Any news on the possibility of a Steam version

[Pereza0]

I am subscribed to the mailing lists, and there was talk about this a few months ago and I was just curious if any progress has been made.

[Sar]

paid mods failed so they won't be able to sell mountain dwarf mod for it, which means there's no point in putting it on Steam anymore

[dowan]

What would be the benefit of this game being on steam? Aside from the obvious benefit of achievements popping up I mean.

Achievement Unlocked: Babby's first meal: YOU ATE SOMETHING! 20 points!
Achievement Unlocked: Optimal Player: You lured monsters back over 200 tiles in a single game! 50 points!
Achievement Unlocked: Cheater!: You won a game in version 16.0 with the double damage bug! -100 points!

I haven't used steam in a long time, so this may not be an accurate depiction of how their achievements actually work...

[nago]

No idea why, but afaik Tome on steam is quite popular

[PleasingFungus]

What would be the benefit of this game being on steam?

Wider exposure, mainly.

The main problem is that we don't have any real development effort put into the offline version, so recent versions have been fairly broken (lots of issues with modifier keys 'sticking', visual lag, stuff like that). We'd prefer not to launch until we've got things cleaned up there, which isn't something any active devs actually have enthusiasm for, so... probably needs some enthusiastic soul(s) to come in and clean stuff up.

It's hard because our community is so strongly focused on the online version - who cares about a version of the game which only a 'silent majority' plays...?

e: apologies for seriousposting in cyc

[Pollen_Golem]

Maybe there ought to be a subforum for serious, miscellaneous topics.

[Sandman25]

The main problem is that we don't have any real development effort put into the offline version, so recent versions have been fairly broken (lots of issues with modifier keys 'sticking', visual lag, stuff like that). We'd prefer not to launch until we've got things cleaned up there, which isn't something any active devs actually have enthusiasm for, so... probably needs some enthusiastic soul(s) to come in and clean stuff up.

Is there an easy way to find those problems in mantis? I didn't know offline is so broken, it always looked faster to me.

[Pereza0]

What would be the benefit of this game being on steam? Aside from the obvious benefit of achievements popping up I mean.

Achievement Unlocked: Babby's first meal: YOU ATE SOMETHING! 20 points!
Achievement Unlocked: Optimal Player: You lured monsters back over 200 tiles in a single game! 50 points!
Achievement Unlocked: Cheater!: You won a game in version 16.0 with the double damage bug! -100 points!

I haven't used steam in a long time, so this may not be an accurate depiction of how their achievements actually work...

Basically, being more easily accessible to the public

People very often ask for good roguelikes, and usually the will have checked out the ones... on Steam - often a mixed bag - because that is where games usually are. Finding those games on the "centralized" library Steam is is a lot easier than finding DCSS, even if they have half the following and quality and they are actually paid for. Also, having a Steam version could provide stuff like DCSS steam groups, actually getting an idea of how many people play offline through it... and maybe it could be cool to implement a sort of offline shared morgue (though it has potential to get abused I must admit)

Some people enjoy Steam achievements, but I don't think they are a priority now or something people care about that much. And how achievements work depends on how the developer design thems basically, there are good examples and bad examples closer to (and worse than) what you parodied.

What would be the benefit of this game being on steam?

Wider exposure, mainly.

The main problem is that we don't have any real development effort put into the offline version, so recent versions have been fairly broken (lots of issues with modifier keys 'sticking', visual lag, stuff like that). We'd prefer not to launch until we've got things cleaned up there, which emergingisn't something any active devs actually have enthusiasm for, so... probably needs some enthusiastic soul(s) to come in and clean stuff up.

There might be some issues, but I wouldn't consider it broken.

Keep in mind, we are talking about an excellent game (lots of content, challenging, free...) in a platform where completely borked incomplete paid-for games have regularly reached the top sellers list (and many people are not happy about this stuff). DCSS (even as it currently stands) will be a breath of fresh air in a marketplace filled with incomplete and broken games.

DCSS will be successful if it releases on Steam, don't doubt it for a second - if DoD managed to do well there, I expect DCSS to be a massive hit. And I am pretty sure it would be easier to get people motivated to improve the offline client if they saw how many people were actually using it.

[n1000]

Is there an easy way to find those problems in mantis? I didn't know offline is so broken, it always looked faster to me.

I believe some (maybe most?) of the major problems are with SDL 2.0 and mouse behavior.

If you look through open bug reports with severities of 'Block,' 'Crash' and 'Major' (there are only a few pages of open ones) you'll be able to see plenty which only affect the offline DCSS client.

https://crawl.develz.org/mantis/view.php?id=9559 is one I've seen discussed quite a bit lately. @Pereza0, this bug by itself is probably enough to make a Steam release uncomfortable.

Anyway, I'm sure if anyone took it upon themselves to tackle any of this sort of bug it would be greatly appreciated.

[dowan]

I didn't realize the devs were so exclusively devoted to the online version of the interface. I suppose it makes a lot of sense, since they're just pulled from the online community and the tavern, which has a lot of online players.

As a technical question, what made switching to sdl v2.0 worthwhile, when it seems to have caused so much trouble?

[n1000]

I imagine the Android/iOS support was at least part of it.

[bcadren]

I'll give this a serious answer too; though it's from my own perspective.

One of the first Roguelikes to sell on steam (Nethack) was a completely different version on Steam than in other locations. Nothing was gameplay different; but it had a unique interface and completely new polished 3D graphics (instead of the original console).

I'd be willing to work on such a project as a 3D artist; though there would need to be quite a few of us to get such a thing done in any sort of timely manner. [There are 26 player races and around 200 monsters, plus weapons and actual dungeon stills; lots of resources to model.]

And, we'd need someone literate enough in the code to either make:

1. Crawl run as a background daemon and give us current screen data to parse into new UI (IE Crawl Console has a tiny patch that passes the necessary information to a new application that just displays graphics).
   OR
2. Make a fork that properly displays mesh data and creates necessary renders from either nothing or imported library.

I -am- a programmer, but I don't know my way around Crawl's source and what's feasible with it at all; nor does writing a graphics shading algorithm all the way from 'display a triangle' level sound fun. (Display a triangle is the mesh renderer version of 'hello world'). And the fact that none of mainline crawl has anything remotely traditionally polished [I mean drag-and-drop mouse control; cool UI for spells and inventory, etc.]

In short; I'm proposing selling a 3D game with music on Steam, which -is- crawl (Web Console) under the hood and simply a UI and graphics patch.

Off-Topic:

Spoiler: show

I'm going to agree with Pollen Golem - We probably should have a separate forum for serious talk that isn't on design discussion of mainline crawl, competitions or giving advice on playing mainline crawl. Bumping all miscellaneous to one forum would be fine if it wasn't the "Thank all jokes, try to derail the topic" forum.

[Pollen_Golem]

3D sucks. Long live 2D.

Crawl can be made very presentable and professional/modern-looking with a little free-domain music, some sound effects, a better-scrolling message bar, and some graphical touch-ups: briskly animated menus, portraits in the char-gen screen, fewer full-screen black backgrounds, etc.

My poorly informed opinion is that if devs don't want to do it for free, it can be outsourced for less than $10k.

[rockygargoyle]

I didn't realize the devs were so exclusively devoted to the online version of the interface. I suppose it makes a lot of sense, since they're just pulled from the online community and the tavern, which has a lot of online players.

As a technical question, what made switching to sdl v2.0 worthwhile, when it seems to have caused so much trouble?

According to sgrunt the code was a mess, so he had to rewrite most (if not all) of it.

[dowan]

Once again, +1 to the idea of a serious proposal but not GDD forum. Or, failing that, at least asking that posters stop trying to derail CYC threads so often.

Oh crap, I'm derailing this cyc topic.

[Pereza0]

Oh crap, I'm derailing this cyc topic.

QUICK

SAY RANDOM THINGS TO GET IT BACK ON TRACK

HOO HOO LALALALALA

[Rast]

What would be the benefit of this game being on steam? Aside from the obvious benefit of achievements popping up I mean.

Achievement Unlocked: Optimal Player: You lured monsters back over 200 tiles in a single game! 50 points!

Not even close to optimal. Maybe luring a single monster 200 squares.

[Pollen_Golem]

Here are some candidates for the fifth Tavern subforum: "The Ecumenical Temple" "The Bazaar"

[PleasingFungus]

As a technical question, what made switching to sdl v2.0 worthwhile, when it seems to have caused so much trouble?

I imagine the Android/iOS support was at least part of it.

That might have been reasonable, but IIRC wasn't really discussed. I remember we kept running into issues caused by SDL 1 that we wanted to move to 2.0 to fix. Our shift to C++11 was also an inspiration in some way, I think, though I don't recall how.

[bcadren]

According to sgrunt the code was a mess, so he had to rewrite most (if not all) of it.

I started to write a review for Crawl as if it was a modern release title at one point; I stopped because it would have taken -far- too long. Most the issues that keep it from being suitable as a modern game have to do with UI though. Crawl is way belong minimum viable product to be sold modernly, because of the sheer menu density and, to a lesser extent, art quality. To the average modern player; even if we optimized with the best artists out there the current -style- tiles is done in would be seen as 'pixelated' and 'dated' and not in the good 'retro' way, just dated. From my own sense with limited pixel art experience, it looks too muddy (too many colors used, doesn't have the crispness of someone intelligently using color theory and limited palettes on every creature.

3D sucks. Long live 2D.

I love good sprite art myself, but I think crawl's full style would have to change regardless; the current tile style is made to line up to console, which doesn't look as good as a perspective topdown like Legend of Zelda titles. Honestly to do it -well- the better call would probably be 3D for a number of reasons.

• Even excellent pixel art will get shot down as 'pixelated' by a lot of common folk. See this excellent article
  • the only non-pixel art option for 2D would be to port everything into flash and rely on vector art; which is a pain.
• Socketing weapons and weapons into multiple kinds of humanoids without literally having the same pixels is FAR easier.
• If applying actual animation (which would be expected, but not necessary), having different creatures repeat similar animation data would be easier.

That said; I did some numbers. Sheer number of monsters there are in crawl would mean that repolishing JUST THOSE without worrying about items or world tilesets would take more people to do than there are currently involved in development total. Quick List of Creatures in Crawl -Categories are based on how much artwork would be necessary to create.

[Pollen_Golem]

Even excellent pixel art will get shot down as 'pixelated' by a lot of common folk. See this excellent article

That article is for pixel art. Crawl tiles are created as pixel art, but that's not how they're perceived. They're not magnified at all, so you don't see individual pixels as squares. I suppose crawl tiles can be magnified for comic effect on some screens, like below, for example when you see some kind of enemy for the first time ever. Otherwise, I don't see the relevance of that article.

[archaeo]

Eh, I feel like Crawl looks better than a lot of roguelikes you pay money for; I'd take Crawl's graphics over ToME or the new and atrocious ADoM anytime. Or even the really gross-looking Vulture's Eye stuff for NetHack.

[ontoclasm]

See this excellent article

I've read this article and personally think it's terrible. The guy who wrote it did so based almost exclusively on negative feedback he got for this game:

What I get out of that is not "pixel art is a dead end, nobody understands it :C" but "that game's pixel art is mediocre."

And in any case, let's assume he's right and that pixel art is viewed as weird and ugly by all but a few hardcore fans. So what? A multi-million dollar AAA studio produced this hideous abomination in the year of our lord 2014. Look at that thing. It's a pile of mush with the blandest color scheme imaginable, no concern for silhouettes or focal points, and higher contrast on the background trees than the things that actually matter to gameplay. And that's been carefully posed and edited (again, by professionals at great expense) to look as good as possible; in motion it looks far far worse. The fact that frat boys look at the dull grey lump of a grenade launcher and go "whoa, so tactical bro" does not impress me all that much.

If we could magically convert DCSS to look like CoD:Ghosts, would it have a larger audience? I dunno, maybe. I doubt said frat boys enjoy roguelikes a whole lot. But while it might look more "modern," I seriously doubt it would actually look better.

[Arrhythmia]

but what about the part where it would take a team of more than a dozen dedicated artists to "repolish" crawl's monster tiles

[Pereza0]

Yeah, that article was quite silly.

I personally don't dislike the art in Auro, the animations look cool I guess? - even if the world is boring and featureless. But it is a terrible fit for that kind of game. Huge characters and huge tiles. Then again, a mobile game so maybe there isn't much option.

Either way, it is silly to blame pixel art for your failures when so many games with pixel art have been a resounding success - try to see what you could have done better but don't blame the world (YOUR CONSUMERS!!!) for your failure

EDIT:

Also, wanted to add. Expecting effort, talent and originality to grant you success in a market where Flappy Bird and Clash of Clans trump everything is stupid. Competition is fiercer on PC when it comes to "hardcore" games, but if your game is good you will have an audience that actually pursues the kind of games it enjoys instead of just playing whatever is hot at the moment

And yes, this is blaming the audience. But I am just a guy on the internet, not a developer

[bcadren]

but what about the part where it would take a team of more than a dozen dedicated artists to "repolish" crawl's monster tiles

You know what, no matter; what I try I get made fun of. I' a bit tired of it. I'm not a pansy or an idiot; I don't have to take it. I have a history of chronic depression and I don't -want- to take it, so stop.

Anyways;

I am better as a 3D artist than a 2D one, so my instinct is to redo it that way. If you want a real reason; I feel like it'd be easier for humanoid enemies to display what weapons and armour they are wearing in a 3D system than a 2D one. Model the weapon once socket it into the enemy and have various positions on different enemies. I could make a couple models to demonstrate, but ultimately there aren't enough people to redo that, so it's probably pointless. IDK a blademaster with two differently branded weapons that have visual effects for their brands could make a cool splash screen, even if a 3D render would look out of place compared to the current splashes. I can easily see 3D conveying all the information we try to cram into one tile without need of the detail cramming. Show that the creature has an electrocution weapon by having sparks coming from the weapon is perfectly reasonable in 3D while there are no pixels for that and it's a very difficult effect to make look reasonably good at the pixel density we are currently using.

Honestly; my conclusions from the article I linked are less 'pixel art is dead' and more 'pixel art shows your hand; your audience will be able to tell how your piece was made, thus you have to pay careful attention to composition to make that into a good thing and make sure every stroke is exactly as intended.'

Crawl, currently doesn't have very good composition (similar point to what ontoclasm made about Auro). Colors used don't necessarily match well; artstyles of various enemies don't necessarily line up and the reigning artstyle gives the impression to trying to express too much data in too few pixels rather than any kind of stylistic impression that draws the world together. We correct blatantly bad problems (player ghosts being nearly invisible on certain backgrounds for example) but different color densities and detail levels live on.

The best sprites in crawl are non-humanoids; they don't have to worry about having a place to socket in a weapon and have a lot more freedom; the Chaos Spawn (ontoclasm's personal image) is probably one of the best because not only does it convey what the creature is, but it looks interesting and unique.

It'd be hard to correct crawl's composition simply because there are so many assets and there isn't any dedicated art direction (closest we got is onto and Bloax and they are more correcting very bad sprites than trying to make the sprites all match a style (despite being the same resolution and color count, a Chaos Demon looks crisp and otherworldly while a polar bear looks realistically shaded and muddier). Neither sprite is bad (bears aren't great, but they aren't bad), but they are pretty well different styles.

In the end, I feel like it's a bit odd; gameplay-wise crawl tries to be the most polished Roguelike out there, Tile-Wise, whatever is good enough goes (it's a hell of a lot better than console, but not an excellent enough style to draw in players).

IDK. I'm trying to be as honest about it as I can. I am an artist and trained enough to do some work for 3D (I don't trust myself as well with pixelart, though I'm not inept), but I'm not an art director and I don't really know how to fix perceived issues beyond the current "remove really bad problems" rule. It's hard to have one defined style when the same zombie can show up in the browns and greens of the swamp and in the stark contrasting red of Gehenna.

[Hirsch I]

everyone makes fun of everyone around here. I really do not see how you would think it is something personal. but that's beyond the point.

I sincerely believe that having your offline games matter, for tournaments, highscores etcetera, would be a great leap for Crawl. we would have the vocal minority having their fun playing online, but reporting bugs that affect the silent majority of offline players.

[Rast]

We would have people cheating like crazy. Not a lot, but some.

[Hirsch I]

oh, yeah, I guess. that sucks.

[all before]

Why make crawl's visuals more like other games? Compared to all the emphasis on realism and immersion in AAA games, I like how abstract, austere, surreal, and silly dcss's visuals are.

[Sar]

The problem with Crawl visuals is that they're made by very different artists using very different styles. They are very very inconsistent.

I wish Denzi made more tiles. I liked his/her ones the most.

[savageorange]

The problem with Crawl visuals is that they're made by very different artists using very different styles. They are very very inconsistent.

I wish Denzi made more tiles. I liked his/her ones the most.

Which graphics are you referring to?

I find the background tiles and item tiles generally good. The player tiles are a little inconsistent. The monster tiles are most inconsistent (particularly in level of realism)

Even excellent pixel art will get shot down as 'pixelated' by a lot of common folk. See this excellent article

This article (and the artist's art) has been discussed by many people on Pixelation, with the general verdict "This is basically sour grapes cause the art for this game is problematic (and, viewed as pixel art, mediocre)."

Having some experience in pixel art, CG, and 3d, I agree with your assessment that 3d makes it easier to 'slot things in'. However, this is not a valid justification for making Crawl 3d ; that would dramatically change how Crawl played, even if there was no extra geometry/dimensions of actual gameplay (ie. like Vulture's Eye or NotEye: a reskinning/viewing of the underlying game only).

We would have people cheating like crazy. Not a lot, but some.

Yeah, Crawl's design doesn't help here. Designs like ToME4 where the actual game is implemented in platform-independent Lua code allow identifying exactly what code you are using via a hash, reliably, which is quickly and easily verifiable even if you have millions of valid hashes.
Obviously this doesn't deter the really determined hackers; nothing does. But it's a start.

[Sar]

Monster tiles are the worst offender by far, yes.

[Pollen_Golem]

There's another big problem with the article. In the past five years or so, there has been a huge resurgence of the pixel-art aesthetic, to the point that many games are gratuitously pixelated, to their own detriment, where there's no good reason to pursue this style. I don't even know how the author managed to avoid mentioning that.

[bcadren]

There's another big problem with the article. In the past five years or so, there has been a huge resurgence of the pixel-art aesthetic, to the point that many games are gratuitously pixelated, to their own detriment, where there's no good reason to pursue this style. I don't even know how the author managed to avoid mentioning that.

Using pixel-art well actually is a good choice in a lot of circumstances; choosing it because you think it'll be significantly less work is a bad idea though; good art is WORK regardless of the artstyle. That said; if you really don't care about art quality and are a programmer wanting to give your game SOME art it is less work to make crappy pixel art than crappy 3D.

[Blobbo]

Crawl is a true roguelike, it's base form is the terminal.

If you stop thinking of tiles as pixel art, and more of "how do I encode as much visual info into 32x32 pixels as I can do with a single colored symbol that is referencing a known concept" then the challenges are clear. Seeing a capital D come down the corridor at your @ is something every roguelike player understands and has an instinctive reaction to, whether or not they've played Crawl specifically.

[ontoclasm]

I am perfectly aware that our tiles don't all match; I tried to count at some point and Crawl contains on the order of 3000 different tiles. And every bit of the work done on them was by a handful of amateur volunteers, all of whom have more important things going on. Given that creating a single tile can take multiple hours, it's a serious project.

Personally, I'm currently engaged in making the tiles for Caves of Qud, so I haven't had time for Crawl the past few months. Once I'm done with that, I'm going to redo all the merfolk and naga tiles, and at some point I'd like to redo the player doll tiles (which is a huge can of worms, since changing the base tiles even a single pixel will invalidate the hundreds of armor bits that get layered on top of them).

However, if you find a tile you don't like, or which you have trouble reading, or which doesn't match the style, then the solution is to let me or roctavian know, not to post "well Crawl looks like shit haha" on tavern. Feel free to minmay up a list of like 300 of them and stick it in here; we won't be offended.

Just as a reminder, Crawl looked like this only a few years ago:

[nago]

Currently I think the biggest offender are player ghost who often are almost invisible.
There are others tiles which are probably ugly but they certanly don't affect negatively the gameplay.

[Sar]

I do appreciate the work put in Crawl, so sorry if that offended you!

The screen you posted doesn't look terrible to me, though, so maybe I'm not the best to ask about what tiles are bad!

[ontoclasm]

I'm not offended, it's just that "they're inconsistent" doesn't really help me improve anything. Naga don't look like vault guards; which should change to match the other?

[Sar]

That's a really tough question considering just how many of these tiles are there! Honestly I don't even expect Crawl looking consistent unless somebody hires an artist to work on it for some time, or some artist has crazy amount of dedication and time. If I had to pick some tiles that stand out, I would point at the monster ones Bloax made. Not to say they're awful or that Bloax is a bad spriter (though I used to joke about it on Tavern), but they are often pretty different. I don't know the technical details, I think it's the amount of colours he uses. I think they'd look better if the resolution wasn't 32x32, which it is.

Edit: also I think vault professionals look pretty good.

[bcadren]

I feel like the creatures with lots of contrast and obvious detail outlines, over more realistic internal shading and muddier detail areas look much stronger. [Outlines really pop; trying to actually look like a bear with complicated shading just looks messy].

Anyways; I felt compelled to try some half-assed depth shading and it's amazing how much of a difference such a small thing made.

Before:


After:


Difference:


For the record, I don't like that particular tileset; the attempt to look like actual rock leaves it a little muddy rather than more striking and there are a lot of details on it that are difficult to discern (is that supposed to be ivy growing on the wall or...?) it makes no gameplay difference, so it doesn't matter, but it's still odd.