Our data is released under various Creative Commons licenses.More information
If you love this content, please consider a donation.
Wall (5,591 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.
deniko
8 hours ago
Pandaa
8 hours ago
deniko
8 hours ago
Ricardo14
13 hours ago
Thanuir
yesterday
AlanF_US
yesterday
Dron007
yesterday
Thanuir
yesterday
CK
yesterday
CK
yesterday
I wonder why when I search by a tag:
https://tatoeba.org/eng/tags/sh..._tag/10420/ukr
I can see the same sentence twice:
https://i.imgur.com/JcrZxfZ.png
Those two are not just two sentences that look alike but differ in some "invisible" or "indistinguishable" character, if you click on each of those sentences you get to the same place, #8033352
Kétszer adta hozzá a cimkét valami okból.
Szedje le, majd tegye fel újra, úgy már csak egyet kell hogy mentsen.
I removed the tag, re-added it, and that actually helped, thank you for your suggestion, Pandaa.
I'll leave the screenshot here anyway as this bug is weird anyway, maybe it will resurface elsewhere. The sentence did show only one tag "autism".
Perhaps it'd be good if we know how many sentences we've added on https://tatoeba.org/eng/sentences/add since most of us* like statics (me included :) ) and maybe get some new ideas.
(As far "most of us" doesn't mean "everyone", I believe this feature can be "optional" (activated only in https://tatoeba.org/eng/user/settings )
What is this list for? https://tatoeba.org/rus/vocabulary/add_sentences
It contains wrong words for Russian, like "пофы́ркала" (with accents which is usually not used). There are also many weird and non-existing words.
а что не так со словом "пофыркала"? в books.google есть несколько примеров его использования.
С "пофыркала" нормально, но оно с ударением записано и скорее всего не будет найдено без него, будет везде так и выводиться, что неправильно. Кроме того, вот обнаружил там слово "сосулька", якобы без предложений. Но поиск находит несколько предложений с ним, правда, в других формах. Добавил в базовой форме ещё, ничего не изменилось. Ну и есть много совсем уже ошибочных слов/фраз: "будет ценить", "шётка", "развликать" и т.д.
на самом деле этот "словарь" опция весьма недоработанная и мало кто ей пользуется. я бы на вашем месте не стал придавать ей такое большое значение.
Понял. Просто изучаю, что тут есть. Я уже понял, что это просто пользователи вводят слова изучаемых языков, беря их из словарей, где слова идут часто с ударениями. Задумка в принципе хорошая, но список слов надо бы модерировать нейтивами.
I wrote a bit on how to use the list: https://tatoeba.org/eng/wall/sh...#message_32400
Nice post. The problem in Russian is with accent symbols which should not be added in the list of requested words. Probably this list should be moderated.
The acute accent symbol is intentionally ignored in Russian searches. However, I agree that it doesn't make sense to add it to words in the list of requested vocabulary unless it's used to disambiguate between two words that are otherwise written identically.
The list does need to be cleaned up, and not only in Russian. Unfortunately there is not yet any way to modify or delete bad entries, even your own.
Moderation or cleaning up the list would require a user interface, which requires programming. As far as I understand, volunteer effort on that front is gladly accepted.
In the meanwhile, you can add sentences to the words without the extra acute accents. It seems that the acute accent is used in learning materials and dictionaries, and with rare words, to indicate stress.
If you really want to work on the list of wanted words, you could add a few sentences which to this use. 'The acute accent in the word "word" indicates stress and is typically omitted.' or 'Books for learning the language use acute accents, such as in the word "word", to mark stress.'
But adding more than a few of these would not be that interesting, I guess.
The content of this message goes against our rules and was therefore hidden. It is displayed only to admins and to the author of the message.
** 2019-12-07 - Count of Links to English Sentences with Audio **
http://tatoeba.byethost3.com/en...019-12-07.html
** Members who have rated sentences OK **
http://tatoeba.byethost3.com/st...2019-12-07.txt
There are currently 1,012,509 OK ratings.
**Traduction des Conditions Générales d'Utilisation (CGU)**
Nous avons mis à jour nos CGU et elles ne sont pour l'instant qu'en français. Si vous comprenez suffisament bien le français et pouvez nous aider à les traduire vers d'autres langues, je vous invite à nous rejoindre sur Transifex:
https://www.transifex.com/tatoe...-terms-of-use/
C'est la plateforme que nous utilisons pour gérer les traductions de l'interface du site web. Si vous ne comprenez pas très bien comment l'utiliser, n'hésitez pas à poser vos question ici sur le Mur ou en m'envoyant un message privé.
Vous pouvez sinon vous référer à notre wiki[1] ou à la documentation de Transifex[2].
Merci!
---
[1] https://en.wiki.tatoeba.org/art...ce-translation
[2] https://docs.transifex.com/
J'ai essayé de rejoindre l'équipe pour l'anglais, mais je n'arrive pas à sélectionner "English" .
Je pense qu'il fallait d'abord que quelqu'un accepte ton inscription au groupe. Est-ce que tu peux réessayer maintenant?
Ça marche, merci !
This is a call for for help to all English native speakers and French native speakers.
The Terms of Use have been translated into English, but they need to be checked. Ricardo14 kindly took care of doing this translation, but he is neither an English native speaker, nor a French native speaker. He did his best so we can finally have some progress on having the Terms of Use in other languages than French, but now we need people who have actual knowledge of English and/or French to proofread and correct mistakes.
The translations are maintained in Transifex but I have copied the English translation here:
https://github.com/trang/tatoeb...h-translation)
Please help ensure that the English version is written in correct English and that the English version isn't saying something completely different from the French version.
Thank you!
I translated the Terms of Use into Turkish, mostly from English. However, I noticed that the most of the English text was taken from Google Translate. So, I had to check the French version too. I found many mistakes (some are very serious) but I only noted down a few (I corrected them in the Turkish translation tho).
I will submit my translation in a couple of days after editing the legal terms glossary that I created for this translation.
Now I'm sharing the mistakes that I found, so it will be easier for natives to check them.
By the string numbers on Transifex:
48. des auteur·e·s. | the authors · e · s.
(extra "· e · s")
79. il est de votre responsabilité de faire circuler avec elle sa licence | it is your responsibility to circulate with it its license
(with it its license)
82. (3Creative Commons3 SA : 4Share-Alike4) | (Creative Commons3 Ltd: 4Share-Alike4).
(wrong "Ltd")
89. double dots
97. Interdits comportementaux entre individus | Behavioral interdependence between individuals
(This is a big mistake. The French says prohibited behaviour)
102. Voir la section 1Propriété intellectuelle1, paragraphe 2Contenu sous droits réservés (copyrights)2 pour plus de détails. | See the section 1Intellectual Property1, section 2Content under copyright2 for more details.
(no "paragraph" mentioned; the title doesn't contain all)
103. Interdits de détournements d’usage | Prohibited use of hijackings
(misuse instead of hijacking)
105. / 109. Interdits informatiques / Interdits circonstanciels | IT banned / Circumstantial bans
("Prohibited" was used for "Interdits" several times)
112. (This one has too many mistakes to write. For example, the English says "tits" instead of "its")
117. (extra spaces)
126. / 135. / 140. / 141. / 150. ‘Conditions Générales d’Utilisation’ was mistranslated.
126. que votre profil soit supprimé, bloqué, radié ou tout simplement laissé inutilisé | your profile is deleted, blocked, deleted or simply left unused
(the second "delete" may be "write off")
136. nous ne pouvons pas nous engager à un maintien en ligne de notre Site Internet sans aucune discontinuité. | we can not commit to maintaining our Website online without any discontinuity.
(cannot commit ourselves)
-----
— It was spelled as "can not" many times.
— Like 126. / 135. / 140. / 141. / 150, there are several wrong translations of titles within the text.
— There are some problems with "site", "web site", "internet site".
— Sometimes it's "prohibited" or "forbidden" for the same French verb.
Thanks! I'm going to check it out.
> des auteur·e·s
Ceci n’est pas français.
Au mieux, le féminin de auteur, est autrice ou autoresse. Au pire, il conviendrait d’écrire : auteurs et autrices, Mais ceci serait ridicule car ce féminin est inutile au pluriel collectif, qui est neutre.
En français, aucun féminin d’un masculin en -eur ne peut se terminer en -eure.
Directrice
Puéricultrice
Actrice
Sénatrice
Conductrice
Doctoresse
Chacun ne peut pas improviser des formes à sa guise...
à ce propos, le néerlandais est dépourvu de genre féminin, faut-il aussi l'improviser pour complaire aux féministes des CGUs ?
I haven' read everything yet, but here is what I noted from line 109 (on Transifex)
112: We prefer anticipating rather than punishing. However we may need means of sanction against an account. They are..., their actual use...
Using them is neither...
124: I'm not sure the capitalized Agreement is needed
126: Instead of "your profile is deleted, ..." maybe "have your profile been deleted, ..."
135: this needs native check, at least the last sentence should be something like "The other parts of the Terms of Use will remain in effect,"
137: "applications" maybe better than "apps"
142: These Terms of Use
141: A moficiation of the Terms of Use...
Instead of "the amendments", maybe "the modifications" as it is used in the previous sentence.
142: Conditions Générale*s* (French side)
Et je suis d'accord avec sacredceltic sur cette forme d'écriture ridicule. Il y a la même avec encouragé·e ligne 130 et sûrement d'autres. Cependant cette forme ne se trouve pas quand les CGU parlent de "nous" côté Tatoeba, alors que pourtant il y a bien des femmes. Ce qui tend à rendre l'usage de cette forme encore plus ridicule.
Alors qu'une forme correcte d'inclusion existe depuis toujours en français pour quelqu'un qui souhaite insister dessus, comme l'écrit sacredceltic: « des auteurs et des autrices ». De la même façon, on peut écrire « vous êtes encouragés et encouragées », si l'on souhaite insister sur l'inclusion en écrivant un français correct.
c'est d'autant plus ridicule dans des CGU, car la loi, elle, est inclusive. Donc les pluriels neutres n'ont pour elles aucune ambiguïté et n'excluent personne.
I took care of the changes suggested by Gulo_Luscus and Aiji that had not already been made.
Thanks, Alan!
J'ai beaucoup contribué à la traduction du site sur Transifex à une certaine époque mais les modalités de contribution ont changé il y a de ça quelques années et il faut désormais être approuvé et avoir un "ranking", dispensé par on ne sait qui de moins qualifié que soi-même... C'est en même temps risible et complètement contre-productif, parce que, conséquemment, je constate que de nombreux éléments de l'IHM sont désormais non traduits en français (y compris le message indiquant qu'on peut utiliser la langue de son choix qui est désormais en anglais sur l'IHM en français...le comble !) mais je ne peux plus y remédier du fait de ces procédures d'approbation.
On marche sur la tête...
(While I can read most French, I'm not good in writing in French. So I hope it's ok for you if I'm answering in English.)
> il faut désormais être approuvé et avoir un "ranking", dispensé par on ne sait qui de moins qualifié que soi-même
Out of curiosity: where do you see that "ranking" on the Transifex page? I've applied for the German Team a few months ago and after approval (there were no conditions to fulfill) all my translations were added to the site without problems.
> y compris le message indiquant qu'on peut utiliser la langue de son choix qui est désormais en anglais sur l'IHM en français...le comble !
Well, that sentence was changed recently and it always needs some time until the translations show up. I've just noticed that this sentence was translated into French a few hours ago. Trang usually updates the translations every week.
Je ne sais pas à quel système vous faites référence, mais actuellement, pour le projet Tatoeba, il suffit de demander à participer à la traduction (rejoindre l'équipe français, l'équipe allemand, etc.) et que quelqu'un accepte pour pouvoir contribuer. J'imagine que c'est simplement un système pour assurer que n'importe qui ne puisse pas contribuer n'importe quoi. J' ai rejoint deux équipes, à chaque fois, j'ai été rapidement accepté.
Quant aux phrases non traduites en français, je m'en suis justement occupé hier. Les traductions devraient apparaître à la prochaine itération.
Eh bien le message que je mentionnais est toujours en anglais...
Il manque sans doute à faire une publication...
Et il y a aussi le « STATS PER LANGUAGE » qui est très irritant...
Patience, patience. Cela sera réglé au prochain déploiement.
Il n'y a qu'une chaîne non traduite en français actuellement et c'est parce que sans contexte, j'ai préféré ne pas le traduire.
Kiitos!
How can I translate sentences random if I want to see the translated sentences in different languages as well (as it used to be)?
Toimiiko tämä?
https://tatoeba.org/hun/sentenc...&sort_reverse=
Käytin oikeassa sivupalkissa olevaa mahdollisuutta näyttää käännökset kaikille kielille.
I would like the admins justify their behavior on the following sentence. The fact that we can't see the history of comments is a shame, but anyway...
https://tatoeba.org/fra/sentences/show/1996169
So is it okay if I go to the English corpus and I replace all US style quotation marks by UK style, because I don't like it really, all "Mr" by "Mr.", because let's be honest not writing a dot is clearly stupid, and also I will probably remove all those useless commas here and there, because each of them takes two bytes of memory, such a waste.
More seriously, the question of duplicates due to the different spaces in the French corpus has been discussed several times in the past. It was discussed by French contributors, corpus maintainers and admins, and the last time it was discussed, I clearly remember exchanging with gillux about an "edition rule". You know, ideas and arguments, not just "gna gna I'm frustrated by those duplicates". We tended to agree that if a space was used, it should be the correct one (espace fine insécable), but we did not reached a consensus on what to do.
Now, some frustrated people come in a corpus that they have nothing to do with.
That's not their business. They haven't contributed French sentences. They don't know the language. And they decide that all spaces should be erased and merged into the solution they chose.
Where does this idea come from? I guess that if a solution was to be adopted, the French contributors, and the French corpus maintainer would know about it. Does the French corpus maintainer knows something about it? Well, no he doesn't, because nothing has been decided. And if something were to be decided, don't you think an algorithmic way of doing it would be chosen? And if not, if really it was decided to do it by hand, that it would be the job of the French corpus maintainer and not some random people passing by?
Now, it doesn't need to be an open-ground battle. People can be reasonable. If you are frustrated, go out for a walk, hit the gym, play the drums.
Oh, and also, as I and others have already mentioned several times, near-duplicate are a non-problem. They get smoothed down as time goes on, and not sharpened up as some people want you to believe. 300 sentences that may be merge would represent less than 0.1% of the French corpus.
https://github.com/Tatoeba/tato...ment-414060133
Now, this is NOT a thread to discuss the use in spaces in French. No, no, let's not avoid the main subject here. This is a thread to discuss the behavior of some people, and if contributors of Tatoeba think that it's okay to do whatever you want.
> This is a thread to discuss the behavior of some people, and if contributors of Tatoeba think that it's okay to do whatever you want.
Okay, let's discuss behavior. You were frustrated by someone's action. You left comments on the sentence in question. So far, so good. Since the log indicated who had taken the action you disagreed with, you could have sent private messages to them, and waited a decent amount of time to see whether you got a reply. I don't know whether you sent any messages, but if you did, you didn't wait very long for a response, because a few hours later you received a comment on that sentence from one of the people you were addressing -- but you had already decided to take the issue to the Wall, an hour after you left your own comment.
It is fine to ask why someone took a particular action, but you will get the best answer if you begin by assuming the best intentions, by posing the question to the smallest number of people involved, and only escalating by degrees. Naturally, that's hard to do when one is upset. But it's worthwhile.
I think we will all agree that it's not okay for anyone to do whatever they want :)
The case you mention is a rather special case and I don't think it should be taken as a red flag for admins abusing their power (if that was the main concern behind your post).
1) The last discussion we had regarding spaces before punctuation was mostly written in French and I think it's understandable that not every admin will have read it in details. Someone who hasn't read that conversation in details will not have the reflex to think that changing a space could lead to a big problem.
2) The sentence that was modified could only be modified with the intervention of an admin because sentences with audio are not editable by anyone. To edit such sentences, an admin has to temporarily unset the audio for the sentence to become editable, then set the audio back once it has been edited.
3) The person who reported that #1996169 is duplicate of #478907 is a French native speaker.
4) The sentence #1996169 was modified to match #478907. #478907 was an older sentence and was actually added by the very same contributor who added #1996169. I'm not sure if this had any influence, but anyway...
Yes, it could have been indeed more appropriate to first ask for confirmation on whether it is really okay or not to edit the sentence and wait for the owner's response, but given the context, I think it's understandable why the sentence was nonetheless edited without asking for approval from the owner. It was indeed a misstep, but a minor and honest one. Nothing really dramatic.
You were perhaps more annoyed by the comment that suggested to edit all other similar duplicate sentences. But let's keep in mind it was only a suggestion. No one agreed to do such a thing and I could be wrong but I personally doubt any admin would have proceeded without asking first if it's really okay. People will usually think a bit more about their actions before doing things en masse.
Aside of wholeheartedly agreeing with what Alan said above, I would also suggest, if you are afraid admins and corpus maintainers start taking too much liberty in fixing the corpus, that you can help create guidelines in the wiki. We have an article on "How to be a good contributor". You are more than welcome to write something in the same fashion on "How to be a good corpus maintainer" which we could ask every aspiring corpus maintainer to read, and which will also serve as are reminder to all corpus maintainers of what are good practices.
> The last discussion we had regarding spaces before punctuation was mostly written in French ...
Perhaps you or someone else could briefly explain this discussion for the rest of us.
What are the major conflicts between French speakers about the spaces used before final punctuation?
Does the suggestion I made on this issue make sense to people?
https://github.com/Tatoeba/tato...ment-560038998
Or at least link to the discussion. Google translate knows French, so the main ideas should come through even without knowledge of the language.
On Github Trang mentioned https://tatoeba.org/eng/wall/show_message/29619
Danke.
> Does the suggestion I made on this issue make sense to people?
No it doesn’t. It’s been years you don’t understand this problem, but we don’t need you to understand it as long as it is solved by people who do.
The reasons why you don’t understand it is because you take Internet typographic rules for granted and universal. They’re not to be taken so.
US admins shouldn’t be part of this debate because they just don’t get it. It’s hopeless to explain it to them. And that is why US developers should have nothing to do with the elaboration of the CSS of an international website such as Tatoeba.
Each time they’re in charge, on any international website, they create havoc with the typography of other languages and the sites end up looking americanised and juvenilised.
> US admins shouldn’t be part of this debate
> because they just don’t get it. It’s hopeless to explain it to them.
This is both extremely rude and extremely unreasonable.
It’s just factual...I can’t put it otherwise...
You should also tell Arnold
https://tatoeba.org/fra/sentences/show/450604
Hence my "too bad the history of comments is not shown" because there clearly was a "I'll do it, although I hope there isn't too many of them" in the comment section. And that is also an answer to Alan. My post on the wall came after that comment disappeared. So of course, it's easy to put the fault on the upset guy (and I do apologise for that) but please realize you didn't have all cards in hand to judge my posting on the wall. If it was only about Pfirsichbaeumchen editing the sentence, I would have kept it low as I imagined she didn't mean bad. But we were talking about massive editing without discussion here. Of course, I'll drop it on the wall...
I see. I didn't realize that a comment had been deleted.
there are professionals of stealthy comments...
It's fine to get cranky once in a while. No one is blaming you for that, don't worry.
The main thing I wish to know is what can we do about this incident. Do we just consider it was a very special case and there's nothing to worry about? Or should we work on improving something?
oui il est urgent de faire quelque chose, parce que très régulièrement des admins s'arrogent le droit de corriger la ponctuation de phrases correctes qui ne leur appartiennent pas sans demander aucune autorisation, comme ici https://tatoeba.org/eng/sentences/show/467940
Je trouve cela parfaitement scandaleux. Il s'agit de fascisme, ni plus ni moins.
Des suprémacistes anti-espaces qui veulent imposer leurs vues alors que même l'inventeur de l'espéranto, Zamenhof, avait déclaré ne jamais vouloir imposer de règles en la matière, parce que c'est un sujet trop sensible.
Ces suprémacistes devraient être sanctionnés en se voyant retirer leurs statuts d'administrateurs et leurs capacités à violer les phrases des autres contributeurs.
Car c'est bien d'un viol qu'il s'agit. Viol du droit d'auteur, viol de la création d'autrui.
Avez-vous des propositions concrètes pour améliorer la situation ?
Retirer les accréditations des admins qui outrepassent leurs prérogatives. Déjà il faudrait faire un exemple. Ca calmerait les autres...
Pour connaître la liste des contrevenants et le nombre d’occurrences des faits par chacun, il suffit de compter le nombre de phrases mises à jour par des admins sans commentaire préalable. Il n'y a qu'à sanctionner celui qui arrive en tête. Simple.
Toujours dans la demi-mesure !
Je vois l'idée, même si je n'adhère pas, mais il faudrait recibler un peu car mettre à jour sans commentaire préalable, ça arrive souvent. Par exemple, je ne laisse pas de commentaires si je corrige des guillemets ou d'autres broutilles sur une phrase d'un utilisateur que je sais inactif. Il m'arrive même de détruire des phrases sans laisser de commentaires, quand on sait qu'il s'agissait de bots par exemple.
Ouais bon en même temps, personne n'est jamais mort de se voir retirer des accréditations...
>je corrige des guillemets ou d'autres broutilles sur une phrase d'un utilisateur que je sais inactif. Il m'arrive même de détruire des phrases sans laisser de commentaires, quand on sait qu'il s'agissait de bots par exemple.
Tu peux limiter l'analyse à mes phrases. Je ne suis pas inactif et je ne pense pas être un bot...
Ça fera déjà un bon échantillon et puis le résultat m'intéresse...
Si quelqu'un veut s'occuper de faire l'analyse, l'export des logs est ici:
https://downloads.tatoeba.org/e...utions.tar.bz2
et l'export des commentaires est ici:
https://downloads.tatoeba.org/e...mments.tar.bz2
Je ne me porte pas volontaire, j'ai beaucoup trop d'autres choses à faire.
Je n'exclue pas l'idée de sanctionner les gens, mais seulement en tout dernier recours. Avant d'en arriver à là, il faut au moins quelques avertissements.
De mon point de vue, il y a seulement un problème de communication. Je ne pense pas qu'il y ait au sein de Tatoeba un groupe de suprémacistes anti-espaces. Certaines personnes ne sont pas conscientes du problème, c'est tout. Il suffit de leur expliquer gentiment.
Oui, vous êtes plutôt tendance bisounours.
It’s a wild world, baby...
Ce n'est pas un format que je sais exploiter (concaténation de timestamp+ ordre insert sans spécifications de nom de colonnes...) . Auriez-vous un mode opératoire, s'il vous plaît ?
Merci
Format for the comments:
Comment ID [TAB] sentence ID [TAB] username [TAB] creation datetime [TAB] comment text [NEWLINE]
Format for the contributions:
username [TAB] contribution datetime [TAB] action [TAB] type [TAB] sentence ID [TAB] sentence language [TAB] translation ID [TAB] sentence text [NEWLINE]
action can be one of "insert", "update" or "delete"
type can be one of "link", "sentence" or "license"
See also https://github.com/Tatoeba/tato...ly_exports.sql and https://github.com/Tatoeba/tato...tributions.sql if you understand some SQL
Yes I'm SQL fluent. But I first need to inject this huge ".CSV" file (which is not a proper csv format...) in a DB I can access.
I couldn't manage to do it so far because the format is not recognized by my Oracle SQL Developer' data import tool. The encoding is wrong with "\N" strings in place of null values and different other issues.
And the same goes with "CSV" files that I export from the site...
These files are actually .dsv files(tab delimited) but what kind of encoding is that ?
The files should be UTF-8 encoded AFAIK.
I agree that these exports aren't easy to handle unprocessed but that's probably why they are unofficial and not listed on the Downloads page.
A few weeks ago I've worked with the wall export file and I had to preprocess it due to the embedded newlines in the wall messages.
I'm currently trying on Windows and that doesn't work :Tabs and LF within the comments themselves are interpreted as the start of a new column or record...
I'll try from a Mac next week...
Too bad that all of Tatoeba's data is not available from a public Datalake that anyone could leverage with data analysis tools...