Comments on Selena777's sentences
Translate Selena777's sentences
Selena777's messages on the Wall (total 230)
I don't see such a page when I download files.
Actually, I didn't know it was a special program to download them. How can I find it?
There is UTF-8 in my browser settings. Should I change the settings in Excel?
I can't download any list that isn't in Latin script correctly. There are only strange symbols instead of normal letters. Is it a bug?
There are some guggestions.
https://github.com/Tatoeba/tatoeba2/issues/48 - Display owner (if any) next to sentence in more contexts - "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 " More info - http://tatoeba.org/eng/wall/sho...message_16640(
I get the idea now. That seems interesting. But "a trustworthed person regarding their contribution in the certain language" and "a person I feel comfortable to talk to and ask things" is not the same. Should only people that match both criteria be included in the friend list?
Could you explain more about the idea?
I don't mean linking any sentences like a "stamp of approval".
I can also link sentences in my non-native languages to each other, if they have obviously the same meaning and there is no chance for subtile differences of meaning, that I couldn't recognize. If I see an obvious mistake, I'll comment it.
I think, if someone considers linking a sentence by the same way as tagging OK, they just don't completely understand how the project works.
Customized pages are really helpful thing.
How can I add some non-native speakers, which I trust, to my customize page?
In my opinion, there are different solutions:
1) Try to block them through their IPs or something similar (I'm not able to give any advice in particular how it can be done);
2) Block those users as soon as you recognize them with their new accounts and mark red and release all their contribution to let other contributors adopt and correct their sentences, and delete those of them which are insulting or couldn't be understood and corrected.
3) If they only contribute "bad" (grammatically incorrect or unnatural sentences), but not insult nor troll anyone, you can let them contribute "as is", but mark it red, and encourage other contributors check and correct their sentences, and give corpus maintainers and advanced contributors the right to remove red colour from their native-like language sentences.
I think, trolls and vandals are something that any open sourse project have to deal with, it's an inevitable evil, but also it's important to not turn people into vandals with some unfriendly attitude. And it would be helpful also encourage admins and corpus maintainers be more friendly to non-native contributors and help them to make their contribution trustworthed.
I support the idea of creation a collection of "perfect" sentences, but I wouldn't call them "perfect", but "standard". It means, they are:
1) grammatically correct according to the current grammar rules of a language (so can be used both in everyday speech as well for school and university purposes without any risk of getting lower marks);
2) "natural" in the sence that lots of native speakers actually use them, not only few individuals;
3) Wide spreaded and understandable everywhere (not dialects);
4) Stilistically elegant.
It means, the sentences that not accord to those rules aren't nesessary "bad", but those that accord are more universal and valuable for most learners.
In my opinion, it must be a kind of consensus, you can mark a sentence like "standard", and if for example three native advanced contributors voted "It's standart", so it's go to a special list, you can browse it, and if any native advanced contributor vote "it's not standart", so we can consider the sentence as doubtful and don't include it to the "standard" collection. In no native advanced contributor voted against it for two weeks, it goes to the "standard" collection. I include only "advanced" contributors here to avoid trolls and fake accounts, that might vote "for" or "against".
Good contribuion. Sounds just like Moscow "official" accent from Soviet movies.
About "skipped" sentences I mean if one doesn't want to rate it, and doesn't want to see in their "unrated" list.
1. Ones, that are grammatically correct, but sound odd or too akward.
2. Yes, I skip a sentence if I don't like how it sounds, but it's not so odd to be marked as "unsure". Also I skip the sentences, which have complicated punctuation, so I can't estimate, if they correct or not.
3. If they are grammatically correct and don't seem akward, so I'll rate them OK, even if I wouldn't use them by myself because of my personal preferences (for example, some of their words aren't in my active vocabulary, but only in my passive vocabulary), but I'm sure some other people use them.
Suggestions: As well we have the features like "favourites", "lists" and "collections", that have many things in common (but not all), there should be a possibility to add a sentence in "favourites" and "collection" by one single click.
I think it should be "add to favourites and marked OK", "add to favourites, but not mark" (I like the sentence and need it for my personal usage, but I'm, for example, not all sure about orphografy or punctuation), "don't add to favourites, but mark OK (the sentence is OK, but I don't need it for my personal usage). Also there should be a possibility to see all the sentences in any language, you haven't already marked, and the "skip" button (if I don't mark the sentence "OK", "not OK" nor "unsure" (other people can't see "skipped" sentences).
Thanks, I didn't know there was such a possibility. So, it's possible to keep all the lists in a text file and paste them in the search line.
Yes, the idea is about creating groups of contributors.
But it's even not about "trustworthed" or "untrustworthed" contributors.
The option "only native speakers sentences" is already exists, and it works well.
But personally for me nativeness is not so important, cause almost all people I interact in English aren't natives, and I consider English as a kind of international language (of course, I talk about advanced level contributors, not about beginners, which contaminate the corpus with weird and grammatical incorrect sentences). For Italian, Guybrush 88 makes a good job to keep the corpus clean and check and tag non-native sentences, so they can be considered trustworthed enough in my opinion.
There is another issue.
Firstly, many languages are spoken in more than one country. So, the ideal case would be if each sentence is marked with tags which dialect is it (if the sentence fits for several dialects, there should be several tags). But I think, it's impossible for now.
So, as I try to stick British English rules in spelling, vocabulary and pronunciation, I would like at least to have a list of BrE speakers.
Secondly, my personal preference is translating sentences, that express some thouths and feelings, tell some interesting facts, or just not trivial. Translating those phrases is much more fun, but not all the contributors create something like that.
I tried to describe the whole idea and my personal preferences, of course, every member could use that feature for their own purpose. The main thing, those lists should be visible only to their owners, and serve for their commodity, and if someone include you into their "white" or "black" list it shouldn't impact on your rating or "trustworthness".
I suggest to make it possible for every user to create their own private lists of contributors and display the sentences, added only by the contributors from the certain list (now we can only browse the sentences, added by one single contributor, but not by the group of contributors). For example, "Native British English contributors", "The German contributors, I like the most", and so on. Also it would be possible to display all the sentences in the language, excluding those, added by contributors from the certain "black" list (like "Chinese contributors, which style I don't like"). I suppose those lists only for private personal usage, so it's not any kinds of ratings, but the feature for user's commodity.
If a person has an email notification enabled, they often appear even after having been absend for a few month, if you comment their sentences. If the option is disabled,that's unlikely they will appear.
There is. If a person leave their homeland as a child, and not master their mother tongue anymore, but only use it for talking to their parents, his or her vocabulary remains rather limited and he or she tends to use unnatural phrases, due to a strong impact of the second language. But he or she is still a native speaker. That's the case of "native speaker with low level".
All the sentences, owned by inactive users should be released. Else there is no chance to make them trustworthed sooner or later.
Just native? In that case it should be "native monolingual speaker".
For bilingual people the concept of a "native language" might be very vague, and some of them actually tend to use cultural or political reasons, chosing their native languages. So, if we don't want those battles here we'd better be ruled by the criteria of actual language skills and give corpus maintainers the right to decrise the level of those fake "natives", if their contribution is not good enough for this level.
I don't know what do you mean as "native level". If you're a native, it doesn't mean you never make any mistakes either. Do you mean "near native level"?