14 ms·

Show HN: Building an open-source language-learning platform

159 points by kantord 7y ago

onyva 7y ago
Wonderful initiative. Would have loved to get involved but I really don't want to use GitHub. Any idea how people may contribute using other platform?
- zadokshi 7y ago
  Git is designed to be distributed, you could push to someone else’s repo and they can then push to GitHub for you.
- andrewzah 7y ago
  See if the maintainer is willing to accept email patches. You can clone the repo from the github link, make changes, and generate a patch which you can send to a maintainer. Unfortunately it's a bit tough to avoid github, it's like the facebook of programmers. I don't see any particular reason why it's bad though? If anything, it might improve with microsoft now owning it.
  - kantord 7y ago
    I can also merge from GitLab (or other server) forks too, I think? I've never tried doing that, but it sounds like something that should be trivial
helsinkiandrew 7y ago
Great effort - my thoughts as a learner of Finnish and perpetual searcher for internet learning content is that the biggest issue facing any learning platform (commercial or open source) is content, content, content. It takes a huge amount of effort to build decent course material - most open source material are from a community of people that are also using the material - will this be the case with language learning?
- kantord 7y ago
  yes. It's gonna be a lot of work to make that happen, but I imagine the course editor platform being open to more people, with different levels of involvement.
- yorwba 7y ago
  One way to get content would be https://tatoeba.org https://tatoeba.org For example, to teach "oso" = "bear", there are already ten different example sentences with audio recordings: https://tatoeba.org/eng/sentences/search?query=oso&from=spa&to=eng&has_audio=yes&trans_filter=limit&trans_to=eng https://tatoeba.org/eng/sentences/search?query=oso&from=spa&... (Okay, one of them is actually "oso" = "dare".)
  - ng7j5d9 7y ago
    The Nordic Language Processing Laboratory has a sizeable collection of open source parallel texts: http://opus.nlpl.eu http://opus.nlpl.eu Click on any of the headings in [brackets] on the top of the page to see the language pairs available for download for that collection.
  - tasogare 7y ago
    The problem with tatoeba is that the content doesn’t seem to be curated. I checked the translation of I love you in Japanese recently and more than half of the listing is hot garbage that would never be uttered by a native speaker. It’s just like these sentences where contributed by learners who just finished their first lesson.
    - yorwba 7y ago
      Japanese is a special case, because most of the Japanese sentences come from the Tanaks corpus, which... consists of sentences written by learners of either English or Japanese who likely weren't all that qualified for the task. However, the content is actively getting curated. E.g. https://tatoeba.org/eng/sentences/show/137571 https://tatoeba.org/eng/sentences/show/137571 is a translation of "I love you." owned by a native speaker. The search form doesn't make it easy to find native-speaker translations, though. The search option to limit results to native-speaker sentences only works if you're searching in the target language. E.g. if you want usage examples for "愛": https://tatoeba.org/eng/sentences/search?query=%E6%84%9B&from=jpn&to=und&user=&orphans=no&unapproved=no&has_audio=&tags=&list=&native=yes&trans_filter=limit&trans_to=und&trans_link=&trans_user=&trans_orphan=&trans_unapproved=&trans_has_audio=&sort=words&sort_reverse= https://tatoeba.org/eng/sentences/search?query=%E6%84%9B&fro... If you're only using Tatoeba as a data source for automatic creation of course material, though, you can filter using whatever criteria you want.
thepete2 7y ago
Love it. Seems a lot like duolingo, but with a lighter UI. Can you please add shortcuts, so you can type 1, 2 or 3 instead of clicking?
- kantord 7y ago
  you can already do that. Maybe it's not very clear from the UI, but it basically has the same keyboard shortcuts
  - thepete2 7y ago
    You're right. Vimium (browser extension) was the problem :)
qwerty456127 7y ago
Cool! Please make sure to fix what I consider the major problem of Duolingo: it asks so much stupid questions (things I already know perfectly, even if I'm going to forget these tomorrow these still are nonsensical to repeat that much during the same session) I get bored and start clicking too fast so I make mistakes out of pure inattention and get even more stupid questions as the result.
- kantord 7y ago
  Maybe your ideal way of learning or type of platform is different. I have thought of experimenting with the possibility of having different way of presenting the same content. For example, instead of a repetitive software solution, exporting the course material as a printable book.
  - qwerty456127 7y ago
    No. I like the repetition way (it works and it provides instant gratification to support the flow state unless repetition rate is t0o intense) and what I see in the current version of LibreLingo is ok. Just don't overdo. Or make repetition intensity configurable.
    - madacol 7y ago
      Maybe what you want are options like AnkiDroid does: https://lh3.googleusercontent.com/QFBCiyA_6WBKfr8Xoo1MqOrQw18t6KE7POlKV3BfA7fLETjnYYu51y2SMAliHTpd6A=w1860-h928 https://lh3.googleusercontent.com/QFBCiyA_6WBKfr8Xoo1MqOrQw1... At the bottom there are 3 buttons that sort of means this: - repeat card more frequently - repeat card - repeat card less frequently Android app: https://play.google.com/store/apps/details?id=com.ichi2.anki https://play.google.com/store/apps/details?id=com.ichi2.anki
- yorwba 7y ago
  One way to distinguish mistakes of inattention from "actual" mistakes is to leave marking the mistakes to the user. So if you get inattentive, you simply forget to mark yourself wrong. My personal system works that way; though not for any particularly clever reason. I just have no idea how to automate the checks. I don't recommend anyone to try it unless you want to spend a few hours setting it up, but there's a screenshot in the README: https://github.com/Yorwba/alphabet-soup https://github.com/Yorwba/alphabet-soup
petargyurov 7y ago
As someone who is learning Spanish at the moment, here is some feedback. - I think a single click is sufficient to choose the correct answer and submit it; alternatively let me double click to choose and submit. - I don't think a "Continue" step is necessary - Without selecting an answer and pressing Back on the browser, the UI freaks out for a moment (Firefox 72.0.2) - I love cheatsheets, so having a section for these would be awesome (finding good, consistent cheatsheets on the web is hard) It is understandably a little sparse at the moment. Take a look at an app called Memrise - I think the style of its exercises would fit well here. I particularly like the "Fast Review" exercise which lets me do a speed review of all the words and phrases I have learned so far. As others have mentioned already, it would be nice to be able to mark words you already know and stop them from showing. Additionally, being able to mark words you find difficult could appear more frequently. I will keep an eye on this project; good work so far!
- melling 7y ago
  I built a cheatsheet for my iOS language app for Spanish: http://h4labs.org/100-words-app-spanish-pdf/ http://h4labs.org/100-words-app-spanish-pdf/ I’ve also created some Spanish Word Search PDF’s. Most are on Github: https://github.com/melling/Sopa_de_Letras https://github.com/melling/Sopa_de_Letras I was going to create a little book of 20 Spanish Word Search puzzles. What’s your idea for Spanish Language cheatsheets? I’ve tossed around a few ideas myself. Verbs, ser/estar, por/para, saber/conocer
  - petargyurov 7y ago
    I find grammar cheatsheets quite useful, for example: https://i.pinimg.com/736x/0c/b6/81/0cb681ce3173c647a39b12db88344207.jpg https://i.pinimg.com/736x/0c/b6/81/0cb681ce3173c647a39b12db8... Anything that explains grammatical rules and gives me a "template" of sorts is great because it doesn't give me the answer straight away but it engages my brain to adapt the word I want to the given rule. Topics like tenses, conjugation, pronouns are what I like having cheatsheets for.
- _def 7y ago
  I agree with your points, controlling the app entirely with the keyboard seems pretty nice though
melling 7y ago
I’ve been working on my own iOS language apps for quite some time: http://appstore.com/h4labs http://appstore.com/h4labs My original idea was to do lots of little games in one app to keep it interesting: Hangman, Word Search, 4 Pics 1 Word, ... Recently, I’ve been breaking out the games into smaller apps. I think that’s better. Take your data and try to make some fun little games. Also, I think noun gender is important. Should learn it at the same time as the noun. It can change the meaning of the word. Finally, is anyone doing verbs? I have a simple Spanish verbs app. There are usually many rules to help make it easier. I was going to open source the rules that I have but there doesn’t seem to be much interest: https://github.com/melling/Spanish_Verbs https://github.com/melling/Spanish_Verbs Need to extract the data from my site: http://www.h4labs.com/lang/es/index.html http://www.h4labs.com/lang/es/index.html
bayesian_horse 7y ago
Good idea. I personally love Duolingo. But I fear that the course data, which is mostly user contributed, will eventually disappear if the company goes under...
- kantord 7y ago
  same, that's my biggest fear
- Vivtek 7y ago
  There's a lot of politics involved in that course data, too. That an open-source approach exists is a very good thing.
- zelly 7y ago
  Some hero should scrape it.
  - lambertsimnel 7y ago
    Would that be legal? I wonder who holds the copyright to user-contributed data in DuoLingo (assuming they're copyrightable). If some of the user-generated content isn't copyrightable, or was contributed by users willing and able to share it with a FOSS project, could only that data be scraped, or would it be too difficult to identify?
  - bayesian_horse 7y ago
    One way is to get Premium and download the course. I haven't looked at it, but I assume they haven't bothered to do any copy protection on those data packages. Not sure if they contain account-bound watermarks of any kind.
RobAley 7y ago
One small piece of feedback : My (UK) keyboard doesn't have accented characters on it, and I don't know how to type them (well, I do, but many/most users won't), so typing "leon" instead of "león" for Lion wouldn't let me continue as there was a spelling mistake. Ideally, have an on-screen keyboard for accented characters, or instructions on how to type them, or less ideally allow spelling mistakes on such characters.
- kantord 7y ago
  thanks! If you have a GitHub account, feel free to create an issue for that! https://github.com/kantord/LibreLingo/issues/new https://github.com/kantord/LibreLingo/issues/new
cerberusss 7y ago
I'm not sure if it's a joke or not, but that web page its title has a typo. It says "langauge" instead of "language".
LaurentS 7y ago
Hi kantord, congratulations on the site! I tried something similar about 10 years ago, called wikibabel. I then joined with another site called wikiotics.org. We were very focused on libre everything. In the end, we gave up after several years, but our content is still up and we've been looking for a new home for it. Please take a look and see if any of it would be useful for you (we had built a bunch of cards very similar to your demo lesson). We had stuff in a few dozen languages, including audio recordings (mostly of low quality). I've moved on, but I still love the goal, and would be happy to share some feedback on the experience if you want (contact on my profile page). In short: - focus on what the user wants, not the tech. - is English->Spanish the best language pair to start with? You're directly competing with the giants like Duolingo. I tried to learn Burmese a few years back, I would have loved to find something like this for that niche. Good luck!
- rjeli 7y ago
  Cantonese also has a ton of speakers across the world, and the best I could find a few years back was the FSI tapes
  - fartcannon 7y ago
    I would also love Cantonese in a Duolingo style, please!
    - computerfriend 7y ago
      Can't agree more. It is amazing how few resources there are for learning this amazing language.
- Vivtek 7y ago
  That's a pretty neat site (Wikiotics, I mean - wikibabel appears to go to multiple places).
skinnyasianboi 7y ago
Good work and the roadmap looks pretty promising. I hope we will see native mobile apps. Would be a cool case to play around with SwiftUI and Jetpack Compose.
dwrodri 7y ago
I am far from the first to say this, but please do push hard for languages that aren't commonplace (or well-developed) on other platforms. My language of choice would be Farsi.
cptwunderlich 7y ago
I also wanted to congratualte you for your effors. I'm a daily and long time user of Duolingo and thought of libre alternatives for a while. I'd agree with some sentiments in this comment section, that you might want to find a niche, as competing with Duolingo or Babbel would be difficult. Duolingo doesn't too well with "smaller" languages and different scripts. Are you using a TTS engine for the voice (I assume)? I was looking for TTS for a smaller language I'm studying, but I couldn't find anything. I hope that something comes out of the Mozilla's Common Voice project.
- johnkpaul 7y ago
  I totally agree with this. I'm an avid Duolingo user with an over 2000 day streak but I would happily use another tool/game/platform if I could learn Malayalam. My family is Malayalee and I know the entire language in my brain because I can understand them speaking to me,but I respond in English. I would pay a huge amount of money for an English<->Malayalam instructional book and I'd pay a huge amount of money for a Duolingo-similar Malayalam learning experience. Niche and rare languages have little to no representation in the popular apps.
- kantord 7y ago
  I still think it makes more sense to implement languages that more people want to learn. However, the long term goal is making it easy for the community to build courses, so once the project is mature, it should be possible to include a way larger number of languages. I am also thinking of things like conlang enthusiasts being able to create courses for their own conlangs.
  - mumblemumble 7y ago
    I just wanted to chime in and say I think there's another good reason to focus on the more widely-spoken languages first: It's not just that there are more people who want to learn them, it's also that there are more people who can help contribute to the materials.
    - kantord 7y ago
mumblemumble 7y ago
Completely tangential: The Rosetta Stone / Duolingo style of language learning has never really worked for me; I generally have more success and an easier time staying engaged with the "comprehensible input" approach. Are there any open platforms based on this method out there? Maybe something along the lines of LingQ, though perhaps without all the attempts at gamification.
- bluGill 7y ago
  I think the best approach is a few weeks with a duolingo style learning just so you get SOME grammar and a FEW words. Once you can understanding things then input is best, but if you don't understand input isn't going to do much good. If you can find English [any language you know] subtitles all input is comprehensible and it is best - but also the hardest and most frustrating. If someone could just create a list of of youtube videos with subtitles in my language it would be huge. Doesn't matter if it is a blacksmith explaining how to shape metal, or a knitter showing some pattern, I just need input.
  - mumblemumble 7y ago
    TBH, I don't even personally like Duolingo-style programs for getting down the base vocabulary, because, while it's a generally more pleasant experience than other options, I find the approach to be relatively slow and inefficient. Here's a video (in French) that I found reflects my own experience quite well: https://www.youtube.com/watch?v=-KQ1qRJ2wQc https://www.youtube.com/watch?v=-KQ1qRJ2wQc In a nutshell, I've got two complaints. First, while flashcarding is wicked efficient (I'd say essential) for review, it's not a particularly good tool for learning things in the first place. And an overly flashcard-oriented approach has real problems with a sort of streetlight effect: You tend to get stuck on what's easiest to illustrate with pictures, which often isn't what's most useful for communicating with others. edit: Also, for what it's worth, the existing research seems to indicate that, while subtitles in our target language can be hugely beneficial, subtitles in a language you already speak actually inhibit language acquisition: https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0007785 https://journals.plos.org/plosone/article?id=10.1371/journal... (There was another study with Brazilian English language learners that I found more compelling, but I'm unable to find it right now.)
songzme 7y ago
When my wife and I first met, we tried to build a language learning app (don't sign up, we stopped working on it) - https://llip.io/landing/ https://llip.io/landing/ To get users, we started a meetup group to teach people Korean. https://www.meetup.com/San-Jose-Korean-Language-and-Culture-Meetup/ https://www.meetup.com/San-Jose-Korean-Language-and-Culture-... We stopped working on the app, but continued with the meetup group (3 years now and almost 1000 members). Our conclusion is that the most effective way to learn a new language is a simple commitment to showing up regularly to language events. Meetup.com or your local public library will usually have regular (and free) events. Once you have that commitment, picking a great app will help expedite your learning. Excited to see an open source language learning app!
- generalpass 7y ago
  What if you live in a region where you can't people speaking or learning the language you want to learn?
  - songzme 7y ago
    That's what I originally thought too (couldn't find an existing group) but shamelessly I started one with my elementary grade Korean. Students much better than me came and I was able to ask them for help. Have you thought about starting a group at a local library or coffeeshop?
  - elliekelly 7y ago
    I've had a lot of success using a Discord server to learn French. The big gaps in my self-learning knowledge were pronunciation and conversation. (It's one thing to be able to read/write the language at your own pace but different to be able to "think" and react in that language in conversation.) Discord chat rooms are great and thanks to the voice channels I've almost got the 'r' down. Most of the servers even have user "flair" so you can let others know if you'd like to be corrected on your mistakes or if you prefer others let them slide.
- tasogare 7y ago
  Great landing page. I’m working in the language learning space but lack the design skills.

159 points by kantord 7y ago

Wonderful initiative. Would have loved to get involved but I really don't want to use GitHub. Any idea how people may contribute using other platform?

zadokshi 7y ago
Git is designed to be distributed, you could push to someone else’s repo and they can then push to GitHub for you.
andrewzah 7y ago
See if the maintainer is willing to accept email patches. You can clone the repo from the github link, make changes, and generate a patch which you can send to a maintainer. Unfortunately it's a bit tough to avoid github, it's like the facebook of programmers. I don't see any particular reason why it's bad though? If anything, it might improve with microsoft now owning it.
- kantord 7y ago
  I can also merge from GitLab (or other server) forks too, I think? I've never tried doing that, but it sounds like something that should be trivial

helsinkiandrew 7y ago

Great effort - my thoughts as a learner of Finnish and perpetual searcher for internet learning content is that the biggest issue facing any learning platform (commercial or open source) is content, content, content. It takes a huge amount of effort to build decent course material - most open source material are from a community of people that are also using the material - will this be the case with language learning?

kantord 7y ago
yes. It's gonna be a lot of work to make that happen, but I imagine the course editor platform being open to more people, with different levels of involvement.
yorwba 7y ago
One way to get content would be https://tatoeba.org https://tatoeba.org For example, to teach "oso" = "bear", there are already ten different example sentences with audio recordings: https://tatoeba.org/eng/sentences/search?query=oso&from=spa&to=eng&has_audio=yes&trans_filter=limit&trans_to=eng https://tatoeba.org/eng/sentences/search?query=oso&from=spa&... (Okay, one of them is actually "oso" = "dare".)
- ng7j5d9 7y ago
  The Nordic Language Processing Laboratory has a sizeable collection of open source parallel texts: http://opus.nlpl.eu http://opus.nlpl.eu Click on any of the headings in [brackets] on the top of the page to see the language pairs available for download for that collection.
- tasogare 7y ago
  The problem with tatoeba is that the content doesn’t seem to be curated. I checked the translation of I love you in Japanese recently and more than half of the listing is hot garbage that would never be uttered by a native speaker. It’s just like these sentences where contributed by learners who just finished their first lesson.
  - yorwba 7y ago
    Japanese is a special case, because most of the Japanese sentences come from the Tanaks corpus, which... consists of sentences written by learners of either English or Japanese who likely weren't all that qualified for the task. However, the content is actively getting curated. E.g. https://tatoeba.org/eng/sentences/show/137571 https://tatoeba.org/eng/sentences/show/137571 is a translation of "I love you." owned by a native speaker. The search form doesn't make it easy to find native-speaker translations, though. The search option to limit results to native-speaker sentences only works if you're searching in the target language. E.g. if you want usage examples for "愛": https://tatoeba.org/eng/sentences/search?query=%E6%84%9B&from=jpn&to=und&user=&orphans=no&unapproved=no&has_audio=&tags=&list=&native=yes&trans_filter=limit&trans_to=und&trans_link=&trans_user=&trans_orphan=&trans_unapproved=&trans_has_audio=&sort=words&sort_reverse= https://tatoeba.org/eng/sentences/search?query=%E6%84%9B&fro... If you're only using Tatoeba as a data source for automatic creation of course material, though, you can filter using whatever criteria you want.

thepete2 7y ago

Love it. Seems a lot like duolingo, but with a lighter UI. Can you please add shortcuts, so you can type 1, 2 or 3 instead of clicking?

kantord 7y ago
you can already do that. Maybe it's not very clear from the UI, but it basically has the same keyboard shortcuts
- thepete2 7y ago
  You're right. Vimium (browser extension) was the problem :)

qwerty456127 7y ago

Cool! Please make sure to fix what I consider the major problem of Duolingo: it asks so much stupid questions (things I already know perfectly, even if I'm going to forget these tomorrow these still are nonsensical to repeat that much during the same session) I get bored and start clicking too fast so I make mistakes out of pure inattention and get even more stupid questions as the result.

kantord 7y ago
Maybe your ideal way of learning or type of platform is different. I have thought of experimenting with the possibility of having different way of presenting the same content. For example, instead of a repetitive software solution, exporting the course material as a printable book.
- qwerty456127 7y ago
  No. I like the repetition way (it works and it provides instant gratification to support the flow state unless repetition rate is t0o intense) and what I see in the current version of LibreLingo is ok. Just don't overdo. Or make repetition intensity configurable.
  - madacol 7y ago
    Maybe what you want are options like AnkiDroid does: https://lh3.googleusercontent.com/QFBCiyA_6WBKfr8Xoo1MqOrQw18t6KE7POlKV3BfA7fLETjnYYu51y2SMAliHTpd6A=w1860-h928 https://lh3.googleusercontent.com/QFBCiyA_6WBKfr8Xoo1MqOrQw1... At the bottom there are 3 buttons that sort of means this: - repeat card more frequently - repeat card - repeat card less frequently Android app: https://play.google.com/store/apps/details?id=com.ichi2.anki https://play.google.com/store/apps/details?id=com.ichi2.anki
yorwba 7y ago
One way to distinguish mistakes of inattention from "actual" mistakes is to leave marking the mistakes to the user. So if you get inattentive, you simply forget to mark yourself wrong. My personal system works that way; though not for any particularly clever reason. I just have no idea how to automate the checks. I don't recommend anyone to try it unless you want to spend a few hours setting it up, but there's a screenshot in the README: https://github.com/Yorwba/alphabet-soup https://github.com/Yorwba/alphabet-soup

petargyurov 7y ago

As someone who is learning Spanish at the moment, here is some feedback. - I think a single click is sufficient to choose the correct answer and submit it; alternatively let me double click to choose and submit. - I don't think a "Continue" step is necessary - Without selecting an answer and pressing Back on the browser, the UI freaks out for a moment (Firefox 72.0.2) - I love cheatsheets, so having a section for these would be awesome (finding good, consistent cheatsheets on the web is hard) It is understandably a little sparse at the moment. Take a look at an app called Memrise - I think the style of its exercises would fit well here. I particularly like the "Fast Review" exercise which lets me do a speed review of all the words and phrases I have learned so far. As others have mentioned already, it would be nice to be able to mark words you already know and stop them from showing. Additionally, being able to mark words you find difficult could appear more frequently. I will keep an eye on this project; good work so far!

melling 7y ago
I built a cheatsheet for my iOS language app for Spanish: http://h4labs.org/100-words-app-spanish-pdf/ http://h4labs.org/100-words-app-spanish-pdf/ I’ve also created some Spanish Word Search PDF’s. Most are on Github: https://github.com/melling/Sopa_de_Letras https://github.com/melling/Sopa_de_Letras I was going to create a little book of 20 Spanish Word Search puzzles. What’s your idea for Spanish Language cheatsheets? I’ve tossed around a few ideas myself. Verbs, ser/estar, por/para, saber/conocer
- petargyurov 7y ago
  I find grammar cheatsheets quite useful, for example: https://i.pinimg.com/736x/0c/b6/81/0cb681ce3173c647a39b12db88344207.jpg https://i.pinimg.com/736x/0c/b6/81/0cb681ce3173c647a39b12db8... Anything that explains grammatical rules and gives me a "template" of sorts is great because it doesn't give me the answer straight away but it engages my brain to adapt the word I want to the given rule. Topics like tenses, conjugation, pronouns are what I like having cheatsheets for.
_def 7y ago
I agree with your points, controlling the app entirely with the keyboard seems pretty nice though

melling 7y ago

I’ve been working on my own iOS language apps for quite some time: http://appstore.com/h4labs http://appstore.com/h4labs My original idea was to do lots of little games in one app to keep it interesting: Hangman, Word Search, 4 Pics 1 Word, ... Recently, I’ve been breaking out the games into smaller apps. I think that’s better. Take your data and try to make some fun little games. Also, I think noun gender is important. Should learn it at the same time as the noun. It can change the meaning of the word. Finally, is anyone doing verbs? I have a simple Spanish verbs app. There are usually many rules to help make it easier. I was going to open source the rules that I have but there doesn’t seem to be much interest: https://github.com/melling/Spanish_Verbs https://github.com/melling/Spanish_Verbs Need to extract the data from my site: http://www.h4labs.com/lang/es/index.html http://www.h4labs.com/lang/es/index.html

bayesian_horse 7y ago

Good idea. I personally love Duolingo. But I fear that the course data, which is mostly user contributed, will eventually disappear if the company goes under...

kantord 7y ago
same, that's my biggest fear
Vivtek 7y ago
There's a lot of politics involved in that course data, too. That an open-source approach exists is a very good thing.
zelly 7y ago
Some hero should scrape it.
- lambertsimnel 7y ago
  Would that be legal? I wonder who holds the copyright to user-contributed data in DuoLingo (assuming they're copyrightable). If some of the user-generated content isn't copyrightable, or was contributed by users willing and able to share it with a FOSS project, could only that data be scraped, or would it be too difficult to identify?
- bayesian_horse 7y ago
  One way is to get Premium and download the course. I haven't looked at it, but I assume they haven't bothered to do any copy protection on those data packages. Not sure if they contain account-bound watermarks of any kind.

RobAley 7y ago

One small piece of feedback : My (UK) keyboard doesn't have accented characters on it, and I don't know how to type them (well, I do, but many/most users won't), so typing "leon" instead of "león" for Lion wouldn't let me continue as there was a spelling mistake. Ideally, have an on-screen keyboard for accented characters, or instructions on how to type them, or less ideally allow spelling mistakes on such characters.

kantord 7y ago
thanks! If you have a GitHub account, feel free to create an issue for that! https://github.com/kantord/LibreLingo/issues/new https://github.com/kantord/LibreLingo/issues/new

cerberusss 7y ago

I'm not sure if it's a joke or not, but that web page its title has a typo. It says "langauge" instead of "language".

LaurentS 7y ago

Hi kantord, congratulations on the site! I tried something similar about 10 years ago, called wikibabel. I then joined with another site called wikiotics.org. We were very focused on libre everything. In the end, we gave up after several years, but our content is still up and we've been looking for a new home for it. Please take a look and see if any of it would be useful for you (we had built a bunch of cards very similar to your demo lesson). We had stuff in a few dozen languages, including audio recordings (mostly of low quality). I've moved on, but I still love the goal, and would be happy to share some feedback on the experience if you want (contact on my profile page). In short: - focus on what the user wants, not the tech. - is English->Spanish the best language pair to start with? You're directly competing with the giants like Duolingo. I tried to learn Burmese a few years back, I would have loved to find something like this for that niche. Good luck!

rjeli 7y ago
Cantonese also has a ton of speakers across the world, and the best I could find a few years back was the FSI tapes
- fartcannon 7y ago
  I would also love Cantonese in a Duolingo style, please!
  - computerfriend 7y ago
    Can't agree more. It is amazing how few resources there are for learning this amazing language.
Vivtek 7y ago
That's a pretty neat site (Wikiotics, I mean - wikibabel appears to go to multiple places).

skinnyasianboi 7y ago

Good work and the roadmap looks pretty promising. I hope we will see native mobile apps. Would be a cool case to play around with SwiftUI and Jetpack Compose.

dwrodri 7y ago

I am far from the first to say this, but please do push hard for languages that aren't commonplace (or well-developed) on other platforms. My language of choice would be Farsi.

cptwunderlich 7y ago

I also wanted to congratualte you for your effors. I'm a daily and long time user of Duolingo and thought of libre alternatives for a while. I'd agree with some sentiments in this comment section, that you might want to find a niche, as competing with Duolingo or Babbel would be difficult. Duolingo doesn't too well with "smaller" languages and different scripts. Are you using a TTS engine for the voice (I assume)? I was looking for TTS for a smaller language I'm studying, but I couldn't find anything. I hope that something comes out of the Mozilla's Common Voice project.

johnkpaul 7y ago
I totally agree with this. I'm an avid Duolingo user with an over 2000 day streak but I would happily use another tool/game/platform if I could learn Malayalam. My family is Malayalee and I know the entire language in my brain because I can understand them speaking to me,but I respond in English. I would pay a huge amount of money for an English<->Malayalam instructional book and I'd pay a huge amount of money for a Duolingo-similar Malayalam learning experience. Niche and rare languages have little to no representation in the popular apps.
kantord 7y ago
I still think it makes more sense to implement languages that more people want to learn. However, the long term goal is making it easy for the community to build courses, so once the project is mature, it should be possible to include a way larger number of languages. I am also thinking of things like conlang enthusiasts being able to create courses for their own conlangs.
- mumblemumble 7y ago
  I just wanted to chime in and say I think there's another good reason to focus on the more widely-spoken languages first: It's not just that there are more people who want to learn them, it's also that there are more people who can help contribute to the materials.
  - kantord 7y ago

mumblemumble 7y ago

Completely tangential: The Rosetta Stone / Duolingo style of language learning has never really worked for me; I generally have more success and an easier time staying engaged with the "comprehensible input" approach. Are there any open platforms based on this method out there? Maybe something along the lines of LingQ, though perhaps without all the attempts at gamification.

bluGill 7y ago
I think the best approach is a few weeks with a duolingo style learning just so you get SOME grammar and a FEW words. Once you can understanding things then input is best, but if you don't understand input isn't going to do much good. If you can find English [any language you know] subtitles all input is comprehensible and it is best - but also the hardest and most frustrating. If someone could just create a list of of youtube videos with subtitles in my language it would be huge. Doesn't matter if it is a blacksmith explaining how to shape metal, or a knitter showing some pattern, I just need input.
- mumblemumble 7y ago
  TBH, I don't even personally like Duolingo-style programs for getting down the base vocabulary, because, while it's a generally more pleasant experience than other options, I find the approach to be relatively slow and inefficient. Here's a video (in French) that I found reflects my own experience quite well: https://www.youtube.com/watch?v=-KQ1qRJ2wQc https://www.youtube.com/watch?v=-KQ1qRJ2wQc In a nutshell, I've got two complaints. First, while flashcarding is wicked efficient (I'd say essential) for review, it's not a particularly good tool for learning things in the first place. And an overly flashcard-oriented approach has real problems with a sort of streetlight effect: You tend to get stuck on what's easiest to illustrate with pictures, which often isn't what's most useful for communicating with others. edit: Also, for what it's worth, the existing research seems to indicate that, while subtitles in our target language can be hugely beneficial, subtitles in a language you already speak actually inhibit language acquisition: https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0007785 https://journals.plos.org/plosone/article?id=10.1371/journal... (There was another study with Brazilian English language learners that I found more compelling, but I'm unable to find it right now.)

songzme 7y ago

When my wife and I first met, we tried to build a language learning app (don't sign up, we stopped working on it) - https://llip.io/landing/ https://llip.io/landing/ To get users, we started a meetup group to teach people Korean. https://www.meetup.com/San-Jose-Korean-Language-and-Culture-Meetup/ https://www.meetup.com/San-Jose-Korean-Language-and-Culture-... We stopped working on the app, but continued with the meetup group (3 years now and almost 1000 members). Our conclusion is that the most effective way to learn a new language is a simple commitment to showing up regularly to language events. Meetup.com or your local public library will usually have regular (and free) events. Once you have that commitment, picking a great app will help expedite your learning. Excited to see an open source language learning app!

generalpass 7y ago
What if you live in a region where you can't people speaking or learning the language you want to learn?
- songzme 7y ago
  That's what I originally thought too (couldn't find an existing group) but shamelessly I started one with my elementary grade Korean. Students much better than me came and I was able to ask them for help. Have you thought about starting a group at a local library or coffeeshop?
- elliekelly 7y ago
  I've had a lot of success using a Discord server to learn French. The big gaps in my self-learning knowledge were pronunciation and conversation. (It's one thing to be able to read/write the language at your own pace but different to be able to "think" and react in that language in conversation.) Discord chat rooms are great and thanks to the voice channels I've almost got the 'r' down. Most of the servers even have user "flair" so you can let others know if you'd like to be corrected on your mistakes or if you prefer others let them slide.
tasogare 7y ago
Great landing page. I’m working in the language learning space but lack the design skills.