An error occurred while saving the commentHernando commented
The strengths of Readlang are somewhere else!!!. As I see it, the tool is not about accuracy of the translation, rather, it is a tool to help you learn word in context. It helps you to "Rapidly translate the words and phrases you don’t know" and it collects those words AND their context, for you to study later.
Actually, for the inline translations I think Google Translate is used, so you can see the point is not getting the most accurate translation. There will always be some words that require more effort and let me tell you, Spanish is my mother tongue and I promise you many native speakers do not know or get right the difference between "como" and "cómo" (básicamente, tienen mala ortografía), but they are normally able to tell the difference because of the context in which the word is written. Besides, dealing with different character sets frequently require getting rid of accent marks. So, your example is actually a very tricky word, and it is very unfair to judge the tool based on it.
In any case, you can always customize the dictionary that appears in Readlang's sidebar, so you can use your preferred dictionary. If you want the most accurate translation, go get the most accurate online dictionary and tell Readland to use it in the sidebar.
But the thing is that if you only care about the accuracy of the translation, you need something different. Readlang is about reading in a foreign language with a tool that:
1) Do not intrude too much on the reading experience ...
2) But still help you quickly find the meaning of unknown words (although not 100% accurate).
3) And let you collect those new words you are learning with their context for you to study later.
Do not blame the tool. Maybe you just need something else!
3 votesAdminSteve Ridout (Creator, Readlang) responded
Thanks. One issue is that Google charge per character that you feed into the language detection API, so if I threw a whole web-page into it it would get expensive. To implement this I would probably look into implementing a method myself using word lists from all the languages.
Currently, I look for some information embedded in HTML meta tags of the web page which indicates the language it’s in. Of course, most sites don’t use this, and it’s not always reliable, so usually it’s up to the user to make sure it’s correct.