You are browsing as a guest. Sign up (or log in) to start making projects!

Frapanol

  • 9 Devlogs
  • 15 Total hours

I am building a website that helps with french and spanish because many people struggle with this language especially the basics so this will be a great help to them.

Ship #1

I named the app Frapanol, which is a combination of french and Espanol

What I made: A complete french and spanish learning app in one html file (with 2000+ lines of code) including 13 learning modes (vocab, flashcards, quiz, listening, spelling, conjugation, phrases, numbers, and grammar), an AI tutor implemented through a Hack Club API, achievements, daily goals, speech synthesis for easier pronunciation, manual voice picker and a warm yellow/red theme that changes depending on the language. All of it works offline (except for the tutor) in your browser, and everything is saved in your device (not in the web app's servers).

Challenges and what's working well: The two main challenges were the API key and the browser speech. With the API key, the problem was that Hack Club's server doesn't allow other web pages to access it (because of CORS policies), so I made a small Node server that runs locally and receives the requests (your API key is only stored in your browser, not in the file). The other big problem was the voices: Chrome and Safari have strict policies about when you can autoplay audio, and the auto-detected voices weren't working properly on real devices (probably because they had old voices installed, or voices that don't work with the API, like Eddy or Flo, the voices made by Apple for testing). To fix this, I made it prioritize the manually selected voice, and then the default browser voice, and report the status to the user. How to test it: Open http://localhost:3400 in a browser, open Settings, put your API key in the "AI tutor key" field, select a voice in the "Voice" dropdown menu (you can try several ones and click "Say hello"), and test it by chatting to the tutor. It should work smoothly after you give microphone permissions to the web app and enable the "Tutor" setting. All in all, it's a fully working PWA (progressive web app) in any browser that lets you configure it with some settings.

  • 9 devlogs
  • 15h
Try project → See source code →
Open comments for this post

4h 22m 17s logged

In general, the described scenario was highly challenging. All features were properly implemented, and the design and color scheme were modified to a red-yellow configuration. Nevertheless, the process of working with a speech feature has taken a lot of time and resulted in multiple issues. The situation where the speech function ceased to work remains unexplained and worrying.

During two or three hours, I tried to figure out what the issue was with my assistant, but all my solutions failed. In addition, I tried several functions to find the one that would work, but nothing worked. Finally, I managed to solve the problem when I realized that the issue was not on my side but on the browser’s side since the speech output in my browser did not function correctly.

I decided to try the program in Safari, and everything started to work fine. What surprised me was that the code worked fine on the phone. Thus, the problem was in the browser. However, as mentioned above, this project also involved a great number of challenging tasks and unanticipated issues. For instance, a long procedure was required to connect an API key, during which I faced multiple obstacles that I managed to overcome due to the help of an AI assistant and my own time investment.

Hence, the project was fruitful but included many issues on the way. :OVERALL THE BACKGROUND FEATURE IS MY FAV Lol

0
0
8
Open comments for this post

4h 18m 34s logged

The progress is going very well recently. I have finished all the core logic of this project, and I had a tremendous amount of fun doing that. I did a lot of research work and testing, and with the help of AI, I automated a lot of tedious work of rewriting and translating. Instead of spending hours trying to find the right phrase, I could focus my time and effort on actual development of the product.At this point, I think that the product is about 75%-80% complete. The major part of the work is done, and what is left is relatively simple, though time-consuming. The next big step is to add the functionality of listening and dictating. This will allow the web application to receive input from the user via voice. With that, the app’s functionality will be complete, and only some final polishing and logic tweaks will be needed to complete the project.I am very satisfied with the progress I made recently, and I am looking forward to working on the project in the future. Most challenging tasks are done, and the remaining ones are relatively simple.

0
0
2
Open comments for this post

31m 29s logged

I have finally completed all the phrases, vocabulary, and the Spanish and French portions, and to be frank, I was able to do that thanks to the AI tools. Firstly, I arranged everything into one neat document so I could easily work with the data. Then, Claude aided me in moving forward in some of the coding parts so I had a good basis for the rest of the work. Then, I did the rest of the job on my own. As for the pronunciations, I looked it up on Google, while the bulk of the work – the translations and polishing – was done by the AI. Additionally there was many errors but then I figured out the issue was one { out of place lol. Overall its going good so far and way faster then I thought it would. I also saved it with local history etc.

0
0
3
Open comments for this post

36m 20s logged

For the Spanish and French vocabulary, phrases, and pronunciation components, I’m planning to combine Google Translate with AI‑based language tools. Since I’m not fluent in either language, these tools will help ensure accuracy and consistency. My role will be to design and structure the entire architecture for those sections – the logic, the flow, and how everything connects – while the AI systems handle the actual generation of translations and pronunciations.I’m still checking whether Google Translate provides reliable pronunciation outputs; if it doesn’t meet the standard I need, I’ll switch to another AI model like Claude to fill that gap. Either way, the core idea is that I’ll build the framework, and the language tools will supply the linguistic content.

0
0
7
Open comments for this post

1h 38m 32s logged

I finished majority of the logic in the website and now everything works so far but I would say I am like 45-50 percent done because I still have to code the css and other parts of it. I also downloaded local host to track my progress. And now comes the tough part…I gotta add all the phrases in the different languages etc so that the users can use it etc with the phrases and everything. “Greetings 👋”:[ [“Bonjour”,“Hello / Good morning”,“bon-ZHOOR”],. I am going for a colourful tone because I want it to look fun and after i am done this I am going to add an thats possibly powered by hackclub which will make everything more easier and good to learn. Ai is also helping me with a few things like blocks etc.

0
0
6
Open comments for this post

22m 18s logged

Currently I am working on making tabs for each part and for now I am doing all these in the screenshot below and also instead of getting emojis manually I lowkey just asked ai to generate me the emoji and it gave it to me which was faster and more efficient. I also learned what non breaking spaces are or for short known as  ! . Overall in these last 20 min ive learned a good amount and I have coded like 60-80 lines of code which is pretty nice and I am looking forward to add more and if I would have to guess I am like 26.873 percent done lol but its fine the rest is pretty straight forward after I finish the logical aspects of the code.

0
0
2
Open comments for this post

1h 47m 56s logged

Currently I am working on the html part and I made different functions which so far include phrases, flashcards, quizes and possibly games. With a bit of ai I learned some more aspects in css and they also helped me code like 200-300 lines of code. So my next goal is honestly adding more and more and adding a language switcher function which can switch languages for me so basically kinda like duolingo but mine is gonna be more fun because I am going to try to add more stuff. PLUS I added a progress function, which is honestly helpful for individuals that wanna lock in and learn with efficiency and effectivity.

0
0
3
Open comments for this post

1h 3m 22s logged

Ai helped me a lot with debugging and fixing small errors. SO I finished like a lot of html atleast 100 or 200 lines of code and it is going good so far but I also wasted time in trying find out what phrases to put in and in the future I was possibly gonna put a tutor entirely powered from ai using the hackclub api which will be a stretch but its fine and I will try to apprehend these aspects.

0
0
2
Open comments for this post

20m 50s logged

I am trying to create a french website that will help individuals learn french and ill add a spanish feature too with it so basically it will teach both french and spanish. Currently I started with the french and I am done around 40 lines of html but I may have to restart and also I learned new things like “antialiased” which is a term I learned from the github copilot which was nice… next I wanna keep working on the html and making everything good. I also connected github and downloaded Localhost so none of my code gets deleted or anything in the future.

0
0
4

Delete project?

Are you sure you want to permanently delete this project? This action cannot be undone.

All devlogs, followers, and associated data will be removed.

Followers

Loading…