Looks nice. The capabilities are intriguing if it works well. At first glance, the homepage should probably say what languages you can learn.

As it is designed solely for English and Spanish speakers, I, being Italian, am unable to use it.

A small feedback is the lack of a login with Amber or similar, only with nip-07 it is limiting

I’m working on that but it’s somewhat painful to solve, originally the app was just for English and Spanish learning and I didn’t imagine I’d be adding more languages until I got interested in some indigenous languages in Mexico. Tech debt from vibe coding 💔

Don’t worry, it’s still very kind of you, and the website is lovely 🫂

Is amber nip-46 or something different altogether? Im looking this up right now

Great,it Very interesting. It would be better in both English and Spanish.

By this do you mean like a bilingual mode?

So much better than commercial apps!!!! 💪🏼💪🏼 Easy, friendly, nostr connect... All I want for christmas 🥳

Nice design.

Thank you for the kind words, that’s validating to hear.

Yes, I think that would be great

Thank you

the AI conversation focus is 🔥 — this is where most apps are heading, and it’s way more engaging than static lessons. I’d want a clearer “aha” moment early on—what makes this special within the first 30–60 seconds?public learning logs / zaps for progress (turn learning into content

Is nip-46

It’s the part I suffer over most, I try to care about every interaction and there’s always something I find that can improve 💔

That’s something to chew on for a bit, thank you

Hmmm the UI doesn’t seem to be showing the zap. Did you receive this?

You can build on this!

GitHub

GitHub - kizuna-ai-lab/sokuji: Live speech translation powered by on-device AI and cloud providers — OpenAI, Google Gemini, Palabra.ai, Kizuna AI, Volcengine, and more

Live speech translation powered by on-device AI and cloud providers — OpenAI, Google Gemini, Palabra.ai, Kizuna AI, Volcengine, and more - kizuna...

I tried it, it was fun as hell, the best way I've ever tried to learn a language.

Oh this is pretty amazing 😻 the languages offered Greek, Nahuatl not only German French, the usual! It was pretty cool. Though, it would be great to sign in with an extension. I like that you’re able to create study groups !

Thanks for the encouragement

I feel like the primal UI is misleading me, did you receive the zap?

I feel like the primal UI is misleading me, did you receive the zap?

I feel like the primal UI is misleading me, did you receive the zap?

I tried it out. A very nice app, I can imagagine it can catch user engagement. In the initial test the AI voice was often interrupted.

Oh no, in what way was it interrupted? Did it not finish reading the text?

Did you receive the zap? It shows in the UI but it’s not showing up in my transaction history on primal 😵‍💫

No I did not, my alby should be on.

Something’s up, it’s showing for some in my transaction history but not for others 🤔

no 🧐 😮

Something is up. Some folks are receiving it but others aren’t in the transaction log

anything on browser is good - it need mic access so will try later

I checked my wallet but there’s nothing there. 😪

I believe there’s something wrong with your configuration on Zeuspay. I went on your profile and pressed the zap button in your header and I got this error

Hmmm sending it directly to your alby address in your profile didn’t seem to work. Someone else had a similar issue but they were using Zeuspay too

Looks great and very straightforward, I like that the web is available in Spanish too, I'd suggest to put somewhere which languages are supported but overall great 😁

Pretty strange i sent myself some sats from another wallet to the Zeus one and i received them 🥹 I’m not using nwc or coinos, could that be it?

Great app!

No, please try this address instead: rodant@coinos.io. Thanks!

Sentences were cut off often.

👀

First, save the sats, I don’t need to be bribed, lol. Second, let me start by saying that you did a hell of a job creating this PWA, it looks absolutely gorgeous. Criticisms, just with the first few minutes interacting with it, maybe you could add the translation of the app name in () next to the person’s native language or whichever language they choose as their primary language? IMHO, that would be a great way to start the experience. Next, in the timer section, allow the user to interact with the graphic to choose their allocated time, using their finger to drag the minute hand around. You could do this several times with some sort of symbol to show you are already past an hour. For example, I want an hour and a half, I would drag my finger around the dial one and a half times. This is an edge case because no one had that kind of time to dedicate to learning a language, Most of those presets are crazy and unrealistic but you could still choose them by simply moving your finger around the dial until it coincides with your choice. Three full revolutions would set the timer at 3hrs, two revolutions sets it for two hrs, etc. That would clean up that whole section while making your choice a little fun. Fun is the keyword here if you want people to actually give this a try, I’m not going to try and learn a new language if the setup is boring. I will offer some more criticisms as I delve deeper into to this so I will bookmark this post. If there was one feature I would like, it would be a physical app with the ability to download all of the pertinent information for my language of choice so that I can use this completely offline as well. You may be able to add this to the PWA, not sure, but if I am being completely honest, I am not really a fan of PWA’s, PTSD from using Android for a decade has turned me off of those types of apps. 😜 Sorry about the novel, it’s how I always interact, I don’t speak emoji, LMAO.

Support Language portuguese is a good Idea (400 million people in the world )

I did, thank you. Following this project.

Not sure what the problem is, I send sats using the zap button and it shows I send zaps, but it doesn’t record in my transaction history, and going into a profile to send a zap results in an error. At first I thought it was a problem with their accounts, but I’m able to send through other wallets. View quoted note →

Your lightning node or provider might not have good peering?

Tried Dutch and it subtracted marks for me answering a bit to slow and then thinking about what to say. Telling about a book I read and what I learned did make me think as I answered. It seem to be a bit hasty on that part. It also didn't recognized a few words correctly. Not sure why. Tried it a second time and with different answers but the same results. What I didn't like was the stacking of questions. What is your name? How are you? What are you doing? Why are you doing it? It doesn't come over natural if you get what I mean.

I think if you make less options and simple flow to fill in the blanks it would be easier especially in case of spanish language.

I could add an option to adjust how quickly it responds, it’s kinda tough getting that right in a way where it feels smooth without too much waiting between turns, but it probably feels worse getting cut off too early. Thank you for the feedback

I had an issue sending you a zap in post so I went to your profile to send it. Where in the app do you mean by having less options? Was it a certain question type you had? The vocabulary and grammar modules have a mix of different question types. Just making sure I understand correctly

I think this could be what I have been looking for. It could be my setup but the mic wasn’t picking up anything I was saying and I had to skip all the speaking portions which was a bummer. Will try again and report back. Thanks for your work on this. One more thing, Where does the data for this live? If it’s built on nostr one could theoretically “take it with them” yes?

So far it’s a partial implementation with nostr using auth because it affords some nice UX gains with instant logins and cross platform accounts (I also work on a coding education app at https://robotsbuildingeducation.com) However it’s not fully nostrified where all of the UI data is fetched and stored to relays, I’m still learning nostr development but between marketing, sales, content creation, design & engineering I can’t always afford to spend so much time on certain implementation details that most people won’t see or feel (most of my time is in consumer spaces like Instagram and TikTok)

Hi, I am working on a learning app for children myself right now (not speech based, though). Here are my thoughts: * The UI is very appealing. * Setup in the beginning is a bit convoluted. Start with defaults instead of bombarding users with options. * BUG: Whitespace between words is not stripped and creates false results if there is more than a single whitespace. * Replaying generated speech actually creates new output instead of merely replaying. This could be optimized. It wastes compute (unless intentional, of course). * I was really surprised when out of nowhere a game popped up. That felt a bit cringe. My general recommendation: Focus on the core functionality. Right now, there is a lot of distraction. That sharply contradicts the idea of setting "focus time" with a timer. Good luck.

Thank you for the feedback. It seems like the primal wallet is having issues sending zaps again. It’s been kind of off & on but I’ll make sure to zap you as soon as it’s available

Tried it since I am learning Dutch, overall very nice exercises which lead you and teach you the same words through different examples. I would appreciate if I could tune in the app the turn detection. Maybe I could but I didn't realize. Would have been quite useful if I could tune the turn detection whilst making the speaking exercises. Second point of feedback is on the exercise that I have to repeat a dutch sentence. I was repeating the sentence then pressing the stop button (I think) but then it was not recognizing that I said what I wanted to say, just later I realized that I had to wait. If I were you, I would make it more apparent in the ui, that I have to wait.

Seems to be a trend with this piece of feedback so I’ll have to prioritize that soon

I gave away about 17,000 sats requesting feedback on an app I’m developing. 1000 sats is admittedly not much to offer for people’s time, but it was insight and good community fun. Anyone that knows design knows that user feedback is absolutely worth its weight in gold. I think next time I do something like this the offer will be 5,000-15,000 sats for feedback, as a way to incentivize even more nitpicking and depth. Highly recommend doing this to any app developer here. View quoted note →