@AlexandreVaillancourt So says, "don't do it" - stackoverflow.com/questions/11658915/…

Do not take Cortex Command's position in my Steam Library as an endorsement. It's horrid.

Oh, if it's not obvious, I posted that because it shows achievements working in my game Cognizer.

Steam release imminent

Oh well, all good things come to an end.

I feel out of touch with the community lately.

SO community mostly, not gamedev as I'm not active on gamedev besides chat.

It really bothers me that people write terrible questions and terrible answers and both get highly upvoted.

It makes me question if SO can live up to the standard I've come to expect.

So I'm making a unity editor tool that reads the public fields of all monobehaviour scripts in the scenes that are in the build settings and lets you set default values for the fileds to be used in different builds, thus illiminating the possibility of building with lets say debug values or something. You guys think something like this will sell in the asset store ?

Are you dreaming of 1-time effort infinite money schemes?

maybe

@DukeZhou yep. I got 7 working NPCs. Also the North Half of the Island is snowy. Aslo added where now you can collect loot from chest. Aslo fixed bug where enemies weren't saving nor loading properly, causing the world to lack enemies when reloading a save.

Also changed how damage is calculated to be more efficient.

That Aslo guy is really hard-working.

Yep. "Also" is hard at work making sure it is at the beginning of my sentences :]

also is a slacker

LIES!

But in all seriousness, how has everybody else's projects been goin?

I had a couple of days where I could concentrate and made a lot of progress.

Nice.

Still some bikeshedding to do, but I'm just about ready to start working on something that may be useful.

Sounds cool!

Nice and scary at the same time to have actual users.

I am actually excited to let others try my game. Just me reviewing the game is biased :P

Sadly, I still have alot of work and refining before that...

@Tyyppi_77 I remember issues with that. Keyboards only let you press so many keys at once before they don't transmit keystrokes anymore, and usually that limit is at 4-ish, which sucks for 2 players. Although I also remember some keys being special, like shift and ctrl and sometimes the arrow keys.

@Tyyppi_77 Are you still working on your metal slug-esque game?

You could also be the cool friend who just gives them a controller or 2 :P

Do you have any kind of automated testing for your game?

ah ok

I'm interested in how stuff like that would be achieved in games

Or whether most game testing is manual

Did you take alot of inspiration from this site?

So you record the user input for 1 succesful runthrough and ensure all psuedorandom will run the same way

That run without the timestep and rendering turned off I guess

That sounds super brittle.

It does

But at the same time

computers are very good at remaining deterministicly

You are testing if you can play through the game with the given input and the given RNG which is a meaningless test because you don't actually care about that.

I think we do though?

For example, the test could fail if the player doesn't match where it should be at a given time, meaning you fucked up part of one of the maps or something like that

Why? If a level gains an extra enemy and suddenly the pre-recorded input fails, what did you learn?

you messed up

and spawned too many enemies

That's literally the point of testing, to ensure that stuff doesn't happen

So the level design must never change? That doesn't seem right.

It can

But you must redo the test for that level

If you change it

^

Oh I guessed that part

Good guess

Same with my game

I wonder if we could use AI to test levels

That would be cool.

It seems to me the input-play-through test is equivalent to the file comparison and misses what should be tested.

The way I see it

@Tyyppi_77 Sure, that would be great, but the test doesn't do that. If you alter jump distance slightly the test will fail no matter if it is still completable.

The only alternative is to play the game yourself

but if your predefined bot does it fine, why bother?

>hmm

It doesn't do it for anything else either. Changing damage, drop-rate, ammo size, anything and the input will most likely fail.

But what if you don't?

I guess it is somewhat helpful to know that the graphics update didn't screw with the game physics, but there probably is a better way.

lmao I am

:(

Another advantage to the prerecorded input for testing a game is that the input can be played at max speed, without rendering turned on. So you could potentially do an entire runthrough of your game in seconds, which could could take 30 minutes for you to do manually

If it fails on a level, you could re record that level (say you forgot you added a box somewhere) which would be add 10 minutes, but that's still less than 30 (or 2 hours)

And then you have the confidence your game works..

@ShaunWild Yeah, no, that test doesn't instill much confidence at all.

If you do it with a by-level test with minimal starting gear that may have some merit.

Although the correlation between "playable" and "there exists an input for which the level completes" is rather weak.

@Tyyppi_77 my favorite rendition is "The first 90 percent of the code accounts for the first 90 percent of the development time. The remaining 10 percent of the code accounts for the other 90 percent of the development time."

deep

Maybe it makes sense to try to measure the difficulty by delaying the input by X frames and seeing if it still completes.

Or one could fuzz-test the game which might produce cool results. Although it probably eats too many resources to be practical.

For a 2d game?

I doubt it would eat many resources?

@Pikalek Yeah, it appears like a bad idea.

Now the next question is what if I build the debug DLL with some parameters and the debug app with some other parameters?

Maybe adding /O2 for the DLL could work.

@Tyyppi_77 I think what Josh said is the way to go. There is nothing that should be done with answers to off-topic questions. Yes, they may be doing it on purpose, and even if we tell them to stop, they can keep on doing it. And if they did it on good faith, well they'll probably realize eventually that the question is closed and that's not the type of question we seek here, and they might eventually recognize that.

@nwp SO has become a rep gaining machine. You're right, why one-line-link answers get upvoted?

@Tyyppi_77 @nwp someone made a tester bot for their game that does random actions until the game crashes

quite cool: blog.thimbleweedpark.com/yay_testertron3000

In my case it wasn't a link-only answer, it was a poorly researched "How do I do X in Y" question with a poorly researched (technically wrong) answer. I still don't understand how it got uncharacteristically many upvotes.

@nwp Yeah, I'll try that!

Some fuzzer libs will then try to reduce the input while preserving the crash. It's pretty awesome.

@Tyyppi_77 Not always :)

I'm trying to think of how well a melee only top down rouge like would work. I just don't know how well it'll work.

Anyone know of any top down melee only rogue like games that play really well?

There should be plenty of rogue-likes with melee classes.

It should be close enough to melee-only to get a feeling for how well it'll work.

@AlexandreVaillancourt that's why it's breaking news

I'd say the same for the close votes review queue.

@nwp lol

@TheMattbat999 seriously though, you should think about naming an NCP "Aslo". It's a terrific name, imo

sort of like "Lazlo" without the the L, and kids will certainly have fun with it.

@DukeZhou ok. I also gotta add a leveling system. Right now, all enemies spawn at first based on the player being lv1. If an enemy dies, the spawner waits an ingame day to spawn an enemy on their level. The problem is new areas end up ALWSYS having lv1 enemies. My idea tho is to have the spawner reset as long as the player is far away enough.

So that new areas include more difficult (and leveled) enemies.

Good call. It' can be fun ploughing through low-level NPCs, Dynasty Warriors style, but those games are 99% adrenaline.

You might think about a logarithmic "mastery scale", where a level 1 master can beat 10 sub-1 NPCs. level 2 master can beat 100 NPCs. (possibly that scale doesn't fit in with the scope, but I like thinking about it.)

In a lot of kung fu films, it's traditional that the hero has to plough through all the henchmen to prove themselves worthy of fighting the boss.

Sounds like you're on it, regardless!

You might also consider a workaround if the glitch does arise--the player character can call out for worthy challengers. This could spawn appropriately leveled NPCs.

Ok.

@DukaZhou another question: which would you prefer: being able to equip more different armor pieces or but just a few? E.G. Helm, chestplate and greaves v.s. Helm gloves, chest, greaves, boots.

@DukeZhou

@TheMattbat999 Having fewer could make it easier to get "full sets" while having more can make it more challenging (and you can vary a bit more the set bonuses if you have more pieces, i.e. "partial set bonuses"). It can also be easier to create more finely tweaked character builds if there are more pieces that can be worn.

I know when I played Mass Effect (1), there was so many options to customize my weapon that I got kind of annoyed by it (dude why do I have to manage all that shit, I don't care and I just want to kill bad guys), and I remember when I was playing Diablo III, I was surfing the web to find the most efficient builds when grinding...

Based on that, I guess you'll have to figure who's your target audience. I think more "hardcore" players will want more ways to tweak and tune their build, while more "softcore" (???) players will just want to play and go forward, so they'll enjoy simpler game mechanics.

Probably gonna go for the simpler route :)

Yeah :)

Nothing preventing you from starting with Helm, Chestplate and Greaves, then add more items later (i.e. EXPANSION PACK!)!

I am gonna have accessories tho. Necklace, 2 rings, and arm bands (?)

I always prefer more options, but Alexandre's suggestion is prudent on many levels!

The thing that really annoyed me about World of Warcraft was how you were locked into specific armor for specific buffs, so you couldn't really customize the look of a character and also be optimized. (They added gems for buffs, but, imo, just a stop-gap workaround.)

Stellaris 2.0 seems to be getting the balance right between casual players customizing their fleets for what "looks coolest in combat" and the hardcore players who are minimaxing every detail. Many debates rage on the forums regarding setups presumed to be optimal, but it's all situational, so many players use the uninspired default ship configurations which hedge the bets by having a little bit of everything.

Point is, the game is well playable and immersive, regardless of approach (casual or hardcore)

I am unsure how I can get accessories to actually work tho, cause I thought adding to a modifier (in which adds to damage, lets say) BUT I am unsure how to remove that modifier, especially if they got a ring and necklace.

In terms of the code?

Wouldn't it just be a T/F for the existence of an object, and if T apply the value modifications?

@DukeZhou Yeah. NOT sure of how to remove the modifier once the item is de-equiped

Ok, that makes sense!

Only process the aggregate value at execution of the attack or defense function

For this type of usage, I don't think you gain much by aggregating the values into a single new variable (array) ahead of time.

Ok.. I think I have an idea of what to do...

WOW did have a good idea in making different regions different difficulties. This allowed me to go back to the Barrens with higher-level characters dressed more like Conan the Barbarian (no armor;) I did it partly because it was cool, and partly for the Chuck Norris jokes, which, at that time, were traditional in the Barrens general chat.

but, then again, that game is an MMORPG, so, different context

Hmmm...

Not sure about that. I am gonna have multiple islands to explore, so that is viable. Well, I gotta go (about to lose service)

I'm just here to "throw spaghetti at the wall" (you decide what sticks;)

Thanks!

Can I throw poop on the walls?

@Sie I guess technically it's every human's traditional ancestral right to do so...

If one goes back far enough

On a more modern topic, I just ordered an Arduboy.

kinky

@StephaneHockenhull Can it play DOOM?

It has a USB port so I guess if I add a "co-processor" it could play doom... And by co-processor I mean an entire PC getting button inputs and streaming the framebuffer back...

I don't man. When my TI83 can give that little thing a run for it's money lol.

512KB of flash.. yeah.

I just don't really get what kind of market they are aiming for with this little thing.

People who want to make games on a dinky little thing

That's a pretty good description of Nintendo.

Ya you could develop for a Switch if you wanted. I mean it's basically a beefy-ish tablet.

But it's more fun when there are big restrictions :P

You could build a computer in Minecraft and make a game on that computer.

32KB of flash, 2.5KB of SRAM, 1KB of EEPROM. 128x64 1bpp screen, 6 buttons, 1 USB port, 1 Piezo speaker.

That's it :P

You could take two rocks and bang them together until you figure out how to make a computer with them.

Anyway it looks cool. Kinda reminds me of this...

Trying to find it.

Pico8 ?

It was called Banana Split or something like that. It was Pi based I think and it was a little mini console.

Damn it. This is going to aggravate me until I figure it out lol.

Pretty sure it was this: getchip.com/pages/pocketchip

yah, the pico8 thing

Ah.

The buttons are pretty terrible and it's not very limited natively. And Pico8 is some LUA-based fake console that never existed with limitations that don't make much sense compared to real hardware consoles.

Didn't sound as much fun.

Ya. I don't know who thought those buttons were a good idea.