Wikipedia talk:Wikipedia Signpost/2013-04-29/Recent research

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

Discuss this story

Font size[edit]

According to the above, four researchers from Barcelona "recommend using 18-point font size when designing web text for readers with dyslexia". Can't dyslexics "zoom" their displays, like the rest of us do? --Orlady (talk) 15:08, 2 May 2013 (UTC)[reply]

Better yet, they should be able to use a font that is specifically designed for those with dyslexia. -- John Broughton (♫♫) 19:38, 2 May 2013 (UTC)[reply]

Opinionated[edit]

As usual, good work summarizing briefly a number of interesting activities. However, a minor grammar error brought me to a hightened state of alert, which made me notice the poorer quality of the next "Mining content removed" item. It's too long for an "in brief" bullet point, because the reviewer spends too many words pointing out what's wrong with the reviewed work. Other than that, it's a well written page, rewarding the usual wait for our overdue Signpost. Jim.henderson (talk) 12:25, 3 May 2013 (UTC)[reply]

I was the reviewer. Thanks for the feedback; this is my first research review for Signpost and I'm still getting the hang of the genre conventions. - J-Mo Talk to Me Email Me 19:07, 3 May 2013 (UTC)[reply]

Usability study[edit]

To increase redability, I recommend a line width of 120 characters. You can do that with this code, copy it on your userspace. Have fun! --NaBUru38 (talk) 17:32, 3 May 2013 (UTC)[reply]

Provenance[edit]

The provenance of Wikipedia articles is per-character, but the W3C PROV descriptor is per-document, isn't it? 116.233.70.143 (talk) 01:00, 5 May 2013 (UTC)[reply]

As said in the brief summary, this tools appears to be based on "metadata from Wikipedia revision history and user contribution pages (e.g. the author of a particular revision, or articles edited by an editor)" - this metadata provided by MediaWiki is per-revision, it does not include information tracking the authorship of particular parts of text. Some external tools like WikiBlame or WikiTrust do this, and there is currently a proposed Google Summer of Code project to integrate an optimized and streamlined version of the WikiTrust algorithm (only the authorship tracking part) into Wikipedia - we'll probably cover the accompanying WWW'2013 conference paper in the next issue. Regards, Tbayer (WMF) (talk) 18:38, 5 May 2013 (UTC)[reply]