Hi John, Could you provide a bit more context? From which language are you drawing these experiences? Did you consider filing a phabricator request for the technical component that can be improved (if so, could you link to it)? Could you also provide some links to these discussions that are causing the internal fighting you refer to?
I'd be curious to understand better what you're talking about before taking a position. Thanks! Best, Lodewijk 2017-05-02 17:20 GMT+02:00 John Erling Blad <jeb...@gmail.com>: > Yes, I wonder if the extension for content translation should be turned > off. Not because it is really bad, but because it allows creating > translations that isn't quite good enough, and those translations creates > fierce internal fighting between contributors. > > Some people use CT, and makes fairly good translations. Some are even > excellent, especially some of those based on machine translations through > the Apertium engine. Some are done manually and are usually fairly good, > but those done with the Yandex engine are usually very poor. Sometimes it > seems like the Yandex engine produce so many weird constructs that the > translators simply gives up, but sometimes it also seems like the most > common errors simply passes through. I guess people simply gets used to see > those errors and does not view them as "errors" anymore. > > Brute force solution; turn the ContentTranslation off. Really stupid > solution. The next solution; turn the Yandex engine off. That would solve a > part of the problem. Kind of lousy solution though. > > What about adding a language model that warns when the language constructs > gets to weird? It is like a "test" for the translation. The CT is used for > creating a translation, but the language model is used for verifying if the > translation is good enough. If it does not validate against the language > model it should simply not be published to the main name space. It will > still be possible to create a draft, but then the user is completely aware > that the translation isn't good enough. > > Such a language model should be available as a test for any article, as it > can be used as a quality measure for the article. It is really a quantity > measure for the well-spokenness of the article, but that isn't quite so > intuitive. > > The measure could simply be to color code the language constructs after how > common they are, with background color for common constructs in white and > really awful constructs in yellow. > > It could also use hints from other measurements, like readability, > confusion and perplexity. Perhaps even such things as punctuation and > markup. > > I believe users will get the idea pretty fast; only publish texts that are > "white". It is a bit like tests for developers; they don't publish code > that goes "red". > _______________________________________________ > Wikimedia-l mailing list, guidelines at: https://meta.wikimedia.org/ > wiki/Mailing_lists/Guidelines and https://meta.wikimedia.org/ > wiki/Wikimedia-l > New messages to: Wikimedia-l@lists.wikimedia.org > Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l, > <mailto:wikimedia-l-requ...@lists.wikimedia.org?subject=unsubscribe> _______________________________________________ Wikimedia-l mailing list, guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines and https://meta.wikimedia.org/wiki/Wikimedia-l New messages to: Wikimedia-l@lists.wikimedia.org Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l, <mailto:wikimedia-l-requ...@lists.wikimedia.org?subject=unsubscribe>
