Script Testing

I would like to start a running list of which scripts work on LingQ and which don’t. I think this would help everyone. Also, if you discover any workarounds, please post them here. For example, if you know a way of transcribing that works.

If you would like to test a script, but don’t know how here are some tips. You don’t need to know the language, but you do need to know where words begin and end. Find a sample text of the script. Make sure that it is at least a paragraph so you get many letters/words. Be certain that it is actually the correct script (or one of, if there are different scripts used). Copy the text (if it is in a different format that forces you to drag a box instead of copying normally then it probably won’t work). Go to LingQ and select a language that you don’t plan on studying from the dropdown at the top left (I don’t believe it makes any difference whatsoever which you pick). Now, hover your cursor over Lessons and select Import. Follow directions. Once it’s saved, open the lesson and try to create lingqs. Pay attention to whether the system breaks the words into lingQable words properly. Edit: Also check to see if words beginning with the same characters are treated as unique LingQs. Report any problems/successes here.

My reports so far.

Hindi is a no-go (and by extension, Marathi and Sanskrit)
Thai does not work (Thanks IYH)

I think the script for Urdu should be tested and if it works maybe the admins will consider adding Hindi-Urdu since it won in the fb poll…but I’m hesitant to test it myself since I don’t know the script and there are many that are similar enough that I might make a mistake.

Hindi+Urdu wouldn’t be good to add as a single language. While they are virtually indistinguishable in their colloquial spoken forms, they are quite different when they are written beyond this level, not to forget the different scripts. Just a note I thought I’d add.

I know they have differences, but they can be written in each other’s scripts without any hitches. I am still holding out hope that Hindi finds it’s way onto LingQ in some form…

Yeah, while that may be possible, it’s a pretty far out way of doing things. Imagine, I have a text in Hindi and I’d like to add it. I have to learn Devanagari well enough then the rules for spelling in the Arabic script, then type it up each time… I wouldn’t even put it into transliteration.

Apart from that, I’m still hoping it gets here soon too. :slight_smile:

I suppose it may be a stretch. For me, it would be no barrier since I know Devanagari and want to learn the Urdu script :stuck_out_tongue:
But more seriously, there would be plenty of “Hindi-Urdu” text already written in the Urdu script available online. I’m not taking a leap calling it Hindi-Urdu…

Anyway, anything would be better than what I tried to do earlier…transcription for Devanagari was so much trial and error and I never got to a reasonable transliteration method. If only one of the standard ways of transliterating worked on LingQ then it would be functional. If words containing these letters: ḍ, ṛ, d, ḍh, ṛh, dh, ṭ, t, ṭh, th n, ñ , ṇ all worked then it would not be a problem for me at all. The most standard ways are to use the dot beneath the letter or to make letters capital. The capitals don’t work since the system does not count them as separate letters. The dots don’t all function properly on this website.

Yeah, I’ve worked a little with older Indian language and I think that the ‘dot-under’ method would fit best to it. Still, not a great solution at the end of the day. What happens if I copy and paste an article that isn’t in that transliteration method? Nightmare… lol

I’ve got a question for Alex or Mark. Do the languages get encoded in unicode?

This blows my mind… It appears that a Punjabi script (Gurmukhi), which is similar in structure to Devanagri (in that they both have horizontal lines along the top that is only broken between words and certain letters) appears to work perfectly.

Whaaaaaaat? haha

@Imyirtseshem - Yep, the languages/texts are in unicode

I don’t know what to say…devanagari also passes the test…I am looking at a lengthy bit of text and can detect no problems of any kind…everything lingqs just fine…Can someone else investigate? Perhaps there were changes to the site? Or maybe the samples I used before didn’t work or the new text is in a different form. I will look into this more myself.

Thanks Alex. I was just hoping that they aren’t being re-encoded to some local variant. That always makes things that bit more frustrating. Converting to unicode, every text which comes in, is the trick.

Oh…wait…I think I jumped the gun. The lingQs don’t appear to be working…

:frowning: Oh well.

I think I’m just going to ignore the problems for now and continue learning. Ignorance will be bliss. haha

That and I’m tooooo tired to worry about it any more today.

When I lingQed हिन्दी (hindi) and then tried to link है (is) it changed my previous lingQ. So I guess it is treating words that start with the same letter as the same word…which means any others doing testing should check for that as well.