Our data is released under various Creative Commons licenses.More information
If you love this content, please consider a donation.
Wall (6,960 threads)
Tips
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.
marafon
4 days ago
CK
4 days ago
sharptoothed
9 days ago
Cangarejo
9 days ago
Cangarejo
13 days ago
Thanuir
13 days ago
ondo
13 days ago
ddnktr
14 days ago
ondo
14 days ago
AlanF_US
17 days ago
Tatoeba beta testing
Do you get to play with the real data set, or only a 'safe' copy?
If you're talking about the "blue" version of tatoeba, it's a separate database with different password , so changing one does not change the other.
Can someone confirm this problem I'm having? I created a few tags yesterday:
- cmt: alternative vocabulary,
- cmt: alternative grammatical number, and
- cmt: alternative gender
by adding them to
* http://tatoeba.org/eng/sentences/show/478530, and
* http://tatoeba.org/eng/sentences/show/478533
They appear at the bottom of
* http://tatoeba.org/eng/tags/view_all
but the pages to which the tags link look like the tag doesn't exist, e.g.
* http://tatoeba.org/eng/tags/sho...rnative_gender
These are the only tags with a colon, so I suspect that's the problem. I tried escaping the colon in the URL,
* http://tatoeba.org/eng/tags/sho...rnative_gender
but that doesn't work. Is this a known problem? Should I file a bug report somewhere?
Yes indeed. The colon is the problem. Thanks for reporting. I created a ticket for it.
Thanks! Looking forward to the update, by the way.
ow to use totoeba
by asking here :)
welcome!
(it's tAtoeba ^^)
If you're looking for example illustrating a word you can simply use the top menu bar in green, for example you want sentence using "example" in English sentences
http://tatoeba.org/eng/sentence...&query=example
If you see a translation is missing in your native language, you simply add translation by clicking on the "translate button" http://flags.tatoeba.org/img/translate.png
If you want to add a really new sentences you can do this by going on the contribute page
http://tatoeba.org/eng/contribute
If you have a specific need, ask, maybe it's already possible in tatoeba, or at least we can think about making it possible :)
Tell us if you have questions
Je pense que les phrases qui viennent d'être créées et ne sont pas encore dédupliquées ne devraient pas être disponibles dans les résultats de recherche, parce qu'alors on exacerbe le problème en créant de nouvelles traductions de copies.
More on tags...
I trimmed the deleted tags from my list and added the number of sentences tagged (according to the latest tags.csv file).
There are still a few that need to be deleted and I've listed these at
http://martin.swift.is/tatoeba/tags.html#delete
Then there's the bunch that needs to be renamed. Can I just send someone a tab-separated file with the old and new names? Most are simple transliterations, but some will need to be discussed.
For example: there are several "check" tags. I suggest we merge "to check", "@Needs Native Check", "Needs Native Check" and "@grammar check" under "@check", "@native check" or something similar ("to be reviewed" as well?).
Then "delete" should be merged with "@delete". I take it this would be considerably simpler with direct database access than through the interface (though I guess one could cURL one's way through that).
[not needed anymore- removed by CK]
Sorry, I failed to mention that I didn't really clean my parsing code up that well; didn't bother with false positives (I'll look into passing grep a more nuanced pattern next time around) and the script doesn't handle tags with parentheses.
I fixed the "British" tag by hand in case you want to resort your page, CK.
PG-13 is a normal tag.
be-1959acad shows that this is applicable only in the Academic variery of Belarusian; it's a normal language tag
Leopolis is a latin duplicate of Lviv ;o
Urdu is to be deleted, since we have a flag now
IMHO spoken to male should be replaced with ‘said to male’
Thanks, Demetrius!
I've moved be-1959acad to the "Language" section, Leopolis to "redundant" and Urdu to "depopulate".
Regarding PG-13, I'm just thinking that a more culture independent tag would be more useful.
Since there is only one "spoken to male" tagged sentence and an existing "said to male" tag, I'll just re-tag that sentence and move the tag to "empty".
By the way, about the XXX tag...
How it should be used? Consider the following groups:
a) sentence describes a sexual intercourse in rude words,
b) sentence uses the rude words with an indirect meaning, to describe something other,
c) sentence describes a sexual intercourse with euphemisms
In which cases XXX tag should be used?
I would say a) and c) and possibly b) (but depending on circumstances).
We'll probably need to fine tune that sort of thing later. It's not really a priority at the moment as the tags don't actually do much yet.
I don't think these should should be tagged under a single label. I think a) and b) could be tagged with something like "rude" or "obscene" and a) and c) with something like "sex" or "pornographic". c) should furthermore be tagged with "euphemism".
I this is actually a great time to think about how we're using and how we'd like to use the tags, seeing how it's not a priority and we can play around with them.
Should we keep proverbs offending other nations in the database? And how should these be tagged?
We already have some Ukrainian proverbs about Russians in the database (sth like “You can ward off the devil crossing yourself, but you can’t ward off a Moskal”). ^^
You could tag them as "stereotype", "prejudice" or even "bigotry".
I'm all for freedom of expression, but I think that clear lines are more useful for a project such as this (oh, and trying to steer around hypocritical stances would be nice). Seeing how the aim of this project is to gather sentences and link them to translations -- not to disseminate facts -- and the web is open to anyone wanting to espouse their hate, I'm not overly concerned about losing offensive content.
I do, however, prefer the idea of simply filtering out the filth one doesn't like.
There are already quite a few labelled as 'Lie'. Filters would be nice (filtering out XXX sentences should be a priority if we want to be 'school friendly'), but I think that sort of content is still needed (in moderation) to give a full coverage of language usage.
> full coverage of language usage.
Talking about which, I'm reminded of a Japanese speaker who was certain that a certain body part was referred to as "pussy cat". Nothing I could say would persuade him that the 'cat' wasn't needed; he just said "I use it all the time with my girlfriend so it must be right."
Feature request - download all in current web.
Suppose you have sentence A, B, C, D, E, and F like below:
....................
.A -+- B --- C ---E.
....|...............
....+- D --- F......
....................
Then I'd like to be able to download all of those sentences as a list regardless of what sentence I'm looking at at the moment.
maybe in a near future but not now looking to the way they're stored, and when it will be possible to do so, then you will be able to see A B C D E when looking to whichever sentence, it will not be restricted to indirect translation anymore.
When that time comes, it would be useful to be able to increase and decrease the depth from which sentences are shown. I've sometimes been translating a sentence and known that it is a translation of a third one, a couple of nodes away. Since it isn't shown, I've had to do a bit of gymnastics to link those. It would be great to be able to show progressively more distant sentences until one finds the one one's looking for.
Also, it would be great to be able to search for sentences in one particular language that are only indirectly translated into another particular language.
Hi,I've just found Tatoeba and would like to contribute.
I could do it with the Esperanto version. I am a teacher of Esperanto at language schools and teached it for two years at the University of Santiago de Compostela as well.
Also I could help with the Galician version, if you want to make a separate section for Galician. Galician is a variety or Portuguese and there are two ways of writing it: with Spanish ortography and with Portuguese ortography (commonly with the other main varieties of Portuguese: Brazilian and Portuguese from Portugal). I have a degree in Galician Philology and teaches Galician for the Autonomous Government of Galicia and at language schools.
I'm a translator and interpreter, too. I work with both varieties of Galician, Spanish, standard Portuguese of Portugal, Romanian and occasionally Esperanto.
welcome aboard :)
Welcome to Tatoeba, Shazzan :)
If you were talking about translating the website into Esperanto, then you will need to register on a website called Launchpad and then translate from this page:
https://translations.launchpad..../eo/+translate
If you were talking about translating the sentences in the corpus, then I suppose you already figured out.
Galician is not yet an officially supported language, but you can simply follow the instructions in the FAQ to have it added: http://tatoeba.org/eng/faq (cf. "I'd like to request a new language. What do I have to do?")
In general, if you have trouble understanding how Tatoeba works don't hesitate to ask.
I often switch between several languages, and it's always somewhat painful having to locate them in the total list of languages. It happens regularly that I inadvertently select the wrong language. Something like a list of favourite or least recently used languages for quicker selection would be very helpful.
I agree that the language list is becoming a bit too crowded but I can't say for sure when we can improve this... We will though, don't worry ^^
But in the meantime, if you want to avoid switching too often between the mouse and the keyboard, you can do the following after you have entered your translation:
- Press tab to set the focus on the languages list
- Enter the first letter(s) of the language of you want to select. For instance "ger" for German (or you can simply enter "g" and then navigate press down until you reach German).
- Press tab to set the focus on the "submit translation" button.
- Press enter to submit the translation.
This way, at least, you don't need to switch to the mouse to change the language.
In the future, we'll try to set up an interface where you won't need to use the mouse at all. It will be less cumbersome.
I made a mock-up version of the Tatoeba search interface moving the search bar in front of the language selection menus and adding the accesskey "4".
That would simplify the process into:
1. [accesskey-combination]-4
2. Type search query
3a. If language settings needn't be set, hit <Enter>
3b. If language settings need to be set, tab and select before tabbing to the submit button.
All with the keyboard. :-)
Any chance something like this could find its way into the interface soon?
Mock up: http://martin.swift.is/tatoeba/home.html
Accesskey combinations for a number of user agents: http://en.wikipedia.org/wiki/Ac...erent_Browsers
Right, right, sysko had told me about this. Thanks for reminding me!
And yes it can find its way into the interface soon :) Next update.
Two quick comments on the search bar update: Firstly, could you disable the javascript that sets the focus to the search field? It currently disables scrolling around the page with the arrow, page up and down, and home and end keys (until one has removed focus). Say one has scrolled down a bit using the mouse but not yet removed the focus from the search field; then any typing on the keyboard will send the viewframe to the top.
In certain cases where entering text into a form is the only purpose to going to that particular page, setting the focus to that form does make sense. So on the "Contribute" page it might make sense to set the focus to #SentenceText.
Secondly, could you add an accesskey ("4" seems to be as standard as it gets for search fields) for the search box? It's fairly unobtrusive as it has virtually no side effects.
Great! Thanks. :-)
Ok, I'm looking forward to the day this gets some improvements. Until then I can live with the current solution, just wanted to get my thoughts across. Thanks.
Do you mean the drop down lists of languages? Most user agents let you jump to the item by pressing the first few characters.
For example, for English you can select the list (by clicking on it with your pointer, tabbing to it or however you select things) and then hit the "e" key twice (the first gets you to Egyptian Arabic and the second key press selects the entry below; English.
You can also type the first few characters. In the case of Italian, you'd be quicker to type "it" than hit "i" four times.
Hope that helped.
I'm aware of that quick navigation feature with the keyboard, but in my eyes that's still too cumbersome. Too many clicks, too many changes from mouse to keyboard and back to mouse. A handful of configurable language buttons somwhere near or some often used entries at the top of the drop down list would be much better.