@BorjaBalle Especially if every journal does that, the (hoped for) effect will neutralize…
#Wikipedia Live Monitor is up again.
Known issue: #Wikipedia Live Monitor is currently down. Apologies for the inconveniences, should be back soon.
Fixed a tiny bug in #Wikipedia Live Monitor where humane dates would be displayed as “undefined” in non-Chrome browsers. Should work now!?
~RT @GuardianJoanna: […] Loving [Wikipedia] Live Monitor! Amazing tool. Great for newsrooms. // Thanks. What could we do better?
@bkeegan Thanks for this great article. I was wondering if you were interested to team up for the next iteration of Wikipedia Live Monitor?!
@francisybea Planned: EDIT classification & categorization to have more precise (& even less error-prone) signals. Also live page VIEW logs.
@mhausenblas Thanks :-) On an unrelated note, I’ve just sent you an email…
@mparramon English on demand. There is demand tonight :-)
@LeaVerou I made similar experiences w/ links (<a>) that you can click() in Chrome w/o appending them to the DOM, but nowhere else.
@lechatpito LOL, I always used bahn.de while I lived in France. I tried hard to use sncf.fr, but it was just impossible…
@WikiWatchBlog Idee ist via wohldef. Kriterien (Paper) zu entscheiden, ob zeitnahe Edits am selben Artikelcluster ein News-Event darstellen.
@nicolastorzec Strong +1 to that. From my experience, you hit the dereferencing issue w/ way smaller datasets already. Fundamental issue…
@abursuc Thanks. I have quality as a ranking factor. In the concrete case, though, it was outweighed by the image’s social signals.
@rtroncy Shaky because shot with my iPhone filming my screen :-)
@sympapadopoulos A crawler? Then your base isn’t the Web browser, but whatever language your server speaks… That’s probably a good thing :-)
@RubenVerborgh World-peace would be nice, too… I’m dramatizing, sorry. <video> is actually quite clever w/ just requesting needed parts.
@sympapadopoulos In HTML5 video, you don’t have access to low-level codec-specific things like keyframes. There’s no notion at all about it.
@sympapadopoulos Supposing this is for <video>. Then keyframes no, fragments yes. Just seek to the position and wait for the “seeked” event.
@PFCdgayo Just sent you an email…
@PFCdgayo Wow’ed by the paper’s outcome. This indeed casts shadows on any & all results of streaming API analysis where filters are applied.
@PFCdgayo Aware of applied filters. Yet, sample seems smallish. Let’s assume completeness, tho. Then enormous!! boosting of matching tweets.
@PFCdgayo Also 528592 / 1280344 * 100 ≈ 41%, which is >> 1%, i.e., the supposed relation of streaming vs. firehose. What do I misunderstand?
@PFCdgayo Thanks for sharing. Data collected 2011-12-14—2012-01-10. Only 528592 tweets Streaming; 1280344 Firehose. Understand why so few?
@rtroncy Some 404’ing project icons, vacancies link broken. Besides that, nice update.