I like to start projects…finishing them is another story.

My major projects I would like you to check out (open source): Chinese Language

‣ Learning App: https://greenants.github.io/HSK-3.0-Study-Game/

‣ An Abstract Board Game: https://github.com/GreenAnts/Amalgam_Webgame

I am always looking for contributors to help out.!

◉ Community Forum: https://unfinishedprojects.flarum.cloud/ ◉ Patreon: https://www.patreon.com/c/UnfinishedProjects

Thanks for reading my profile :D

  • 11 Posts
  • 29 Comments
Joined 10 days ago
cake
Cake day: February 4th, 2026

help-circle
  • I honestly wasn’t aware of this, and I looked into it and NodeBB will be a great option if/when we eventually migrate off of the free hosted forum (currently using freeFlarum for the initial community building, as I don’t want to spend money on this until i am sure a user base is able to be built up).

    Thanks for this info though - I think that this will be the best bet moving forward - as this way its pretty much the best of both worlds (a more personalized platform outside of lemmy/piefed, but also able to integrate into the Fediverse)



  • It’s not really that I have an “idea” more so that I am shifting my community (well shifting is a poor choice of words - as my small community had been abandoned for a while, and I am reattempting to start it again) to a forum. Discord’s structure was great when I used to play video games and needed a voice chat server to talk with friends or share screens, but for my current purpose of trying to gather a community around open source and creative projects - I think a forum is more fitting (for the majority of scenarios at least). The forum I’m currently trying to use also has an instant messaging chat built into it, for more instant communications.

    If the community grows to the point where people are collaborating in real time and need voice chats/video sharing, I think at that point they can DM each other and organize a 3rd party software to conduct those meeting.

    I just think that for my specific purpose, a forum is suitable - while for others it might not be.




  • I’m just going to copy and paste this here from my other cross post:

    lol, yeah - its not lost on me that Lemmy is a forum, if that is what you are getting at. But I personally equate lemmy to reddit, and while there are subs for niche topics, the idea of a standalone forum for my specific purpose seems like it has more of an opportunity to create the “small, close knit” type of community that doesn’t seem to fit within the Lemmy sphere.

    I could totally be wrong, and maybe its nostalgia, but something about a good old forum seems to bring something different to the table in my eyes.








  • (Weird - I sent a reply but now its not showing up.) I hadn’t heard of “clozemaster” but tried it out and it seems like a nice resource! The fact that it has the vocab in the context of sentences, while still only focusing on one word seems quite useful as to not learn the word in isolation.

    And great, yeah as both being beginners we can stumble through it together if you like - feel free to pull request whenever you like, or however else you want to work on it. As for being a native Spanish speaker - that helps out a ton :) - I did forget to mention in the main post that I was wanting to focus on Mexican Spanish rather than Spanish from Spain, but I don’t think should matter too much because the actual vocab should be super easy to just swap in or out with new JSON data.

    But yeah, if you decide to contribute (of course no pressure), I look forward to seeing some of what you come up with.


  • I think it depends on the language, but specifically for the Spanish version I was initially brainstorming the following idea(s):

    1. Separating parts of speech (verbs, nouns, adverbs, etc) into different “word decks” and for each deck you play as a different character (warrior, miner, etc.). This would create some variation to keep the player interested.
    2. The words you are trying to guess will be above the various things that character is trying to accomplish (enemies for warrior getting killed - ore that the miner gathers - etc.) 2a. For Spanish, Verbs have multiple conjugations - so there might for example be one Large enemy you kill for the infinitive version (eg: “ser”) - and there is a bunch of his minions you must kill in succession afterwards which are the conjugations (eg: "soy). 2b. You have a health bar, and getting answers wrong takes away health. The large enemies (infinitive verbs) for example could do more damage than the conjugations.
    3. occasionally, other things that are not spanish words will be sprinkled in that might be little power ups or something that you can use (maybe they can be used for hints or something)
    4. You go through levels, gain character experience, etc. - which of course is all superfulous, but might make it feel more like the player is actually making progress - trying to get a high score.
    5. As you gain experience, you could upgrade the visual look of the character or unlock things.

    –These are just some ideas I was brainstorming, and if I am doing this myself, it will take me a while so it would be a slow and incremental process - but I think this might be one of many possible ways to keep it vocabulary focused, while still making it feel more like a game than learning. I’d of course be open to any ideas though.

    The mandarin version already made feels more like a flashcard app, with minimal gamified aspects, but I think if I was to start over on the Spanish one, I would lean more into the game feel - with more visuals and such. I think it would also be nice to add a toggle option to either type out the answer or select it with the mouse. Of course, depending on the time and effort, there are a lot of avenues to take, it just is a matter of actual implementation. Getting artists on board would be nice as well, since we wouldn’t have to rely on the Creative Commons sprites and could get a more unified theme. Probably wishful thinking, but I am thinking of best case scenario.

    Sorry for the long answer, I realized I ended up rambling on a bit. :D

    EDIT: I threw together a super quick mockup to kind of see how it would work out - and I’m not actually 100% sure how it would work out. . . it definitely brings to light the fact that it will need some more brainstorming. The fact that there are so many conjurations (present and past tense) complicates it a bit. I hate to have a new word for every single tense, and prefer to have them grouped so that it is easier to process, but it might be information overload with how many there are? Super quick mockup







  • The hanzi drawing needs to be a bit more lenient, I can’t draw that accurately in a small touch zone on my phone

    The Hanzi section is using a open source library (https://hanziwriter.org/) - so I’ll have to see if they have an easy way to adjust this.

    杯子 and 杯 can both show up at the same time as “cup; glass” but only one button will work for each

    Ahh yeah, I have noticed this issue - and am not sure yet the best way to tackle duplicates. I guess for now I am going to just say “it’s a feature, not a bug” /s - lol. It’s like those memory games where you have to remember the correct one lmao. No, but in all seriousness, yeah - I need to tackle this. (there are a few other issues still needing fixed too - like the font selection to choose different fonts)

    This is awesome!

    Thanks! It definitely needs work - but I think its at least functional enough to be useful(ish).

    [EDIT]: Thinking more about the duplicates: I am not a Chinese speaker, so I wouldn’t feel comfortable doing it, but someone who has more nuance to the differences in the words could probably just update the JSON translations so that duplicates don’t exist. The JSON data for all the translations were harvested from APIs - but could easily by hand tweaked by a knowledgeable person.

    [EDIT 2] Fixed leniency - easier to draw and not be as precise on the hanzi writer.





  • Yeah, I’m not a huge fan of the AI approach, I would love to build it without, but I simply am not a good enough programmer to do it without (I have tried in the past, but eventually give up).

    AI has allowed me to actually get the digital version farther than I have ever gotten it before - and on one hand it’s nice that I can actually create something, but on the other hand I am afraid it will turn off people who would have otherwise possibly been interestes in contributing. . . But without anything created at all, I would have probably a harder time getting contributors - so a catch 22 in a small way.

    But yeah, my current plan is to try and get the project as far as I can using AI, in the hopes that an experienced programmer will eventually be interested in contributing and cleaning up the mess I’ve made. Probably not the smartest strategy, lol, but it’s the only one I’ve got. ¯_(ツ)_/¯


  • Just a couple if days ago I converted my Chromebook (~4gb ram) to Linux with coreboot and installed antiX, then changed over to MX Linux, and I think MX is a great for what it is. I’m using the fluxbox window manager (provided as one of the options on MX) because it’s lightweight enough for the Chromebook, and my Chromebook runs far faster on Linux than it ever did with chromeOS.

    If nothing else, I suggest looking into MX - I’m a happy customer. I think antiX was actually great too, and a bit more minimal - but the graphical interface of MX out of the box felt a bit more polished and was worth while for me.