Skip to main content

Indie game storeFree gamesFun gamesHorror games
Game developmentAssetsComics
SalesBundles
Jobs
TagsGame Engines

Alloyed

22
Posts
1
Topics
24
Followers
59
Following
A member registered May 23, 2015 · View creator page →

Creator of

Recent community posts

hey, I'm making <thing> and for <thing> I want to have some kind of user account. I also don't want to do my own user accounts, so the itch serverside api seems like a perfect fit! So this is just a bunch of notes on my experience getting everything integrated, from the client to the server side.


I started here, which is pretty clear enough:
https://itch.io/docs/api/serverside

To know where my jwt token (yeah I know it's json web token token you can't stop me) will come from, it links me to https://itch.io/docs/itch/integrating/api/. This is a little odd, because the information I actually want is on https://itch.io/docs/itch/integrating/manifest-actions.html. Ideally I'd be able to post an anchor link to the "API Key and scoping" section, but I can't tell if the headers actually have anchor links? either way, this is the most pertinent info and I can work backwards to find out what a manifest even is.

When I get there, all the documentation there uses the term API key. In practice, the thing that's in ITCHIO_API_KEY is a jwt token, which is what I would expect, and not a persistent API key.

Then, next step in implementation is to make and test a client that can do the right thing with an ITCHIO_API_KEY. The only way I could find to do this was to actually create an itch project for my app, upload it with the manifest, and then run it. That makes sense, but it adds a lot of time to the iteration cycle while developing. If there were a way to use butler to treat a random folder as an itch app and run it, or even better a way to have butler just return a jwt token, it would be much simpler for me to test my auth code, and maybe even do some automated testing. 

So in addition to authenticating users, I'd like to make sure that a user should have access to my game before letting them in. The way to do that then, I guess, is to use the download_key endpoint. So I go ahead and test it, and to my surprise the account that created the itch project to begin with doesn't have a download key, which is true, I guess? I just went ahead and generated a key and claimed it on the same account, and then my code worked fine. Is there/should there be a more inclusive endpoint that can include anybody who can download the game on the app, including maybe press accounts? For testing, I can just turn this piece off, that's not a big deal, just wanted to check if this approach is valid.

And then some not-discussed things that I figure I should ask. I 100% don't expect a problem with these at my scale, just that it puts my mind at ease knowing that an answer exists.

1. Rate limits? Right now, I just fire off a request, and if there's a failure of any kind that's bubbled up to the player. If I knew there was a rate limit, I could enforce it on my side and save some traffic, or even better implement a queuing/retry system.

2. Revoked keys? the API docs mention the possibility of a revoked download key. This makes me sad, because I wanted to cache off the download keys: if a player has authenticated successfully once I assumed I could just treat them as having owned the game and save an http request. What kind of situations would this happen under and could I get away with just querying to get a full list of revoked keys every once and a while?

(1 edit)

I tried to make a video but I'm in game crash town

This happens when I try to start the game from the main menu, no matter what, using windows 10

https://hastebin.com/etecivuyi...

I'll still try to post me playing through the tutorial when I'm done with this batch of videos <3

https://youtu.be/JQ8fIBlIL3Q
https://youtu.be/ggTYk1VD6EM

^

https://youtu.be/16sOCMWbuy0

https://youtu.be/16sOCMWbuy0

https://youtu.be/Fdx4nSo7YMc

I don't like the RE games so I'm kind of predisposed to dislike your game too, sorry. I would definitely take my opinion with a grain of salt. I think I did find a few technical things in there that you can work on tho

https://youtu.be/3uB7rUdvBoA

recorded impressions ^

I fall into the bucket of people who didn't realize that there was a gem mechanic :V

ty

<3

the framerate thing is a stupid mistake on my part, I think: if you have the spare time can you try this build and see if it has the same problem?

windows

mac/linux

noted on the other points <3

On the tech stuff, understood, I think I know everything I need except for #3. What do you mean by "arrows from below"? as in you fire directly upwards somewhere around 90 degrees?

Re longevity: I was kind of hoping that people wouldn't notice that the map is static when it's mixed in with a bunch of other minigames/mutators. I'll probably do it anyways so I can trick more people to playtest the game with me

Don't bully the birds :c

(they actually already have a hitbox, I still haven't decided what should happen if you shoot at them tho)

Thanks, got bug A, I couldn't reproduce B but I think I could just rewrite that bit of code to avoid that class of problem

My feedback is apparently too long for itch so here's a link to it instead

http://hastebin.com/ucezequxax.md

tl:dr: I had fun but ended up with a pretty huge list of nitpicks along the way

(1 edit)

Attempt one: http://b.1339.cf/flovilc.mp4

Attempt two: http://b.1339.cf/xafsihf.mp4

I did attempt one cold, no instructions, I kind of vaguely looked at them for attempt 2

This is literally the first slenderman-type game I've ever played, so I'm probably more than a little off base here, but:

Observations

* I didn't know about the vault mechanic for the first two attempts, but realized it by the time I tried it off-camera. It seems to work pretty well, but hiding its existence in optional instructions is not great.

* spent a lot of time lost, and not in the way where I felt threatened or really worried: since I knew I was playing a game, was very specifically looking for a sound-related trigger to progress, and that the thing chasing me had to actually chase and touch me instead of coyly looking from behind some twigs, it took a lot of the early tension out. ingame exposition would have helped.

both sound like they would be fixed with notes laying around. it's kind of silly but ehh, if it works elsewhere

* every time I've seen the light from far away it had some visibility problem, where you could only see it if you looked at it from your peripheral vision. not a 3D guy so I can't really explain it better but I do it in attempt 2

* resetting from the menu seems to softlock at the loading screen. Happens on attempt 1. reproducible.

recording: http://b.1339.cf/pvskoyy.mp4

observations:

* I didn't get to see much of the game thanks to skelebro, maybe make his bone arc more vertical so even when it's on path to get you you can strafe out of the way?

* making jumps a limited resource seems like an odd choice IMO, since you can only progress by jumping: it makes it possible to make a level unfinishable for actions you took well in the past. Either give the player all the info they need to beat the level up front by making the game last a single screen, or do something like add a double jump and make that limited so the game only becomes harder, not impossible, if you mismanage it

recording: http://b.1339.cf/rqkgzkv.mp4

observations:

* maybe I'm bad, but it seemed like some situations were unwinnable, and they were unwinnable due to random movement as opposed to my direct actions. Frustrating

observations cut short because I just realized that I could have been shooting bones the whole time

...whoops

You can have the recording anyway I guess


(1 edit)

recording: http://b.1339.cf/qgbhliu.flv

observations:

* too easy for too long, by the time I started to need to pay attention I was already out of it

* the blocks _almost_ fell in time with the music, which was a little disconcerting. sort of an explicit sync would have helped

* it seems like an ideal strategy would be to just circle the arrow keys really fast. Adding mechanics that didn't make that true, like a "blocked just in time" combo, would have made it more compelling to play