Tatoeba is a platform whose purpose is to create a collaborative and open dataset of sentences and their translations.
Home Page: https://tatoeba.org
License: GNU Affero General Public License v3.0
The languages selected for the "From" and "To" drop-down lists for a phrase search are not retained when a new window is opened. Instead, the selection always flips back to "Any". However, switching the interface language does preserve the selected languages in the same window. In addition, when opening a new window, the selected language for adding new sentences is retained, and the interface language is also retained.Reported by Griza Leono and marafon on the Wall in January 2014, after the December 2013 crash.Additional note: pne points out that the same is true for random sentence::::Example: http://tatoeba.org/eng/sentences/show is supposed to show an English sentence, but it shows any sentence.Choose a language from the drop-down at the top (e.g. Czech) and click "Random", and you get a Czech sentence... but the drop-down has turned back into "Any language".
Send automatic notifications for a wider variety of events, including when translations are linked to or unlinked from one of the sentences that I own.
AlanF_US, Wall
Original ticket: https://www.assembla.com/spaces/tatoeba2/tickets/404
./add_lang.sh kal Greenlandic 3277
Tatoeba is difficult to browse and use with a mobile device. We should either make the design responsive, or have a stripped-down mobile version, or maybe a dedicated app.
orion17, Wall
Original ticket
There were two instances of __() applied where they shouldn't be. Here are the changes I've made and am about to check in:app/views/helpers/comments.php:292:--- app/views/helpers/comments.php (revision 2080)+++ app/views/helpers/comments.php (working copy)@@ -289,10 +289,7 @@$this->Html->image(IMG_PATH . 'edit.png',array(- "title" => __(- $tooltip,- true- ),+ "title" => $tooltip,"width" => 24,"height" => 24)app/views/pages/help.ctp:248--- app/views/pages/help.ctp (revision 2080)+++ app/views/pages/help.ctp (working copy)@@ -245,7 +245,7 @@),$html->image(IMG_PATH . 'translate.png'));- __('');+ '';?>
Change: "The user %s you to want to send this message to does not exist. Please try with another username."to: "The user %s to whom you want to send this message does not exist. Please try with another username."http://subversion.assembla.com/svn/tatoeba2/trunk/app/controllers/private_messages_controller.php:168
./add_lang.sh hat "Haitian Creole" 3278
./add_lang.sh tuk Turkmen 775
Provide a function that lets you search for existing duplicates or near-duplicates of a sentence that you're about to submit, ideally without taking you to another screen and disturbing your workflow. There are various ways that this function could be executed:
Original ticket: https://www.assembla.com/spaces/tatoeba2/tickets/396
Select list for a sentence instantly rather than requiring OK button. Perhaps this could be controlled by a preference setting.FeuDRenais, Wall
Original ticket: https://www.assembla.com/spaces/tatoeba2/tickets/390
Going to "Wall" on the virtual machine gives this error (possibly because the Wall is empty):Warning (512): SQL Error: 1054: Unknown column 'hidden' in 'where clause' [CORE/cake/libs/model/datasources/dbo_source.php, line 527]Query: SELECT Wall.id, Wall.owner, Wall.parent_id, Wall.date, Wall.title, Wall.content, Wall.lft, Wall.rght, User.username, User.id FROM wall AS Wall LEFT JOIN users AS User ON (Wall.owner = User.id) WHERE hidden = 0 ORDER BY Wall.date DESC LIMIT 10
Jim Breen e-mailed Team Tatoeba on 2014-02-16 to say:The weekly generation of the wwwjdic.csv file is not taking place. Theone at http://tatoeba.org/files/downloads/ is dated 6 Feb.
When I go to Contribute/Discuss sentences on my virtual machine, I get this (possibly because the results returned are empty):Warning (512): SQL Error: 1054: Unknown column 'SentenceComment.hidden' in 'field list' [CORE/cake/libs/model/datasources/dbo_source.php, line 527]Query: SELECT SentenceComment.id, SentenceComment.user_id, SentenceComment.text, SentenceComment.created, SentenceComment.modified, SentenceComment.sentence_id, SentenceComment.hidden, Sentence.id, Sentence.text, Sentence.lang, Sentence.user_id, User.id, User.username, User.image FROM sentence_comments AS SentenceComment LEFT JOIN sentences AS Sentence ON (SentenceComment.sentence_id = Sentence.id) LEFT JOIN users AS User ON (SentenceComment.user_id = User.id) WHERE 1 = 1 ORDER BY created DESC LIMIT 50Warning (2): Invalid argument supplied for foreach() [APP/controllers/components/permissions.php, line 144]
al_ex_an_der posted on the Wall:From time to time tags are created by mistake. It would be helpful if tags would be deleted automatically (without a necessity of human interference) as soon as there's no longer any sentence connected with them.
Add "-20" and "+20" to the existing targets for a long list ("<<", "<", x, x+1, ... x+5, ">", ">>").
Original ticket: https://www.assembla.com/spaces/tatoeba2/tickets/416
Exchange between Trang and me:
I wrote:
(Phrase 4) Role of moderators
app/views/pages/contribute.ctp:66
We should probably change this to something like:
Role of advanced contributors-> Role of corpus maintainers
Trang responded:
The status of "Corpus maintainer" used to be called "Moderator" but we changed the label because people were expecting the moderators to moderate the community, while we were only expecting them to moderate the corpus (cf. http://blog.tatoeba.org/2011/05/new-users-status.html).
Anyway I think we should remove these links from the "Contribute" page (http://tatoeba.org/eng/contribute, in "Important to read") and replace them with links to the wiki instead.
Offer an online method to search one's Tatoeba mailbox for a string.
AlanF_US, Wall
Original ticket: https://www.assembla.com/spaces/tatoeba2/tickets/403
Improve handling of sentences to which an OK tag currently cannot be added. For instance, modify the current message to explain that you cannot add an OK tag to your own sentence.
Original ticket: https://www.assembla.com/spaces/tatoeba2/tickets/397
quote of email from dpsilver7
Ar / ara (Syrian Accent) (it's not classic Arabic, no one speaks Classic Arabic anymore!! :D, it's vernacular, all Arabs and most foreigners I know (that were learning Arabic) wanted to learn Syrian Accent.
Segregate sentences by learners from sentences by natives. This can be done either via a two-category approach (natives and non-natives) or via a rating on a scale.
alexmarcelo and others, Wall
Original ticket: https://www.assembla.com/spaces/tatoeba2/tickets/393
CK writes:
The suggestion:
Make the brackets [ ] part of the clickable part of the link for the "shorthand" used on comments.
In other words, instead of showing this on the page.
3030038 誰か鉛筆持ってない？...
show this.
[#3030038] 誰か鉛筆持ってない？
The reasons:
Initially requested asked by "ken p" via e-mail to Team Tatoeba. He included the following example:
what is your name?...............English
तुम्हारा नाम क्या है?......................Translation into Hindi
Tumhārā nāma kyā hai?........Transliteration of Hindi
તમારું નામ શું છે?...........................Translation into Gujarati
Tamāruṁ nāma śuṁ chē?....Transliteration of Gujarati
Rolling over a tag should show username rather than (or in addition to) user number (ID) of contributor who left tag.
AlanF_US, Wall
Offer an online method to search the Wall for a string.
AlanF_US, Wall
Original ticket: https://www.assembla.com/spaces/tatoeba2/tickets/402
sharptoothed posted on the Wall:"It would be also nice if tag auto-suggestion drop-down list get worked again."
Offer an option to let automatic notifications go to one's Tatoeba mailbox (private message inbox) instead of or in addition to an external mailbox.
AlanF_US, Wall
Original ticket: https://www.assembla.com/spaces/tatoeba2/tickets/401
This is the ticket for the API component as it is in CakePHPA few things:1) The methods throw their own exceptions, containing at least a code and a message, which then the component catches. These errors are mainly to indicated to the user which parameters the method accepts. So, for example, if the client provides the wrong arguments for a method, the client will get an error message saying "method_name only accepts the follow arguments: arg1, arg2, arg3".2) Each method manages its own version3) Still need to create routing rule for API and figure out permissions schemeDetailed information can be found at the wiki:https://github.com/trang/tatoeba-api/wiki/Tatoeba-API-specification-2
See attached file. Two examples:Adopting a sentence is also part the "quality process". You can find more information about it here: app/views/pages/help.ctp:315Sentences lists app/views/pages/help.ctp:328"part the" -> "part of the""Sentences lists" -> "Sentence lists"
gleki, Wall (though gleki posed this as a question rather than a request)
It would be nice if you could retrieve and display a page containing more than the default number of sentences (currently 10) for a given request . Possible ways to specify the number of sentences:
(1) Build the number into the request URL as a parameter.
(2) Allow the user to choose from a set of possible values via the GUI.
(3) Set the default value from your profile.Since setting the value too high might cause problems with our traffic load, we might want to think about setting a maximum value, and if so, what it should be.
Allow unsubscribed users to see the sentences of a particular user.
tommy_san reported (in private correspondence) the following issues with the set of UI strings to be translated in Launchpad:
Here's the list of missing strings I found:
(1) On every page at the bottom: Blog, Twitter, Facebook
(2) Names of some new languages:Algerian, Arabic, Ancient Greek, Hiligaynon, Picard, Standard Tibetan
(3) http://tatoeba.org/eng/sentences_lists/index create (the text inside a button)
(4) In the title of pages, as in http://tatoeba.org/eng/contributions/latest Tatoeba (This must be the same in the languages that use Latin alphabet, but not always so in other languages.)
(5) The whole "Improve sentences" page (and I believe this page needs to be updated) http://tatoeba.org/eng/activities/improve_sentences
(6) In the menu More, Quick Start Guide, Tatoeba Wiki
And here are some other things that I noticed:
[a] http://tatoeba.org/eng/contribute The link to "Role of moderators" must be deleted or changed, since there's no "moderator" anymore.
[b] http://tatoeba.org/eng/help I think the sentences "Note that translations are limited to registered users. Click here to register." should be deleted. Almost all other things written on the page are what only registered users can do anyway.
[c] http://tatoeba.org/eng/terms_of_use In the terms of use, there are some places where the French and English versions don't really seem to match. For example, a whole sentence beginning with "En d'autres mots, vous ne pouvez importer du texte que s'il est" is missing in the English translation. I'm not sure if it's great problem, but when we think of the fact that all other translations are translated from the English, it might be a good idea to make a more strict translation of the French, which is the only valid one.You might be able to see the difference clearly if you take a look at the French version of this page: http://tatoeba.org/fre/terms_of_use where the English translation is partly re-translated into French so you can compare them both. (This meaningless translation on this page must be deleted at any rate.)
Improve the algorithm used to automatically identify languages.
sacredceltic, Wall
JimBreen, Wall
Jim wrote that he recently came across two cases where sentences were overwritten by versions in different languages. He thought it occurred because the person/persons who did it didn't know the processes that well. I asked him to clarify whether he meant that a warning should be displayed to the user or the admin, or both. He said that he meant both. In further discussion, he and tommy_san said that whereas changing a sentence's flag was likely to be legitimate when the sentence was new, it was less likely to be legitimate later, so perhaps a warning should only be issued if the sentence was old.
Original ticket: https://www.assembla.com/spaces/tatoeba2/tickets/413
From e-mail from "moustache sans barbe" to Team Tatoeba:
Toutes les transcriptions phonétiques des exemples (il y en a plusieurs) de phrases comprenant "...に入っていた" est retranscrite par erreur en "...には/いっていた" ce qui est un contresens sans rapport.
Je vous laisse corriger cette transcription orale.
Cdlt
In English:
All the phonetic transcriptions of the examples (there are several) of phrases containing "...に入っていた" is [sic] incorrectly retranscribed as "...には/いっていた", which is an unrelated mistranslation.
I'll leave it to you to correct this oral transcription.
Cdlt
If a newly created sentence is assigned a language with one directionality, then it will continue to be displayed with the same directionality even after the language is changed. For instance, say that I add an English translation "The sky is blue." to a Hebrew sentence, but I mistakenly mark my translation as Hebrew. I see my mistake immediately because the final period clings to the beginning of the sentence (the left) rather than the end (the right). Now I click on the flag to change it to English. Even though I've changed it, it will still be displayed on my current page with right-to-left directionality (with the period clinging to the beginning rather than the end). If I actually go to the URL of the sentence, it is displayed correctly, however. Thus, this is a transient annoyance rather than a permanent problem. It would still be nice to fix the problem, though.
Original ticket: https://www.assembla.com/spaces/tatoeba2/tickets/407
There is no help for the search engine under the FAQ or help links. You can only find the help after you actually perform a search (e.g., search for "blue" in English). After the search is completed, help for the search engine is displayed in the upper right-hand side of the page. This help should also be incorporated into the FAQ and help links. I'm considering this a bug rather than an enhancement because searching is basic functionality that should be covered in the FAQ and help links.
Display owner (if any) next to sentence in more contexts. For instance, display it next to each translation of a sentence in a list of sentences (e.g., sentences in English not translated into language X.
AlanF_US, Wall
Display a list of all tags sorted alphabetically, not only by frequency.AlanF_US, Wall
TODO
CALL add_new_language('sag', 0);
Language name: Punjabi (Eastern)
Language info: https://en.wikipedia.org/wiki/Punjabi_language
ISO 639-3 code: pan
Public List: https://tatoeba.org/eng/sentences_lists/show/3279
Icon: http://goo.gl/QgFcfo
Requested by: sabretou
Previous messages:
TODO
- Also change string for pnb from "Punjabi" to "Punjabi (Western)".
- icon for pnb: http://goo.gl/F3rBNK
ead http://tatoeba.org/eng/wall/show_message/17381#message_17381
The changes mentioned in that thread involve splitting the existing Punjabi language sentences into two, each with a different script:"(pan) is Eastern, or Indian Punjabi, predominantly written in Gurmukhi and spoken in India, while (pnb) is Western, or Pakistani punjabi, written in Shahmukhi and spoken in Pakistan."Once I get two lists from sabretou (one for Eastern, one for Western), I'll add a link to this ticket.
./add_lang.sh sin Sinhala 3262
Currently, the search bar issues the same message when it's not working at all as it does when it fails to find results for a given query. This should be changed.
freefighter, Wall
Very long words get cut off, as in this Turkish example:
http://tatoeba.org/eng/sentences/show/1532594
The answer might be autohyphenation, at least for some languages.
alpha44, private communication
alpha44 would like to have a "See translations for" button for each of the sentences in a list of favorite sentences, whether the list belongs to the user or to someone else.
Original ticket: https://www.assembla.com/spaces/tatoeba2/tickets/414
Make a new option "Display orphan sentences" in the settings page and set it "off" by default.
tommy_san, Wall
Original ticket: https://www.assembla.com/spaces/tatoeba2/tickets/391
On the Wall, orcrist posted the following::::For Japanese and other languages with character sets as opposed to alphabets, when people are learning the languages, we start with a phonetic alphabet. This is often written in small letters just above the actual text. These small letters are called "rubies", and for Japanese specifically, the phonetic reading written adjacent to a word is called "furigana". Before the server trouble at the end of 2013, Tatoeba showed furigana (albeit with some errors).Whether it is worth the admins' time to work on re-enabling furigana support is an open question. If there are other high quality free resources that work well in replacement of it, perhaps it is not necessary. Rikaichan works well as long as you are reading everything from within your web browser.
Recently, Ticket 384 was addressed, which was nice to see. However, I think an additional change needs to be made.
The subject of the ticket was "Place 'Help' and 'FAQ' links at top of page, not at bottom". The problem was that "Help" and "FAQ" links are extremely important but had very low visibility (at the bottom of the page). A change was made to put them at the top of the page. However, now they're underneath a "More" link. They were not obvious at all to me there, and I'm sure they would be less so to someone who is unfamiliar with the site.
There is plenty of room for another item on the top bar. I would suggest that a "Help" link be added between "Wall" and "More". Then under "Help" you could have "FAQ", "Quick Start Guide", "Tatoeba Wiki" and "Help" itself, though perhaps there should be some way to distinguish between top-level "Help" and the actual "Help" link itself. Maybe that lower-level link should be named something like "General Help"?
Meanwhile, I think that "Downloads" and "Tools" should remain under "More".
In most e-mail clients, there's a way to save messages as drafts without sending them. This lets the user work on them until they're ready to send. This would be nice functionality to provide in our private message area as well.
Original ticket: https://www.assembla.com/spaces/tatoeba2/tickets/408
Bring up a randomly generated page within a set (e.g., English sentences with audio not translated to Portuguese). marcospcruz, Wall
Offer a mechanism to request translations of specific sentences by natives.
tommy_san, Wall
Original ticket: https://www.assembla.com/spaces/tatoeba2/tickets/394
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
Django
The Web framework for perfectionists with deadlines.
Laravel
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
Microsoft
Open source projects and samples from Microsoft.
Google
Google ❤️ Open Source for everyone.
Alibaba
Alibaba Open Source for everyone
D3
Data-Driven Documents codes.
Tencent
China tencent open source team.