Last modified: 2012-02-24 02:51:57 UTC
While visiting Haifa, I set my locale to English and ran "nearby". The phone showed a marker for [[Technion]]. The summary showed the literal word "ndash" instead of the en-dash character.
Here's another interesting detail: The actual article says "The Technion – Israel Institute of Technology". It has an actual en-dash character and not the HTML entity –. So i suppose that the software converts it to an entity at some point.
Created attachment 9854 [details] screenshot on galaxy tab 10.1
It looks like the extract isn't up to date; the opening paragraph has been rewritten since. Older versions like https://en.wikipedia.org/w/index.php?title=Technion&action=edit&oldid=362964628 appear to include the – as a character reference, so it's probably got one of those old versions. I think the extract comes from the geonames lookup service...
There is noting we can do about this - except replace Geonames with our internal APIs.
Shouldn't we leave it open as an upstream bug?
I think this bug is specific to the article in the example. We get articles in delayed fashion via Geonames, which is an issue though not a bug, but that will change soon.