{{}} No language found.
{{}} No language found.

sabretou's messages on the Wall (total 192)

2018-11-27 13:20
That the Wall appears to be dominated by North African languages is not really the users' fault as much as a technical limitation of the Wall system.

I think a switch to a more conventional forum/discussion board interface has been a long time coming. Being able to visit only the threads you want to see and ignore the rest will be a massive step-up as far communication on this site goes.

The problem of a debate on the Wall drowning out other discussion is a problem that's only going to get worse as Tatoeba picks up more contributors.
2018-11-20 07:03
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.
I agree. If a language learner were to chance upon this site and look for Mozabite, I think it is likely they will look up the language list, see that the language is not there, and go away.

For a user trying to make use of those sentences, they will have to first be aware of the status of Berber on Tatoeba, then they will have to know about tagging on Tatoeba, then they will have to know that Mozabite sentences have been tagged within Tatoeba, and finally, they will have to understand how to use the appropriate search tools to find Mozabite sentences.

Granted this seems like a technological issue, but I think that if an ISO code for the language exists, we should use it. Using our own definitions of dialects and languages could open a can of worms.
2018-11-18 12:43
I understand. Still, we follow ISO conventions for consistency here on Tatoeba, so we will have to add these creoles separately until and unless ISO decides to combine them.
2018-11-18 07:40
Wikipedia lists three different ISO codes for Antillean Creole variants:

gcf – Guadeloupean Creole
acf – Saint Lucian / Dominican Creole
scf – San Miguel Creole French (Panama)

Perhaps theachaean can clarify which variant they are contributing in.
2018-11-06 13:25
I can confirm that I'm having auto-detect being set to default every time I open a new page on Tatoeba. That is to say, the contribution language is not being remembered.

Windows 10 64-bit
Waterfox 56.2.5 (64-bit)
2018-10-09 15:51
Use No.1, it's cleaner and has better contrast.
2018-10-09 12:58 - 2018-10-09 12:58
I've resized the flags to match Tatoeba's format. Here's how they look:


I think 1 looks the clearest.
2018-10-08 13:49 - 2018-10-08 13:53
Malayalam: I have a suspicion that the flag represents the seal of Travancore, as opposed to the state of Kerala. They're very similar, so it's hard to tell.

Mohawk: This flag represents an organisation, The Mohawk Warrior Society. It is possible that not all Mohawk contributors will accept this flag.

Tamil: The last I checked, years ago, this was the flag of some informal international organisation promoting the Tamil language. It doesn't seem ideal, particularly given that (to my observation), Tamil flags typically involve red, black, and white.

Telugu: The present flag is plain, and the font is squashed. While we could create a non-squashed version of the flag, perhaps we can also wait for a native speaker to suggest any appropriate colours or symbols.

Like I said, each of these cases would have to be discussed with native speakers, as I don't think I'm entirely qualified to comment on these.
2018-10-08 08:14
The icons for Amharic and Hawaiian look a little squashed at present, so I've created updated versions:

The icon for Guarani was assymetrical, so I've centred it:

Also, keeping with the updates to Latin and Ancient Greek, I think the icon for Sanskrit could be beautified as well. I don't have any ideas on how to proceed with that, however.

The icons for Malayalam, Mohawk and Tamil *may* need to be changed, as I suspect they may not be properly representative. Separately, the icon of Telugu could perhaps be beautified. I have been waiting for active native speakers of these languages to discuss this with them.
2018-10-07 03:03
1) I typically go straight to a custom URL advanced search that's saved in my browser history. I almost never use the regular search.

2) I use the same custom URL for most of my searches (random sentences not translated to Marathi), but sometimes I use specific words.

3) It wouldn't make a difference in my specific case, because I would probably continue to use the custom URL (it's only a couple of keys away for me). But if such a thing were to be implemented, it would be more convenient for me in case I lose the custom URL for some reason, or want a quicker way back.
2018-10-06 01:53
That may be because the user who contributed that sentence used Auto-Detect, and the sentence got assigned to Algerian Arabic. The user must have noticed this and corrected the language appropriately.

You will notice a similar situation with haziqmir's sentences under Urdu:
2018-09-25 13:27
There was no exemption, as that was from before we enforced the ISO 639-3 policy. You will notice we also have some constructed languages that are not covered under the ISO 639-3 specification.
2018-09-25 13:06
New languages added to Tatoeba must be listed in ISO 639-3.

'sem' is not listed in ISO 639-3, so it doesn't look like we can add this.

2018-09-21 12:44
I think it would be best to have a plain white flag with the ISO code on it, similar to how we dealt with Kashmiri and its two competing flags.
2018-09-13 06:13
This is pretty great, and works perfectly for me on Windows 10, Waterfox 56.2.2.

A few thoughts:

- Since the entries are so large now, perhaps language flags can be incorporated as well? This may or may not lead to quicker identification for some people.

- Can a spellcheck be built into the system? I can foresee some people mispelling or using unconventional spellings for languages. For instance, if someone types 'Panjabi', it would be ideal if the results for 'Punjabi' are displayed nonetheless.

- Alternately, perhaps we can set 'hidden' alternate names for certain languages, to account for those alternate spellings. For instance, a lot of people know Persian as 'Farsi', but typing in Farsi shows no results.

- The present light grey highlight for matching text is hard to see. I would recommend using a different colour, maybe the default green that Tatoeba uses.
2018-08-14 13:51
वाढदिवसाच्या हार्दिक शुभेच्छा, लिसा!
2018-08-12 20:43
The accents are rendering correctly for me as well.
2018-08-05 17:13
I'm in favour of #2 for Latin and #2 for Ancient Greek.
2018-06-09 13:23
Are there any emoji that are exclusive to a particular language? Is this something that a language learner would have to learn while picking up a language? I'm no expert, but I do imagine that ":-D" is universal, and can be used in English, German, French, Russian, and so on. It can theoretically be used even in Latin or Sanskrit. I think emojis are independent of languages.

Allowing emojis on Tatoeba could lead to a lot of pointless, extraneous sentences that add nothing to the corpus. Imagine:

Tom went to the market.
Tom went to the market. :)
Tom went to the market. :(
Tom went to the market. :D
Tom went to the market. :'(

And so on. Is there any language where these emoji-laced sentences will be translated differently than the base sentence? Conversely, is there any language that forbids emoji explicitly?

Emoji are a part of culture, and might even be considered a language unto their own. But I do not think they are a part of any existing language, and so they serve us no purpose on Tatoeba.

Acronyms and abbreviations are a different story. I think modern acronyms like 'irl', 'brb', 'cya', etc. all have a place here.
2018-05-20 17:29