Mark Williamson wrote:
1) Videos take up heaps more space than audio. A very
short sample
article (the English version was only 4 or 5 sentences long, a total
of 88 words compared to the average 348.6) turned into a 1.5 minute,
16MB video. Now, imagine a much longer article -- the average article
on enwiki is 348.6 words long. Now, multiply that by 100000. Then, add
to that the fact that there are over 100 signed languages in
widespread use in the world today, and you get a perfect nightmare.
That ends up with over 600TB of video... an astronomical amount of
space, not to mention the bandwidth costs. Angela said that space and
bandwidth aren't a concern -- I sincerely doubt that anybody has that
much space to donate to such a project. And that's just the first
hundred-thousand articles, at the level plwiki is today.
Mark there are people that have a terabyte in their home computer. Also
you make it seem that all these articles will miraculously apear. I am
sure that it will take a long time before we have 10.000 articles in ANY
of the signed languages. To get there, we need people who are seriously
committed to the project. So the problem is not what you make it seem.
2) It's my opinion that if Wikipedia were to be
based more on spoken
than on written materials, we should rely more on speech synthesis. Do
some people have problems with speech synthesis? Yes. Is some software
of better quality than other? Yes. Can some software be difficult to
understand if you're not listening very carefully? Yes. But it's
certainly better, I think, than having to re-record an article every
time a revision is made.
It is your opinion, yes. But why not have the deaf decide it for
themselves. We can start a project that does not have the "ideal"
infrastructure. We have done that before with Commons and we can do it
again for signed languages. When you make the video for an article in
multiple parts, eg a paragraph at a time, the article is modularised and
only those paragraphs are replaced that need replacing. I think there
are many ways of making it work more easy. I am not going to learn
signing, but I will also not make the road to the realisation of a
signed wiki unnecessary complicated by insisting on things I am not
aware off or that are not for me to decide.
3) What about unwritten spoken languages? Should they
get their own
Wikipedias, or should they just be uploaded as spoken translations of
enwiki articles? It doesn't make sense to me to limit ASL to videos
uploaded to enwiki.
Well Mark, let us cross this bridge when we meet it. The difference is
that people ARE asking for ASL.
Mark
On 16/09/05, Timwi <timwi(a)gmx.net> wrote:
Kelly Martin wrote:
Then it sounds to me like ASL is not yet mature
enough, as a language,
to merit a Wikipedia of its own. If ASL signers are not willing to
accept synthetic signing, then I suggest that they need to adjust
their attitudes (or else improve the quality of sign synthesis
software).
Interestingly, though, English speakers are usually not willing to
accept synthetic speech when they can have real speech from a real human...
It's not practical to record hundreds of
thousands of
videos to facilitate access for a relatively small community.
... and yet, the [[Wikipedia:WikiProject Spoken Wikipedia]] aims to
record hundreds of thousands of audio files to facilitate access for a
relatively small community (the blind).
Personally I think videos of ASL translations of Wikipedia articles
should just be a WikiProject like Spoken Wikipedia. Just like our sound
files, the video files would be a translation of a particular revision.
When the file becomes out of date due to heavy editing of the article,
re-recording should be considered, but since we haven't got very far
yet, we are concentrating on recording new articles first.
Timwi
_______________________________________________
Wikipedia-l mailing list
Wikipedia-l(a)Wikimedia.org
http://mail.wikipedia.org/mailman/listinfo/wikipedia-l