'A', but at higher resolutions you can have more space to have smoother lines, line thickness can vary, as can more accurate placement and maybe the presence of serifs can be refined. That's the way I would approach it.

I think you are absolutely right that you are the best person to do a mockup as you have the best understanding of your game.

I'm looking forward to seeing what you come up with!

[Kyzrati]

I see it as not adding extra detail as you would for a sprite, but refining the shapes of the glyphs at the higher resolutions. For example at small sizes there are not a lot of different ways to do a upper class 'A', but at higher resolutions you can have more space to have smoother lines, line thickness can vary, as can more accurate placement and maybe the presence of serifs can be refined. That's the way I would approach it.

That is an excellent analogy. Thank you. I can see how that would work, it would just need a truly "stylized" type of style, which could theoretically fit Cogmind well.

I think you are absolutely right that you are the best person to do a mockup as you have the best understanding of your game.

I'm looking forward to seeing what you come up with!

The whole point of hiring an artist was so that I wouldn't have to spend a bunch of time doing pixel art . Wouldn't it be hilarious if I go through all this and then just end up doing it myself (extremely unlikely, mostly because I don't want to do the larger pieces, but you never know...)

[happymonster]

 

[happymonster]

BTW: I hope one of the glyphs is a Cog

[Kyzrati]

Well, that is the most logical way to represent the PC, and is sorta what I was going for with the semi-stylized 'C' thing in my original 7DRL sprites. Also, that whole first group of sprites in the 'L' concept were intended to be concepts for Cogmind, which I thought were pretty neat.

[UmutD]

In Ascii mode, it is a bit hard to read the walls (as in, when you zoom out and try to figure out your overall situation), not the items/enemies.

In all tilesets, it is the quite opposite (due to 10x10 size being too small). It's hard to distinguish items/enemies in 10x10, besides most of them don't feel "cogmind" as people said (though they are all beautiful in their own way, props to the artists)

I can't imagine this will be super useful since most players will pick a mode and stick to it, but you can swap between ASCII and tiles on the fly, and of course it's animated :D



(This is a quick tileset I made one day a couple years ago for the prototype.)

Why not take the wall/block/non-movable item textures from here, and keep all items/enemies as ascii symbols? That would be enough for me. I think, in such a game that has a lot of variety, designing new, small and legible symbols is both hard and making a little difference even for the players who won't like ascii anyway.

As I said before, I might prefer a tileset, but that is only to improve legibility of the space and increase situational awareness when playing the game. I already like the aesthetics and I believe it would be hard to introduce a new tile-set that doesn't change this visual feel, especially if the tile size stays this small.

[Kyzrati]

Why not take the wall/block/non-movable item textures from here, and keep all items/enemies as ascii symbols? That would be enough for me.

The reality of the broader market (beyond those already familiar with roguelikes), is that using letters at all is a big turnoff. I'm not sure (though I'm very curious to find out) that Cogmind's aesthetic is strong enough to attract these people as well. Plenty of people have told me they don't normally enjoy ASCII but would definitely make an exception for Cogmind--but these are people who already *know* what ASCII even is--that's not saying anything about potential players beyond that. Maybe more market research? (Or are those players possibly unattainable anyway?)

I think, in such a game that has a lot of variety, designing new, small and legible symbols is both hard and making a little difference even for the players who won't like ascii anyway.

This is what I'm worried about. More tests are needed...

I should point out that while I almost always show the game using rather small 12x12, which is what I designed it in/for, many players will be able to play with larger tiles--whatever fits their desktop. So there is a *little* more room to work with. (10x10 is quite likely to be too small to make an effective ASCII for, which is why I've been under the assumption that mode will likely require ASCII, which still works fine even at that size.) At those larger sizes it might possible to create interesting, readable sprites that reflect the detail, but still:

I believe it would be hard to introduce a new tile-set that doesn't change this visual feel, especially if the tile size stays this small.

I think this is a key point, and one which is making the decision much more difficult.

About the walls being hard to read, that can be adjusted if necessary later (simply thicken the '#'), though I find that thicker hashes tends to draw too much attention to the walls. Combining ASCII and sprites in that way is a pretty interesting idea, though...

[oodavid]

Since you can select between themes already (tiles / ASCII) would it not make sense to let people be able to skin it easily and hotswap within the game? That way it would free you up to keep working on your side of things while letting artists have free reign to try out new ideas and concepts...

[Kyzrati]

Since you can select between themes already (tiles / ASCII) would it not make sense to let people be able to skin it easily and hotswap within the game? That way it would free you up to keep working on your side of things while letting artists have free reign to try out new ideas and concepts...

Yep, that's already possible--hot-swapping has been implemented. Modding in and enabling new tilesets is also incredibly easy. All that needs to be done is to drop sprites into prepared templates.

The main problem is where "my side of things" stands right now :D. I'm actually working up to an alpha launch in a couple months, and figured it makes sense to have the tiles ready for that version. But an artist can't really move forward until I give clearer guidance, which at this point is apparently going to require a greater time investment on my part :/.

Didn't see that one coming

[oodavid]

The main problem is where "my side of things" stands right now :D. I'm actually working up to an alpha launch in a couple months, and figured it makes sense to have the tiles ready for that version. But an artist can't really move forward until I give clearer guidance, which at this point is apparently going to require a greater time investment on my part :/.

Didn't see that one coming

I see you've adopted the stone soup style of project management!

TBH, looking at the art submissions a page back it does appear people are willing to inject their own spin on things, unless by "guidance" you mean storylines, themes or whatever. Personally I love the ASCII style, not because it's kitsch, but more that you're able to make it look so goddamn dynamic and pretty!

[happymonster]

I think an alternative ASCII set based on symbols rather than detailed sprites is perfectly possible. It should take less time too.

[Kyzrati]

TBH, looking at the art submissions a page back it does appear people are willing to inject their own spin on things, unless by "guidance" you mean storylines, themes or whatever.

Guidance as in more or less the ideas we've been discussing here, in terms of art direction (less pixelly, more icony). That said, I think an artist can definitely do a better job once they do understand the story and content better.

Another incredibly important thing I failed to mention in the original ad was the fact that Cogmind is supposed to be an immersive experience where you "are" the robot. In this way the game really embraces the ASCII aesthetic, and any tileset should both maintain that feeling and mesh with the rest of the UI without appearing too gamey.

I corrected that oversight last night in an e-mail to a subset of the artists, and hope to further hone the concept with some tests of my own before deciding the next step. This includes possibly larger mockups using the submitted concepts, just to see, but I think an icon-like approach will be more successful.

I think an alternative ASCII set based on symbols rather than detailed sprites is perfectly possible. It should take less time too.

Less time is good :D. Heck, if it's less expensive I'll pay multiple artists to do complete sets.

I see you've adopted the stone soup style of project management!

How so? I'm sorta familiar with DCSS development, though I never followed it closely enough to know much about their project management.