Toucan vs Merid, written by the person who built one

I am the author of one of these two extensions, so treat this accordingly. I have put the part that reflects badly on me first, which is the only way a comparison like this is worth reading.

Which one should you install?

It depends on what language you read the web in. English, learning something else, Toucan. Vietnamese, learning English, Merid. That is not a diplomatic non-answer. The two extensions genuinely run in opposite directions.

Toucan takes the page in front of you and swaps a few words into the language you are studying. Merid takes a Vietnamese page and swaps a few Vietnamese words for English ones. Same original idea, mirrored, and each mirror serves a different reader.

The comparison table

That last row is the most honest line in the table. Toucan has a company, a team and revenue. Merid has me.

Where Toucan is straightforwardly better

Language coverage, maturity, and the likelihood of still existing next year. Toucan handles roughly twelve languages. Merid does exactly one pair, Vietnamese to English, and will not be adding a second any time soon.

Toucan has also been through years of releases and now has Babbel behind it. If you want something stable with someone accountable when it breaks, that is a real advantage and not a courtesy.

I am being clear about this because I would rather you install the right thing than install mine and be disappointed. If you are learning Japanese, use Toucan.

Where Merid wins, for this specific reader

On the only thing that decides it, which is whether the extension runs on the pages you actually open. That is not a feature, it is a precondition. An extension that does nothing on VnExpress or Vietnamese Facebook has no features at all, as far as that reader is concerned.

The word lists differ too. Merid ships SAT, C1 and C2, which are the three targets the people around me are aiming at. You can also build your own list if you are studying for something specific.

Then there is the piece I spent longest on: context checking. A dictionary gives you synonyms, but a dictionary synonym is not automatically a valid substitution in a particular sentence. Merid can ask a model whether the word it just inserted actually fits, and correct itself when it does not.

In that screenshot it had used amorous, decided that was wrong for a sentence about two footballers, and swapped in captivate. Both map to the same Vietnamese phrase. Only one of them belongs in that sentence.

How aggressive should the swapping be?

Gentle, at least for the first week. Merid has three intensity levels and the lowest one is deliberately very low. If a page comes back with thirty words changed you have stopped reading it and started decoding it, and whatever you opened the article for is gone.

There are also three display modes: replace the word outright, only highlight it, or place the English beside the Vietnamese. The side-by-side mode is the one to use when you need to be certain you understood something and do not want to guess.

What happens to the words you meet?

They go to a deck, if you save them. Mark one as known and Merid stops swapping it in. Here is my deck at the time I took the screenshot: twenty-five words, fifteen in progress, ten known.

It contains hardy, innovate, copious, amalgamate and confluence. I did not sit down and study a single one of them. They turned up while I was reading things I was going to read anyway, and I kept them.

That is the entire idea, and it is why I abandoned every vocabulary app I installed before this. They all wanted their own slot in my day. This one does not ask for one.