Keegan Peterzell via Commons-l, 23/02/21 22:36:
> Special:Search also functions at its best in English, and is not
> necessarily friendly to other languages.
This is not true. I think you meant to say that the text in Commons is
mostly English, while the new structured search is expected to be able
to leverage more multilingual content.
What's the rationale for changing the search bar autocompletion from the
title prefixes to the structured search? The non-prefix search has
caused much trouble in Wikidata for years. What kind of use cases were
considered for this part of the matter? (I know you can disable that in
the preferences.)
Federico
Thomas Tanon, 06/10/2014 11:00:
> Thanks a lot Federico for this email and sorry for the late reply.
>
> I'll try to get some time to see how METS works on the next few days.
>
> If this standard is compatible with ProofreadPage assumptions it would be very nice to implement in the extension an API that would allow to do import/export in this format.
Great, it will be an interesting discovery whatever the outcome. Please
let me know if I can help in some way, e.g. with an example file in METS
format from BEIC (they're not publicly exposed).
How to "speak METS" is already a problem for me/us because BEIC would
agree to publish (somewhere on Wikimedia projects) their METS for
hundreds books, but we don't know how to store them. :)
Nemo
«Received 36 books today. Totally thrilled with the selection. I’m
paying $10 or even $20 for books that sell for 150 rupees, but many of
these are old and I’d never be able to find them except on the market.
Going to have the definitive set of high-quality scans of GOI books.»
«5 scrapes of sites are going well. State library crawl of 1903-2014
southern gazettes is about to hit year 2000, the crawl of 23k books from
Bengal half-way done (so 3 more months), the download of videos up to
700 films out of 5k, about 200 gbytes so far.»
https://twitter.com/carlmalamud/status/989975678207455232https://twitter.com/carlmalamud/status/989974421476491264https://twitter.com/carlmalamud/status/990096301361545216
Federico
Nicolas VIGNERON, 28/04/2018 18:19:
> Can we have the context here? I'm not sure to understand what this is
> about...
Sorry, I thought you'd know him:
https://en.wikipedia.org/wiki/Carl_Malamud
Many of us know him for his work around PACER. Looks like his archival
projects are expanding to India. He seemingly gets Indian documents and
then gets them scanned, some from the Internet Archive I think.
Federico
Klaus Graf, 25/03/20 18:18:
> Where is the proof that there was a hiatus?
I doubt anyone has proof of anything with Google Books. But many had
written about it, see for instance:
<https://www.newyorker.com/business/currency/what-ever-happened-to-google-bo…>
> [...]
> And so on.
Yes, they're all examples from 2019 aren't they? I didn't mean to say
Antwerp was the first, sorry.
Federico
Deryck Chan, 30/08/2018 16:30:
> without any prior notice on their user talk page nor any known process
> on how to regain that right.
This is more of a problem on wikis without bureaucrats, or with limited
'crat bandwidth. I agree that such wikis may deserve additional help,
including direct talk page messages and a backup avenue for flagging by
stewards.
I would be surprised if the English Wikipedia needed to be included in
such a class of needy wikis. But yes, restructuring user rights often
has unintended consequences, including an increase in instruction creep
and a decrease in overall efficiency of the community. (We've had
multiple examples in the past... from the English Wikipedia itself.)
Federico
Benoît Evellin (Trizek), 10/05/19 14:34:
> I would suggest you to bookmark the recent imports link on Translatewiki
> <https://translatewiki.net/w/i.php?hidehumans=1&translations=only&namespace=…>,
> and check it time to time to catch up on new messages for MediaWiki.
The usual way to follow new messages to translate is from within
Translate itself, with the "new additions" group:
https://translatewiki.net/wiki/Special:Translate/!additions
This does not seem to be a problem of slow translation, though. If
Raymond's link is correct, the translation was made 5 days ago. It's an
unacceptable regression to have to wait so many days before a
translation is deployed. We need to restore the usual functionality
we've had for something like a decade, or this little unnecessary shocks
to translators will continue.
Federico