Wall (5,865 threads)
Before asking a question, make sure to read the FAQ.
We aim to maintain a healthy atmosphere for civilized discussions. Please read our rules against bad behavior.
6 hours ago
11 hours ago
14 hours ago
14 hours ago
15 hours ago
15 hours ago
15 hours ago
17 hours ago
17 hours ago
18 hours ago
I think it would be nice to grade sentences. Some kind of pluses and minuses. For what?
There may be more than one version of sentence, more usable than other ones. Choosing of the right one, for person does not familar to the language rules can be tough. This kind of rating should help.
Who should rate? Original language carriers (highest rating - this could be done through user profile), people who knows reading on the other language well, could be done over some kind of qualification procedure.
Sentence rating has been discussed before. It's not a bad idea, but I think there are too many sentences (at least for the top few languages) for it to work well. Take English. If, say, 500 English sentences are rated every day it would take around one whole year before every English sentence was rated.
If there was a rating system I think it would be better if a select subset of the popular languages were singled out for rating. 5,000 is probably a handy number. After a while, low rated sentences could be dropped out of the 'select few' and others selected to replace them.
I'm going to use this space to again push forward the idea of an optimized algorithm that invites users to review sentences/translations that they are most fit to review via a popup sidebar ;-)
I usually use google translate as a dictionary.
needs native check
When the new languages will be added?
When the bugs are fixed?
Dear Allan Simon (sysko),
Your name is not like a French name.
And I have got a Twitter.
Since the last update the 'Tatoeba-style' autocompletion system has disappeared. Not too bothered by that. However now the Firefox autocompletion doesn't work on it either!
Could we have the autocomplete="off" removed?
my bad, I've forgotten to restart the autocompletion server. It should be back now
Why are Japanese sentences doubled every time in grey characters under black ones?
for historical / architecture reason
Some languages, mostly non latin alphabet one, sometimes require a "latin alphabet" version or something to help the occidental learner to read it. So the second grey line is used for this purpose (you can try with chinese or shanghainese).
Before for japanese we used romaji, but it was really innacurate, the software we used was producing bad result etc. so we moved to another one, and then we decided to used katakana (sorry if I used the wrong terms :s, I mean the things in Japanese which can be assimilated to an alphabet) And then Trang decided to switch to a ruby version of this. So now for japanese it's true it's a bit "double". But for architecture reason, we cannot "merge" them for japanese, as I don't want to make a special case for japanese in the code (harder to maintain etc.)
I will update tatoeba soon, in 5/10 minutes, so don't worry if the server is down a little time.
done, search and several random should be a little faster :), and now it takes 5 minutes to reindex all the content against 2 hours before.
so now the search engine will update its index every day instead of every weeks, I've also corrected a bug with Armenian alphabet which was not handle ...
Why is it impossible to select & copy sentences? Is it on purpose?
it is not on purpose, but consider the following,
we need a way to make people go to the page of a given sentence / translation, there's the little button on the left, but when we did it, few people clicked on them, and so a lot of people have missed this page.
So we've decided to make the text of sentences link, so yep it has the drawbacks to make difficult to select inner text. If someone know a easy and standard solution to make the "link" to become simple text when select, I will be glad to integrate it.
For the moment the workaround is to go on the page of the sentences, and on the right you have the historic of the sentences, and there you can select inner text etc.
So as you can see we had to choose between this two things which has both drawbacks.
> If someone know a easy and standard solution to make the "link" to become simple text when select, I will be glad to integrate it.
Well, /could/; not necessarily /should/. This will break standard browser features such as open-in-new-window/tab, copy-link-location, etc. I'm not sure it'd be worth that.
Actually, no; a simple onclick would have the browser follow the link after text selection. This would require a slightly more complicated solution.
I'm out of decent ideas.
... well, except trying to lobby the browser developers to make selecting link text easier.
Oh, now I see whe Google would index Tatoeba sentences. Because they are not simple text, right?
No, the reason is just for the moment tatoeba is not really known, we have few website which reference us. Moreover tatoeba as a lot of pages, and in different languages. So it divides the page rank between all these pages. Moreover the more a page is linked by other website the more chance you have to find it. So as you can imagine, no website links "Hello, I would like to buy an hamburger". Moreover the more "keyword" a page have, the more chance it has to appear on this keyword, but Sentences in tatoeba are 1 sentences. So we don't have this repetition of keyword Wikipedia has.
All these things to explain tatoeba is not "google indexing" friendly yet.
This is a feature of your browser, not Tatoeba.
In Firefox, you can press down the Alt-key while selecting text. This will trigger the broser to "Save-As" after you've selected the text, but you can just cancel that and then copy the link.
In Opera this Just Works(tm).
Oh, it actually works! :)
Yeah, I have this problem too. I've tried Safari and Chrome on Mac OS X.
Thursday WWWJDIC examples update summary
24 records deleted.
6 records added.