@EricTressler It's happened twice so far, so I'm not sure if I could reproduce it. If it happens again, I'll try to make a video

shouldn't a screenshot show it?

@EricTressler I suppose, but I'd like you too see how it gets fixed after un flagging and re flagging old mines

@DJMcMayhem ok. I guess I can also try to replicate it; do you remember a puzzle # where you saw it?

@EricTressler Yeah, I saw it on level... Uhmmmm.... 19? I don't remember the number of the top of my head

The hexagon shaped one with 3 colors right before the comb one

It happened near the upper left corner. I have no idea if that's helpful at all

ok, thanks. i'll play through it a few times. do you have any other options turned on (darkening tiles, nonstandard bindings)?

It also happened on an earlier level, but I don't remember which

@EricTressler I had the first three options on, but no the first

@EricTressler Just checked. It was 13 and 19. Also, I doubt it'll matter, but I used the flag style on the left

I can't reproduce it. I checked to make sure the redraw is triggered after the state change, and it is. That's frustrating. Please just send me a picture if you see it again.

Will do

Looks like we have a Heisenbug!

I used to redraw every tile every frame. That's safe and stupid; then I switched to only redraw when a tile changes state. But counting down doesn't count as changing state, so every time you flag or reveal something, I set all of the neighboring tiles (including the neighbors of extra tiles that can be revealed by 1 click) to an explicit redraw state. That happens after the click, so I can't really imagine what could be going wrong.

@Mego it took me a long time to even mention that I had made a game, because self-promotion embarrasses me. probably never.

The only interesting algorithm is a little too involved for PPCG, I think

Thanks. I think next time I will just make it free and open source, and then it will be easier. This isn't on Steam for the money, but mostly just to see if I could go through the whole "real" process. I'll probably let the price drift down toward the minimum over the next year or two.

@Mego It's a bit of a mess. I wonder if you'd still feel that way if you saw it. On the other hand, a lot of the code is just there for level development -- none of that needs to be included for a working game.

That's an interesting idea. I might end up open sourcing it. Have you made any games?

@Mego Heck, it's enthralled a good half dozen of my IRL friends and family. After I just got it a say ago

@EricTressler I played around with game Dev in Python back when I first started learning to code

I looked at pyGame a year or two ago, and I wasn't impressed by any of their showcase games

@Mego what sorts of games have you tried? I picked this because I like puzzles, but also because I knew graphics and stable frame rates and things would be less critical.

Just to chime in from a quasi-legal-definitely-not-a-lawyer-standpoint: most places open source their games by releasing the code and letting it be distributed in source or binary but only for free yada yada, and then keeping their copyright on the assets

@EricTressler Yes, pygame is pretty simplistic. But it is good for learning to code with a topic that's interesting without being too challenging

@quartata I have no assets except for 3 sound effects, so that would not be protecting much

@Mego I beat all of the Zorks in college. There's a SCUMMVM(?) engine you can use, I think

In that case maybe what you could do is polish the level generation tools into something other people could use (perjaps combined with a way of making new tilings) and do a Steam Workshop project

I've mentioned this before at least a couple of times, but I put The Talos Principle right up there near Portal too.

hmm I suppose the only trouble is when community maps make it into TF2 Valve gives them money...

from sales of map stamps

@quartata what you just described is more work than the game has been so far, I think.

@Mego Oh yeah definitely no "bundle community levels into DLC"

Map stamps are optional, they're a way of basically tipping the map creators

Portal 2 is probably the only game that I've played user-created content for (well, and Skyrim mods)

valve encourages said behavior via hats, rather Pavlovian

Mind you these are for levels that are accepted into the actual game and come with it, but were made by the community

I think the last arena shooter I played was Quake 3

@Mego was it just a browser game somewhere, or was it possibly on Kongregate/etc.?

Because I know of at least one game similar to that on Kong, but it's not great

@EricTressler Agreed, Talos is fantastic. Have you played antichamber? Or the witness?

@DJMcMayhem Yes. I didn't like antichamber that much (never quite finished it, but got the third gun); I liked the Witness pretty well, but I thought the puzzles were kind of annoying at times

@Mego Same

I think the concept is great, and I would love to see something like the Witness but 10x bigger and more puzzle variety

Especially because of the nostalgia. First game I ever purchased on steam

@EricTressler The witness is already pretty massive. Isn't there like 2500 puzzles?

Another puzzle game I really like is Desktop Dungeons, but it's involved enough that when I put it down, I have to start totally over when I play it again

@DJMcMayhem I.. don't know. I would guess there are maybe 200-300, but half of them are trivial? It's not really the # puzzles, something about them was just a little unsatisfying to me.

I don't have a specific objective complaint to lodge against The Witness

I think I'm imagining something like a cross between the Witness and the Primer from The Diamond Age

@Mego strong claim :P

@EricTressler I played two or three hours and then got kinda bored because there seemed to be a lack of a red thread in terms of the mechanics. It all just seemed like a collection of random ideas. Does that ever get better?

Fair enough. It's a great game but it's never among the first few I think of when thinking about amazing puzzle games.

Yeah, Antichamber was pretty good

Yeah, I just usually prefer 2D mechanics.

Fez has a similar overall feel to it as Antichamber, and I liked it a lot more.

Yeah the story of its development is quite unfortunate.

@Mego funny enough, I was just thinking of an interesting one:

guess the placements of the flags given a somewhat-solved puzzle

I'd probably abstract it a bit so that people don't have to worry about parsing the topology. (You're basically just give a list of index sets and a number of mines for each set, and need to output which indices have mines.)

I'd just use the same topology every time

likely Moore

Right, but people would still have to parse the input and represent the topology.

yeah, but for PPCG I feel like they'd be more comfortable with that than an arbitrary graph

I should have been clearer too: when I say "guess", I literally mean that

they'd be given a point for every flag they correctly guess

I'm talking about a probabilistic solver

that only gets a single state of the board

Oh right

@MartinEnder Are you referring to Desktop Dungeons (for incoherent mechanics)? I don't know if it gets better, but it begins to behave a lot more like NetHack. Many things are available all at once, and you have to decide which race/class/god will help you beat the level.

@EricTressler this is sort of the thing you don't have to worry about with JavaFX

if you do end up making it open source, I'd also love to take a crack at it :)

@NathanMerrill That's something I could probably try, if I can draw in a JPanel. The drawing code is only about 1300 lines, 300 of which are dead. What "sort of thing" are you talking about, though?

diff checks

JavaFX is a totally new framework for graphics. You either pick Swing or JavaFX, they don't mix

(which was intentional)

Ah. It's probably a lot harder to make the replacement, then

yeah.

many of the concepts are the same, and even many of the component names and functionality is the same

for example, JavaFX's "Label" is identical to Swing's "Label" for all intents and purposes

I had(?) to do some really dodgy things to get the launcher to work properly. I'm not sure if I circumvented intended behavior, or if you're really supposed to have to modify things to the degree I did.

For example, I had to override getTableCellRendererComponent() to make "Press any key or mouse button" italicized in the key binding table.

@EricTressler in JavaFX you can either do a CSS stylesheet, or call setCellRenderer()

but for swing, that sounds about right

How did you get so familiar with it? Do you use it for work, or classes, or just personal projects?

work

given, I'm doing mostly standard gui stuff, no drawing

lots of dropdowns, tables, and other various things

What do you work on? (If it's not private)

with JavaFX? A point of sale system. The company does most of their sales through an online website, and my system allows them to do purchases in person by connecting to the store

its not terribly unique, but my work is a PHP shop, so the fact that it's in Java makes it my favorite project to work on :)

Cool. I have no experience writing production code

production code is interesting because when projects start, its often very intentional and well designed, but client needs after the initial launch make the product messy and poorly designed. Personal projects are often times different: they start off as a one-off, but then the people care to take the time to refine them as they grow

so, production code starts off great, and goes downhill, while personal code starts off terrible and goes uphill

I'm currently looking at github; I'm considering upgrading my account so I can make a private repo and share this code (with no license for the moment).

are you in school?

No, not for quite a while. I do data science work for the IRS; previously, defense research.

ah ok. you don't write code for your data science work?

I do, but it's to do clustering, or run a PageRank algorithm on a graph, etc.; the data that comes out is important, but the code is mostly disposable.

I have a core graph library for basic manipulations and visualization that I wrote in Java, but I'm the only one who uses it. That's the closest thing I have to production-quality code, but it's still not really.

@NathanMerrill Do you have any experience with Eclipse + git?

nope. I use IntelliJ for everything

I was trying to share the code, and it made some alarming changes to the project directory. It seems to still be okay, though still not shared.

I am familiar with git. You can just go to the project directory, and do git init followed by git remote add origin <url>

oh, the gitignore is clever

I'm going to push a copy of the the project folder for right now, with no license (so I'll let you look at it if you're curious, but I don't want to commit to open sourcing it)

Okay, I put a clone of the working code up there. I don't intend to open source it any time too soon

@Mego @NathanMerrill if you're interested, I can add you to the repo. If not, this will be a useful backup until I'm comfortable actually using it for version control.

I am interested. I'm nathanmerrill on github

You're welcome to make general suggestions, but there are already a lot of obvious things that should be changed. Primarily, puzzles need to be moved out of the source code into XML or similar files. But anyway, poke around if you want.

Since the code is working right now, and people are using it, I'm hesitant to fix things like that on the grounds that it might backfire.

yeah. Just as information: source code is usually in the directory /src/main/java instead of directly in /src

its mostly for organization purposes. You'd put stuff like your icons in /src/main/resources

and tools like gradle and maven automatically assume that that is where your code is

@Mego I don't want either of you to contribute; this is not a live project, it's just a clone. I'm putting this up there because: 1) maybe it's interesting/useful to you, 2) I should familiarize myself with git, 3) if I make a sequel it will probably be open source

This is reminding me how many versions of the solver I wrote, too. I kept them all around, which was a questionable decision. A lot of this became a black box for me early on.

@NathanMerrill maybe I should keep it private until I've done some more refactoring and I'm ready to open source it?

@Mego I didn't know you were that far along

@EricTressler What you do with it is totally up to you. I do like reviewing code, so I'm happy to do that

What's 45 again?

@NathanMerrill I'll keep it private for now, sorry for waffling. Thanks for taking a look, though.

45 is TMI

oh man, I'm not even at 45 yet

Stuck on 42 still

I feel like I made a mistake, as I have no idea where to progress from here

Want to share a screenshot? I'm going to be going for a while pretty soon

what's the total count

13

so... that 2 sitting there with all of the unrevealed tiles around it

ah, I figured it out

lol yeah

ok, cool

its the 3 row, right?

right

I do like the combination locks

@EricTressler no, I meant Talos

@MartinEnder I don't know; I didn't feel like there were too many mechanics, it pretty quickly gets down to using old ones in new ways (like all good puzzle games)

I think the original game unlocks something new for the first few chapters, though, so there are new things still trickling in for a while (if I remember correctly)

I'm going to head out for the afternoon, but I'll be back later.

@EricTressler yeah, I don't mean that there were too many mechanics, just that mechanics that it had seemed like a fairly random assortment of some puzzle ideas.

I didn't really see any unifying theme to the puzzles (other than "spatial 3D" I guess)

@MartinEnder Remind me, how far in Talos did you get?

I'll have to look it up when I'm back home in a week or so

or wait, I might have mentioned it in a chat somewhere

hm 4 or 5 chapters in or so? I think I did maybe half of the first hub? I believe the one of the last puzzles I saw was two of those floating mines moving in opposite directions along a narrow corridor, in a level set with dark/bad weather?

My main thought on Talos is that the puzzles are certainly very high quality, but they're not really unique. Portal, Fez, Antichamber, The Witness, etc. Are all a lot more interesting in terms of unique mechanics. Talos still does a really good job of making interesting puzzles with ordinary mechanics though. And the puzzle quality gets way better when you get to the tower because there comes a point where you are working both with and against two other AIs, and that was super interesting.

Some later mechanics are really interesting (fans, using the vertical dimension to solve puzzles, and time travel ala braid). But overall, the real reason I enjoyed it so much was because I found the story really intriguing and it drove me to beat more of it to figure out what the hell was going on. I really liked the philosophical themes and religious metaphors. The ending was very satisfying

I also somewhat dislike puzzle games that mix the puzzle aspect with real-time stuff (like navigating skill) especially in 3D. most of games like that have a hard time communicating whether you just didn't execute your intended solution well enough or whether the solution can't work.

I guess that's what makes Portal so amazing (that this isn't as much a problem in Portal as in other such games)

However, if you're not enjoying the game mechanics, the story isn't necessarily a great reason to push through it

@MartinEnder Hmm. Not entirely sure I understand. Could you give an example?

But yeah Talos certainly relies pretty heavily on spatial awareness, navigating and 3D puzzles

Well for example the moving mines. For some of the levels, you need to move past them at just the right time but without getting too close. If you have a solution to the puzzle where the time window is very narrow, you might have the correct solution but you might fail to execute it successfully, which is quite frustrating.

That comes up again when using time travel. If you don't do things quickly enough, or accidentally end up in the way of a laser you'll ruin it even if your idea was correct

yeah, that kind of stuff

On a completely unrelated note, I recently found an awesome mobile puzzle game. It's also on steam. It's called Snakebird. It's like if snake had gravity, and was thrown into the world of Fez. The mechanics are dirt simple, but it gets surprisingly challenging later on

The first 10-12 levels are free if you want to try it out without paying

oh, I think I've seen this before. thanks for the recommendation

Always happy to recommend games I enjoy :)

One of the best minimalist puzzle games I've played on PC was Sokobond.

Woahhhh... It's sokoban with chemistry :O

@EricTressler I feel like I reached an ambiguous situation on 85 (endless knot), I'm probably overlooking something though. I think either the squares or the circles are a valid way to put in the 3 remaining green mines: i.sstatic.net/vImU7.png ... the only way I can think of resolving the ambiguity is that I have a green false positive somewhere, but I can't find one.

@MartinEnder I'll take a look. I can't promise to find anything quickly on that one, though

it's almost entirely solved

I see that. Give me a moment to stare at it

You can partition the cells

let me draw a different picture (ignore your red)

oh

wait

yeah I know what you're gonna say

I had a third solution with different placements, but one of the green ones is never covered, so it's safe

ok. Yeah, you just find 3 nonoverlapping regions that each have to contain a mine

not sure why I didn't think of that, I use this all the time

I assumed you'd know about that (otherwise how did you get to the end of the game? :)

yeah, probably been staring at this level for too long...

Yeah... I take it you think it's a bit excessive?

well... I think it's probably one of the toughest overall. but on this second attempt I actually got through just fine.

I just think it should probably be a bit later. in terms of difficulty it seems like it could go in the last 5 levels

I had a much easier time with it after practising some more on the levels after it

only Solomon's Knot left now

I probably shouldn't move levels now, it could cause weirdness with people's games. I don't think it would technically break anything, but it would be confusing for players.

don't worry about it, it's not that important since levels can be skipped

for most of the middle of the game, I tried to put minor difficulty spikes on the multiples of 5. That falls apart in the last row

well you succeeded at 85 :P

got it :)

I think that took 28-30 hours in total

definitely keep me posted if you do any content updates or write a sequel :)

were there any mechanics you considered adding but scrapped?

@MartinEnder congrats! Yes, a couple; I liked the {#} hint type from Hexcells, and I emailed the creator a while ago to see whether he would mind if I used it.

He didn't answer, so I didn't use it

what did that represent again?

It means that the mines are adjacent. So for example a {3} in a Moore neighborhood tile means that there are 3 mines in neighboring tiles, and they are consecutive (if you think of the 8 neighbors as being in a ring)

ah right

yeah those are interesting

I also considered putting in long-range tiles that tell you how many mines are within distance 2, but I decided that's too confusing combined with the different tilesets

they add a bit of a nonogram component

ah yeah I was gonna suggest distance 2, but I can see it getting really messy for hex/tri or square/oct grids (I guess you could only use it on hex and Moore levels or something)

The {#} hints would have been a different type of hint, though; as it is, the columns and colored hints are equivalent to having an invisible tile and knowing how many neighboring mines it has.

So taken as the appropriate graph, there is really only 1 type of rule in the game.

Well, and "?"s

I'll definitely ping you if I add any new levels or anything. I'm glad you enjoyed it

How did you go about playtesting?

I mostly played it myself, and I was able to get a couple of friends to try it out too as I went. I'm not sure what else to add to that. I think I probably made 130 levels and cut the worst ones as I went

Ah okay. Just curious because I'll probably have to worry about that in a few months (not sure whether I mentioned it, but I'm actually planning to work on a puzzle game myself over the next few months).

Oh, and I also had to play most levels a few times to iron out minor issues, like getting the starting conditions to be appealing

It was really hard to get people interested in it enough to play past the tutorial (even though it was in pretty good shape early on)

I guess it was difficult and niche enough that a random sampling of my friends didn't work very well.

yeah, I imagine I'll have similar issues. I don't know anyone in person who'd be interested in the kind of niche puzzle game I'm going for (okay, like maybe one person).

If you want testers, I'd be willing to put significant time into it. Let me know

That would be great. :) It's gonna be a very different brand of puzzle game though (programming puzzle game... think Zachtronics or Manufactoria if that means anything to you).

Yeah, sure. I like infinifactory, and SP3000 has me convinced to try Shenzen I/O again (it's intimidating)

I got through almost all of human resource factory, and I also have... what's the other zachtronics game...

haha, Shenzhen was great (although I didn't actually finish it, as opposed to TIS-100, but I suppose that's even more intimidating)

SpaceChem? Or the new one, Opus Magnum?

TIS-100, and I think spacechem too

I like them all in theory, and they go on sale enough that I basically pick them all up, but it takes a bit of a push to get me into the games

I'm going to go cook now, but I'm definitely interested in testing. My primary email is my last name at gmail, if you don't see me on here at an opportune time.

I have this problem with SpaceChem... I really want to get into it, but every time I try I lose interest after the first couple of chapters. But I completed TIS-100, got to final level set in Infinifactory, the bonus campaign in Shenzhen and completed Opus Magnum, so I can definitely recommend all of those.

(Also completed HRM, I think even with all optimisations)

@EricTressler it's probably gonna be a couple of months before I've got something I want to put in front of testers, but I'll definitely take you up on that when I get there :)

@EricTressler feature-request from my girlfriend: When the level finishes and the colors invert, it's kinda harsh on the eyes. Could that use less bright colors? Or a smoother transition or something?

Btw, she's been obsessively playing it the last couple days, so take that as a compliment. :)

@DJMcMayhem that's such a minor thing, but adding any option takes time (mostly in incorporating it into the launcher and then the config file), and then pushing the update to Steam. I'll consider it -- I also have to decide what to do about "?"s going dark when you have "darken finished tiles" turned on

@DJMcMayhem I might make that change, but it would be something like Monday evening, and no promises

for the record, I think adding this is a good idea. after a few levels I just ignored it, but for the first few, I found the change upon completing the level quite jarring.

Alright. I'll find a different, darker color for it. But still probably Monday

(as an option, though)

@EricTressler Thanks for the quick response! No pressure to work on it soon. Enjoy your Christmas! Personally, I don't think it would be worth adding as an option because I don't know how that would come to do preference, but if you'd rather have an option for that, it's your choice.