Wall (1 discussion)
Consilios
Ante de poner un question, assecura te de haber legite le FAQ.
Nostre intention es mantener un atmosphere salubre pro discussiones civilisate. Per favor lege nostre regulas contra mal conducta.
marafon
un hora retro
felix63
heri
AmarMecheri
heri
Shishir
heri
AmarMecheri
heri
AmarMecheri
heri
deniko
heri
deniko
heri
frpzzd
heri
araneo
heri

Hello tatoeba community. This project is a great tool for multiple purposes, unquestionably the most common of which being a reliable source for those in need of example sentences for their foreign language studies.
I'm not very much familiar with the project and how things go on, but as my short experience has revealed there are a few points that needs to be made clear. You can consider it partially a suggestion or a question.
Like any other use I guess, I use the search area above to find some sentences which are related to my keywords. Those keywords are sometimes (for some languages) very very tricky. Let's say I search for a verb to find sentences that contains this verb. Yet in the search results there are a lot more than that: some nouns which look like this verb but have nothing to do with it, another verb which looks like my verb when conjugated for a speacial tense, person etc.
At this point, the question is: is it really a good idea to keep a sentence as a simple text in the database? Wouldn't it be better if we create a metadata (more developed than tags) for each sentence. By doing so we can use this database more efficiently and define the relations between languages more clearly. It will also help us to maintain more developed search facilities.
Just to be more clear I'm giving a very simple example. When I add a sentence in German:
"Meine Geschichte ist zu grotesk, um eine Lüge zu sein."
I should add an extra information about the verb for instance: The verb is "ist" and the infinitive form is "sein"
So when a user wants to see a sentence which contains the verb "sein", s/he can find it. Otherwise s/he will search for "sein" and many other irrelevant sentences will come up, like:
"Ich liebe seine Schwester" As one can see; this "sein-e" is not what we are looking for.
Long story short, the badly organized example of mine above is just a little demonstration of how metadata added sentence storage will help users find what they are looking for.
I'm not quite sure if this is the right place for such a suggestion/question, but here it is.

If you search the wall's history you'll get more information on what has been discussed already on the metadata subject.
You can also get info here http://blog.sysko.fr/post/13
Bienvenue sur Tatoeba !

Hoşbulduk.
Is there a search facility for wall posts? You are not expecting me to read every single post in 190+ pages, are you? :))

Actually yes in the future we will have something dedicated to metada and one of this meta field can be about this (the new architecture I'm building will not restrict us on what kind of key=>value metadata we can provide)
In parallel of tatoeba I've already started with a friend to think and code about some software to decompose the sentences into their grammatical strucutre /semantical field of the words
After we can imagine with a mix system of computer-learning/user proofread we can easily increase a lot the rate at which we add this kind of metadata.

Thank you! This is just good news. The news may not be new though, for other users :D as I just found out. As a matter of fact I took sacredceltic's advice and made a google search on tatoeba wall for metadata. I was very surprised to see years ago several other people had been talking about it, using very similar words to mine. Like this one by Zifre: http://tatoeba.org/tur/wall/sho...7#message_4637
Now I can sleep in peace :))

you can search in Google "Tatoeba metadata"
Wall posts are indexed by search engines.

I agree it's a very good idea, but not everybody would be able to "categorize" words. Believe me or not, many people don't know what a verb is...
It would be GREAT to have this resource! It wouldn't be easy to categorize more than 1.5 million sentences, but a very interesting challenge.

@alexmarcelo, It's true that not everyone can do this. But as you said there are a huge number of sentences. This is why such an initiative should be taken. Because without an accurate and clear meta-information, such a huge resource could turn into a useless bunch of sentences. Then what is the difference between tatoeba and a search result I get through a web search engine? (sorry if I sound critical, this is not what I'm trying to do. I just want to stress the necessity of that practice, as you already agreed)
@xekri, I share your stand when you say different languages have different strucutures, therefore differet categorizations. I am not familiar with Lojban, but if you say that such a mechanism would be useful for Lojban too; it means that a common denominator needs to be found. A common solution which can apply to every language.

The Japanese indeces contain exactly the information that you want, but only for Japanese. Also, unless it has been added recently, there is no general UI for updating the Japanese indeces. I would dearly love this to become a general feature of Tatoeba. There are definitely a lot of issues that have to be thought out, but IMHO this is a must have feature.

I think there should be a way to see which user attached a certain tag to a sentence. For example, it would be useful in order to know whether somebody put an "OK" tag to a sentence which isn't in his native language, or if he put it to a sentence he wrote himself.

When you hover the tag, you see a usernumber. One of my scripts translates this number to a username. I think it's part of [Tatoeba suggest tags of siblings](http://userscripts.org/scripts/show/104647) (im not on my computer now, so i cant check).
Im going to work with sysko on integrating my userscripts into the new version this summer.

I don't know why but when I hover a tag I only see an emty rectangle in which, I think, the ID is supposed to appear. This was the situation before I installed Greasemonkey/the script and it still is after.

Do you have the latest version of Firefox? What platform do you use?
Maybe it's just on some sentences, so please try N°579065, as I know it should work there.

There is probably something wrong with my home computer, as I see no problem at work. (Both Firefox 3.6.2, yes old one :D but home computer is a w7-x64, work is an xp-x86)
I will just check it and try to fix it. Thank you.

Thank you both.
I may be very stupid or lazy, but I have yet to figure out how to use those scripts (I'm using Chrome, by the way).

In chrome you have to install [Blank Canvas Script Handler](https://chrome.google.com/webst...mnpdfpdobpnkk) or use firefox with greasemonkey. After installing Blank Canvas Script Handler you have to go to my script's page and click the "install" button there.

and thank you, jakov, I love this most useful feature !

When you hover mouse over the tag, a tip appears saying the ID of the user who added this tag. Having the user ID you can find out the user himself, though it isn't straightforward: http://tatoeba.org/rus/users/show/<ID>

I realise there are few sentences in English featuring the word "straddle".
Could natives please create a few ? Thanks.

sentences_detailed.csv seems to be incomplete, still

accounting for arcticmonkey's leaving and the reduced number of active moderators (specially for German, a language with more than 100000 sentences), I was thinking to propose al_ex_an_der and Pfirsichbaeumchen as new moderators, if they or any of us have no complaints
what do you think? also do you have anyone else in mind?

we also need active moderators for Russian, French and Arabic languages
for Russian I'd strongly recommend Biga, if he agrees.

I'm already an advanced contributor, thanks to you. What do you mean by "moderator"?

Being a moderator means you can correct sentences from nonresponding users, delete tags and sentences
that's pretty much it.
Without active moderators, neglected sentences simply pile up one over the other. Also, the moderator is a thrusted contributor, since when there is a neglected sentence (no answer from the contributor for a month, let's say) the moderator has all the faculties to correct it as he wishes, provided that he acknowledges the contributor through a comment.

Thanks for the explanation. (Probably we should have a sort of FAQ on this.)
It's ok with me. I hope others don't mind.

ah, there is
right here
http://blog.tatoeba.org/2010/05...n-tatoeba.html

i agree for al_ex_an_der, he also haves good contributions for italian and i believe he really understood how tatoeba works

*has

As far as I know, they're both good contributors. So I agree.
Well, I am a little bothered by the Shakespeare quotes that are only in Esperanto and German!(http://tatoeba.org/eng/sentences/show/1294087) But that's OK, I've done the same for German authors. ;)

al_ex_an_der would be a great corpus maintainer! But we should ask him first...

Thank You for Your Confidence and Support!
If there are no objections against it, I am ready to join the team.
I noticed that Pfirsichbaeumchen is doing a very conscientious work and that's why I absolutely support the recommendation of marcelo.

You have my support Alexander! :)

I support the idea. Both Pfrischembaum and al_ex_an_der appear to know how Tatoeba works.

Possible spammer.
http://tatoeba.org/tur/sentences/show/1629462
http://tatoeba.org/tur/sentences/show/1629459
http://tatoeba.org/tur/sentences/show/1629457
http://tatoeba.org/tur/sentences/show/1629454
http://tatoeba.org/tur/sentences/show/1629450
http://tatoeba.org/tur/sentences/show/1629449
http://tatoeba.org/tur/sentences/show/1629448

Maybe s/he is just trying to understand how things work here, I don't know. But I left a comment at a sentence , changing nothing.

It seems to be a turkish user. Maybe one of our turkish contributors can write a PM and explain a little and be there for further questions?

I'm plannig to add a *typography* feature to my *MarkEdit* userscript and i would want to have some feedback about my plans:
In essence i would like to make something similar to [smarty pants][1], which is from the same author as [markdown][2]. The script replaces ASCII simplifications, that look similar to the real typographic symbols but aren't always available, with the corresponding proper unicode symbol, e.g. "--" would be replaced by a en-dash ("–") and "---" with a em-dash.
As we don't have unicode problems on Tatoeba, plus these typographic symbols are important for thesentences themselves, I will not implement it like the markdown code, but like the replace while typing feature that many desktop word processors have: e.g. when you type "(c)" there should be a popup that asks you if you mean to write the copyright symbol.
In the end i should be able to write {"Qu'est-ce que c'est_?"} and the script turns it to {« Qu‘est-ce que c‘est ? »} (without the curly brakets) if i'm in french mode (which is at the same time the "flag" for the sentence) or {"was ist los?"} with {„Was ist los?“}. It should handle the quotation marks, the apostrophe, non-breaking spaces (like {20_kg}), some special characters like {~n} should get {ñ}, {^2} converts to a superscript 2 etc.
[1]: http://daringfireball.net/projects/smartypants/
[2]: http://daringfireball.net/projects/markdown/

tio ŝajnas bone...

My [userscript](http://userscripts.org/scripts/review/134238) now comes with beta-functionallity for typhography replacement. There’s a bar above the textarea that indicates what you can replace at that position: e.g. if you move your cursor to "can't| wait" it will display `can't → can’t` (notice the right typographic apostrophe).
For now there’s just a limited set of replacements, that i have arbitrarily added, and just a few language sets. I am probably not going to add the languages I don’t know myself until someone tells me what typography is used there. Please look into the code and find the `replacementarray` to see how it is structured and tell me what you want to have added.
Also, at the moment you have to set your language in the code, so change `lang = 'fra';` to whatever you want. Later this will be achieved by clicking on your favorite flags.

I would like to ask the opinion of the contributors about whether I could become an advanced contributor. I think Tatoeba has a lot to offer but also needs a lot of work on its infrastructure, like joining the correct links, checking the sentences, managing the tags and so on, additionally to the regular sentence adding and translating.
I don't always have the ideas I need to post new sentences (I tend to only want to add sentences that are relevant to the language I write them in, since trivial sentences are in a more than satisfying number already - that's why I don't add tons of sentences and mostly translate them instead of inserting ex novo entries), but I most likely have the time to check sentences and directly link correct translations to each other and categorise them with the right tags, in addition to checking my native language sentences changing the tags as needed, and all this kind of work.
When I'll solve my health issues I'll be glad to contribute with some audio when possible.
I sent a private message about this to sysko some time ago, but he probably has too much to do on his own to bother about this, so I'll publicly ask the opinion of the project members as a whole, as it should be.

I'm very much in favour. And I appreciate that you present a profound outline of your views and aims both here and in your profile.

I'm also in favour
I was just about to recommend you as you soon as you'd collect some more sentences

Saluton, mi pensas ke Tatoeba povus esti bonega studa helpilo - ekz. mi havas grupon de frazoj el mia nova leciono en mia indigxena kaj studata lingvo. Mi entajpas ilin en Tatoeba kaj en "studada" modo mi povus esti vidi frazon nur en unu lingvo. Poste mi prononcas gxin lauxte, mi povus klaki la butonon kaj la traduko estus montrata. Ankaux havi grupojn de frazojn rilataj al diversaj temoj estus ege bone (eble povas uzi tagojn por tio, sed mi ne scias detalojn).
Hello, I think Tatoeba could be a great learning tool - e.g. I have a set of sentences from the new lesson in my native and studied language. I input these to Tatoeba and in "learning" mode I would be able to see a sentence only one language. After I say it aloud, I could click the button and the translation would be displayed. Also great thing would be to have sets of sentences which relate to a specific topics (I think it is partly done by tags, but dont know the details).

Welcome to Tatoeba!
The main idea of Tatoeba is to collect sentences. The sentences are free and open for everyone (just need to show the names of the contributors) so anyone can build a learning tool using certain sentences.
We are steadily working on cathegorizing the sentences through [tags](http://tatoeba.org/deu/tags/view_all) (btw you can always propose tags in the comments), but you can also use [lists](http://tatoeba.org/deu/sentences_lists/index) for general or personal purposes.
However, your proposal for is a good idea, which could be easily achieved through javascript or CSS in a userscript. I have to learn for an exam until july, but you can come back to me after that and i will write a userscript for you.

Bonvenon al Tatoeba!
Por kolekti frazojn pri specifa temo bone taŭgas listo.

Bonvenon ! Jen la listo de listoj http://tatoeba.org/fre/sentences_lists/index

An die anderen Vertreter der deutschen Sprache:
Vergeßt nicht, hin und wieder ein paar verwaiste deutsche Sätze zu adoptieren. Es scheint, daß der Berg überhaupt nicht kleiner wird.