Yeah, but then I read the Terms of Use and found a section that seemed to contradict my understanding of the situation. I thought that something may have changed since I was away. This is the bit in question:

***********

Information for re-users:

You can re-use text content from Tatoeba project freely, with the exception of content that is used under "fair use" exemptions, or similar exemptions of copyright law

***********

That implies that Tatoeba contains text under 'fair use' exemptions.

Although it may be best to discourage excessive use of copyright material - fair use or not. I know that commercial (for profit) corpi (corpuses?) are usually much more careful about that sort of thing.

http://tatoeba.org/eng/sentences/show/1895984

Maybe examples like this should be tagged "Text used under Fair Use Exemption".

As far as I know the legal situation with fair use and copyright varies depending on country and is not particularly clear cut but taking small extracts for non-profit educational use and making their original source clear should be OK.

Long contiguous quotes would not be a good idea. Also song lyrics are particularly thoroughly protected and probably shouldn't be used. (I recall a case where an author chose to use a single line from a different song for each chapter title in her book and had to pay for every song used).

Yeah, kind of back. I don't know how long I'll be around because it depends on my health and stuff.

I'll give you a bit more detail first.

Here's a screenshot:
imageshack.us/a/img210/3470/screenshottxs.png

At the top you see the *previous* question (example sentence, translation, BIG kanji with reading and stroke order / stroke count).

Below you see the *current* question with part of the sentence highlighted in red.

In order to answer the current question you have to
1. Think about the kanji and picture it in your mind (or scribble it on a bit of paper).
2. Decide how many strokes it has.
3. Enter that at the bottom left where it says 'Stroke Count'.
4. Select the correct kanji from the ones shown at the bottom.
5. Click 'Check Answer'.

If you can't do that then instead click 'Don't Know' and you will be shown the answer.

I'll need your email address to send you the program - you can send me that in a PM to me.

NOTE: For the stroke order to be shown you need to download and install the Stroke Order Font from
www.nihilist.org.uk

Thanks for showing an interest. :-)

Anybody want to beta-test a program that helps you learn the Kanji for Japanese words? Post a reply if you do.

Unfortunately you need to have Excel 2007 (or a more recent version).

Not dead yet. :P

Hi everyone. Just a quick post to say I'm going to see if I can get back to doing some work here again. I've got some ongoing health issues and I pretty much dropped off the internet for a long time and I'm still not particularly well. I haven't been checking my PMs, so sorry if you've sent something to me and didn't get a reply. If you need to get in contact with me please try again.

> Quality cannot be measured by speed.

I agree 100% with that. Somebody might see something and a brilliant translation come to them in a flash, or they might puzzle over the exact nuance of a phrase for days before coming to their final version.

They display fine for me. (Firefox, Windows 7)

Looks like it's working now.

Tag auto-completion script turned off?

It looks like the auto-completion script when entering tags on sentences seems to have gone. I'd actually got used to it as well.

> I'm not sure the "could equally use..." as, while
> grammatically equally correct, they aren't completely
> situationally equivalent.

The thing is that I see nothing wrong with "Mother bought me a bike for my birthday." So I don't think it should have a tag that essentially says it should be 'My mother' there.

There are situational differences - for example the sentence above would, IMO, be more likely when the speaker is talking to his brother or sister - but there's nothing about the sentence to say which situation applies.

More tag changes:

I don't like:
[more natural if change to "my mother"]

Because I disagree.

However I would say
[could equally use "my mother"]
for those sentences.

The
[possible copyright violation]
tag should have an @ as it is for admin / moderators to decide if these _do_ violate copyright. (And delete them if needed).

^ On the same basis as 1. above 'delete' tags should be globally renamed to '@delete'.

Tag change request.

I don't like the following tag:
"not a sentence, should be deleted"

Reasons for disliking it:

1. It's aimed at moderators (because only they can delete sentences) but it does not start with @.

2. It combines the factual (not a sentence) with a value judgement (should be deleted). It has been established that certain kinds of non-sentence are of value to Tatoeba (although they should be annotated appropriately).

I think they should all be changed to "non-sentence" (if non-sentence is not already a tag for them).

They could also all have a @delete tag added. However, there is something of a 'delete crusade' going on at the moment. Sentences are being tagged, on a stricter basis than I would like, much faster than they can be properly evaluated and dealt with. The large '@delete backlog' encourages snap judgements that often don't do the examples justice.

I had an idea last night.

I could (or you could ;-) put up lists of pre-generated 'add link' links to join translations of identical sentences to each duplicate.

e.g. Take the case where you have

A -- B

and

P -- Q

where B and Q are identical. This idea would have the pre-generated html links to add a link between A and Q as
well as between P and B.

Trusted users and moderators could then click down the list.

Notes of caution:

* Clicking a whole lot of links quickly might slow down or even crash Tatoeba! So we should probably advice taking it fairly slowly.

* You don't want everybody doing the same links. They should probably be numbered and people can volunteer to take on a range of a few hundred so the same links don't get clicked more than once.

Member's page display.

I think if you drop the 'hours:minutes' from the 'Member since:' line the page will display better. (Depending on fonts and browser settings) the hours and minutes wrap over to the next line and make it look untidy.

Slight glitch on 'comment' link from front page.

When you click the magnifier glass icon on a comment to 'view comment' it doesn't work on sentences that have been deleted. Instead an unrelated sentence that hasn't been deleted is shown.

The big thing is dealing with the sentences linked to duplicated sentences.

e.g. Sentence A is a duplicate of sentence B.

P-+
...\
Q---A

R---B
.../
S-+

Before deleting 'B' you should link R to A and S to A.

P-+
...\
Q---+
.....\
R-----A
...../
S---+
.../
T-+

Removing duplicate sentences.

As I understand it there are problems with the duplicate removal script at present. As an interim measure I suggest the following:

1. A script that links all sentences that are identical and in the same language.
2. Making a list of all sentences that have identical sentences in the database.

That would make it easier for people to manually remove duplicates.