Live data from Hacker News

Ask HN: What service do you use for human translations of your website?

news.ycombinator.com

11–20 of 51 posts

Re: Ask HN: What service do you use for human translations of your website?

#12

I've used, with plenty of success (ending around 5 years ago), both welocalize.com and lionbridge. We sent them .JS files, XML files, PDFs, and Word Documents. Some things to look for: Across multiple transactions with these [edit: localization] companies, they end up building a database of phrases, of which some companies might try to claim ownership over. Watch out for that when negotiating. You're also most likely…

>You're also most likely going to pay per word, so if you can figure out a good way to not send them sentences you've already had translated you can save some money.

That's why they build those databases of phrases you were referring to, since they apply discounts on matches against those DBs and on repetitions in your document.

Instead, ask explicitly for a Translation Memory file at the end of the job, so that you can retain your personal piece of database of phrases that was produced during the localization of your file (it's your right to have it and serious companies will hand it without fuss).

Re: Ask HN: What service do you use for human translations of your website?

#15
Contracted local university CS / Math students from China. Done on an exchange basis where I helped with English and tech issues. They helped proofread machine translations. Made some excellent friends. And now I have an invitation to visit Henan Province as well ;)

What I am really interested in investigating is what is the minimal subset of words or concepts that makes a website or service useful. Is it 10,000? Or closer to 500. Because if is the latter, localization by standardizing the interface around these usability 'primitives' common to every culture would make the translation process very fast.

Essentially localization becomes a kind of "growth hack" itself. Having English, Spanish, Russian, Mandarin, Hindi / Urdu, etc. Covers a vast swath of humanity. But the marginal addition of say, Polish or Dutch, can result in an unexpected boost!

Re: Ask HN: What service do you use for human translations of your website?

#16
Berlin based Applanga has helped us use human translations across 35 countries in 6 languages for a big multi-platform (ios, android, Amazon Alexa, web) app. We are based in Philly, but their service could not be any better. A professional team of (funny and cool) folks, excellent sdks, with human translators, and verification protocols. Also, their handholding through both tech onboarding, and the human translating part of the process has been great.

www.applanga.com

Re: Ask HN: What service do you use for human translations of your website?

#17
post #9

DeepL ( https://www.deepl.com/ ) is a pretty good translator. It's not perfect, but it's much better than google translator.

I don't see pricing or an API from DeepL that could be classified as a 'service'. Am I missing it?

Re: Ask HN: What service do you use for human translations of your website?

#20

I've used, with plenty of success (ending around 5 years ago), both welocalize.com and lionbridge. We sent them .JS files, XML files, PDFs, and Word Documents. Some things to look for: Across multiple transactions with these [edit: localization] companies, they end up building a database of phrases, of which some companies might try to claim ownership over. Watch out for that when negotiating. You're also most likely…

>You're also most likely going to pay per word, so if you can figure out a good way to not send them sentences you've already had translated you can save some money. That's why they build those databases of phrases you were referring to, since they apply discounts on matches against those DBs and on repetitions in your document. Instead, ask explicitly for a Translation Memory file at the end of the job, so that you…

"Translation Memory". Thank you I could not for the life of me remember that. 5 years was so long ago !
Post reply on HN