Wall (6,213 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.
7 hours ago
10 hours ago
13 hours ago
13 hours ago
22 hours ago
How are the tags used by WWWJDIC handled?
For example the following sentence should have a sense 1 tag on 誰も, but I don't see it in the edit sentence options.
If by "tags" you mean the indices, or "B line", then they are accessible from a special page. I'll send you the link in a private message, you can also ask Jim for more details.
If you are referring to the feature from the old version, that was using the indices to link words from the Japanese sentences to WWWJDIC, then I haven't had time to re-implement it yet.
Yes, I meant the "B line".
We could do with an option to show sentences in language X that do not have translations in language Y. (E.g. Japanese sentences with no English translation).
Yep we plan to do so, in fact we plan to add an "advance research panel" with this kind of research option
As an stop-gap measure how about creating a public list of sentences that need Japanese translating to English? That would let me get started contributing.
yep sure you can, anyway even with the capability to search sentence untraslated in language X, it's always better to directly have the possibility to see which sentences is reclaim by others,
(by the way some of our users used to post comment like "if someone could translate it in japanese it would be great", so the need is already there)
> yep sure you can.
I was sort of hinting that someone with admin access could produce such a list for me. ;-)
Someone with admin access has heard you and here is the list:
(And I notice that some sentences in this list should be deleted)
As how translations relation works in Tatoeba it's not so obvious and I've explain it a lot of times, i post it here, waiting a better place
some of you have been puzzled by sentences which are disapearing etc...
that's come from the fact that a sentence has "translation" and "translation of translation"
and you always have a "main sentence" (at the top and bigger)
so to be a bit theoric
AA is a sentence, i add a translatio "BB"
now we have AA <----> BB
if i translate BB as CC
now we have
so BB is a translation of AA
and CC a translation of translation AA (it will displayer a bit more gray when AA is the main sentence)
so now if translate CC by DD
we will have
and when clicking on sentence AA , DD will not appear, but when clicking on CC or DD it will
in fact you know that when you translate, even how good you are in translation, due to language/cultural limitation/difference, we can't have a 100% exact meaning (for example in english you have only "they" but in france you can translate it by "elles" (female) or "ils" (male)
Ils <--> they <---> elles <---> etc.
as you can see "ils" and "elles" are both correct translation of "they" but they have not the same meaning, that's why we need when "ils" is the main sentence, to say "hey elle is an undirect translation meaning is likely to be a bit different but still enough near to help you"
so now you can imagine with complex sentences, translation of translationf of translation etc.. are really likely to be totally different, so that's why we don't display them
that's why it's important to only translate sentences in the language you know, not translate it because it has a translation you know the meaning of
anyway in the next update of tatoeba we will review a bit the way it's displayed in order to make it more obvious
so that's globally why somes translations dissapear depending of which is the main sentences and why it's important to have the right main sentence
(I Hope i was not too long)
feel free to ask questions, I know I'm not really good for explanating :$
Hi, I'm new here, so this is tentative, but I have 2 gripes:
1. Some entries are just phrases, not "sentences", where by
sentence I just mean complete thought, informal though it
may be. It seems to me that we should sell whole loaves.
2. Each sentence should make sense on its own, shouldn't it?
Sorry if these were already discussed and decided; please
point me to such information. Perhaps I'm just too rigid.
The idea, it seems to me, is that if one goes to Tatoeba
or WWWJDIC to get a sample sentence, then one should GET a
comprehensible complete sample sentence.
Yes, ideally each sentence should make sense on its own, even though everyone can pretty much submit anything...
But entries that are not "sentences" are likely to be deleted in the future. They will remain here for now because no one can delete them (except me) and we haven't really decided yet on what should be deleted and what shouldn't.
Hmm, how about making them into sentences? I mean, the phrases are here because they are significant, right? If the principle is agreed upon that entries should be context-free full sentences, then I and others would feel comfortable changing them into sentences (it would seem pretty easy in most cases).
That works too.
Please add support for deleting and editing comments. I just keep on posting translations as comments by mistake ... I've seen others do this too.
It's in our todo list.
Like I said here : http://tatoeba.org/eng/wall/index#message_142
"We're currently in a phase where we're basically cleaning up the code, optimizing the system and fixing bugs. And then we have a bunch of small improvements to integrate."
The possibility to delete comments is part of those small improvements. You can expect to have that sometime by the end of the month.
Thats great. Good luck with that!
This site has the potential to be the #1 site for anyone learning a language! I can't wait to see it improved.
Now only full database download is avaliable.
I wonder if there will be tool (some kind of checkbox), that you check some sentences and then download them?
F.ex. I check spanish, english, and japanese and then I download generated CVS file with TAB or semicolon separated sentences.
Yes, someday there will be something like that (that is, more customizable downloads). But it's not in our priorities at the moment (sorry).
We're currently in a phase where we're basically cleaning up the code, optimizing the system and fixing bugs. And then we have a bunch of small improvements to integrate. So I'd say you can't expect to have such a tool before April or May...
But you know, if you have time and some programming knowledge, feel free to code yourself such a tool ;)
And in case you actually do decide to code your own tool and put it online, let us know, so we can add a link to it.
hi, I'm new at this, and I wonder if there's an easier way to do simple edits. In using wwwjdic, i spotted some obvious errors and Trang told me I could fix them, so I did. Just now I tried the "Contribute" link, then 15 random Japanese sentences. But how do I get back to the sentence page to see if others need to be fixed? Page back gave a lot of Resends.
Also, I'd really like to go through them systematically, say a 100 at a time, not random so I wouldn't redo stuff. I'm not talking rocket science - my Japanese isn't good enough for complicated stuff - just things like my last few, "grand mother" -> "grandmother" and "he seems to very happy" -> "he seems to be very happy". Not sure if these are that useful, but I'm a finicky type and would be glad to do them.
I admit it's extremely annoying to do simple edits at the moment and believe me, if I had more time it would not be like that. It will be much easier someday, I cannot tell you when, but I can promise you it will be.
In the meantime, you can of course use WWWJDIC as you figured. But I want also want to point out that if you are using the "serial translators" page, the best way to proceed is to open a new tab whenever you adopt (that is, instead of left-clicking on the "adopt" icon, right-click on it and choose "open in a new tab"). This way you keep the list of sentences in one tab while editing the incorrect sentence in the other tab. And when you're done editing, you can just close the tab.
Also, don't be afraid to adopt as many sentences as possible, and remain their "owner". Of course it's better to understand the translations of the sentences you own, but it's NOT a requirement.
The thing is, by adopting a sentence, you will prevent others from changing them into something that might be incorrect. So you can also adopt sentences even if you're not going to edit them. I'd even encourage you to do that.
Ideally, each sentence should belong to someone. Even more ideally, each sentence should belong to someone who understands it perfectly and will respond quickly if there's a comment on their sentence.
And don't think that you're not helpful, all the corrections you are doing are actually very useful. Translating is not the only way to help in Tatoeba. We want to provide quality content but we know there are still many mistakes, and it's a huge task to correct them.
Thanks! I tried opening a new window and that helps, tho by now I'm in the habit of just left-clicking, so I'll have to unlearn it. I have to disagree in a good way with "extremely annoying" - this site is so addictive! It's fun. On the adoption point, I read a book a while back on the early wikipedia that tried to have people controlling subjects, and it was an enormous drag on the project and slowed everything down. So they opened it up and things took off. That's why I release all my sentences. I could be wrong. But you know, I feel that I may be correcting valid Briticisms, for example, so open may be best.
Yes, I agree that if there is too much control, it slows everything down. It was actually what I experienced as well, back in the old version of Tatoeba where there were moderators and each contribution had to be validated by a moderator. I had written a little bit about that here : http://blog.tatoeba.org/2009/01...on-system.html
Anyway adoption was not intended to increase control, it was more intended to increase the involvment of contributors, as well as their responsibilities. And as stated in that blog post, it's also part of the validation process. The fact that a sentence belongs to someone doesn't mean it's error-free, but at least it indicates the sentence is less likely to have a mistake and you can trust it more.
But imagine: I want to translate English sentences into French, and I see this sentence which sounds strange to me, but I'm not sure if it's actually incorrect or if it's just that still lack of vocabulary. If I post a comment to ask about it, but there's no owner, the comment will just appear for a short time on the homepage and if no one answers quickly enough, it will basically go unseen. But if there's an owner, (s)he will receive a notification email and I will have more of a chance to get an answer to my question.
Also, if someone disagrees with your correction and post a comment on it, then you may learn something that maybe you wouldn't have seen if you weren't the owner of the sentence.
If you are correcting valid Briticisms, it doesn't really matter. If someone wants to have the British version back, instead of reverting your correction, they can just add it as a new sentence.
PS: Glad to know you're addicted :P
hmm, sorry, wasn't thinking - of course you can get many sentences compactly arranged by going directly to WWWJDIC.
Are the moderators planning to add sound support for sentences in the future? I am very curios. I don't think its necesarry, but it would definitely kick things up a notch!
it's planned (like a lots of things :D) but for the moment i can't give you a precise date, but yeah we will also really enjoy providing people the possibility to hear what they read. And we're eager to hear our dear contributors' voice.
that's so awesome! love this site.
Wow sounds great! This site is going to be a great hit! Keep up the good work!
Hi, I just spotted what I believe is an error with the katakana to romanji automatic traslator. チェ translates "CHIE" instead of "CHE". Hope it helps.
I was just wondering. The corpus is under CC:BY licence but when you're downloading the database dump, there is no way to re-associate a sentence with its original contributors.
Anyway, I guess it could be interesting to know who's contributing on which languages for collaborative translation research purposes.
Well, the fact that the corpus is under CC-BY means you have to mention Tatoeba if you are going to reuse it, not that we have to indicate the original contributor in the dump.
It works like this : people provide their work under CC-BY, so the attribution of the work of each user is mentioned in Tatoeba itself, through the logs. But then we're reusing the work of everyone to make something else: the corpus, which we also redistribute under CC-BY.
But now, if you also need to have the username of the original contributor, or any other information that is not provided by default, you can just ask and I'll see what I can do =)