Replying to @invisiblecomma and @thePeerJ
Wicked, thanks! Great support :-)
February 2016
241 tweets
Replying to @RenaudCamus
Bach, c'est celui qui composait pour Maurane c'est ça? Je ne savais pas qu'il y avait une suite! @ConquetFC @Lesmatinsfcult
♥ 1
Replying to @invisiblecomma
And I’ve now tried it: it’s a massive improvement, thanks! @thePeerJ
♥ 1
«The Buried Giant», K. Ishiguro. Not as bitterness-inducing as previous ones but a neat post-Arthurian variant on his usual memory narrative
awesome-js-drama:
a collection of JS topics with nothing but drama, drama, drama :) github.com/scottcorgan/aw…
↺ 1
Is there a trick to reading the @FiveThirtyEight feed without it reloading back to the top of the page after each paragraph and a half?
Replying to @anablzk and @osezana
Note that the Clinton-Sanders difference keeps dropping with every reporting increase. Quite the nail-biter.
Clinton at 50% and dropping, the suspense is killing me. 0.7% ahead of Sanders.
Replying to @janl
You’d be surprised… when we built the journal style guides DB, we found a *lot* claiming to accept floppies :) styleguides.science.ai/?limit=25&q=fl…
♥ 1
Replying to @janl
It gets more depressing: …99d-1e911c0a6431-bluemix.cloudant.com/styleguides/_d…
♥ 1
Our #DCIPBoston article: «Web-First Data Citations» scienceai.github.io/data-citations/
♥ 4
Replying to @rdhyee
Yes, same people and same principles, but explained in greater detail. Scholarly HTML will get an update soon.
♥ 1
#DCIPBoston @dalapeyre talks about <source> to describe structured data relationships in JATS, I wonder how it compares to schema:hasPart.
♥ 1
Content negotiation has already been mentioned a few times at #DCIPBoston. Maybe a short article on why it’s a terrible idea would help?
♥ 4
Replying to @mrgunn
It is largely covered in scienceai.github.io/data-citations…, but I’d be happy to extract that into something standalone.
#DCIPBoston Unsure if it correlates to JATS usage, but the number of style guides mentioning «data citation» is low: styleguides.science.ai/?limit=25&q=%2…
♥ 2
Replying to @domenic
Sweet, thanks!
Replying to @domenic
That’s good, but very browser-centric. The case here is more about automated agents. I have some bits in scienceai.github.io/data-citations…
♥ 2
Replying to @mrgunn
There is also wiki.whatwg.org/wiki/Why_not_c…, but it’s very browser-oriented.
Replying to @rdeltour
I think all forms of conneg are bad, really :)
Replying to @rdeltour
Oh, linking to two different things is not conneg — it’s just linking.
Replying to @rdeltour
Conneg is whenever the protocol does it for you.
Replying to @rdeltour
Oh, you mean multiple choices. Gotcha. That still suffers from the fact that it tends to hide data.
Replying to @rdeltour
Unless you expect to trigger and receive a 300, you won’t ever see the alternatives. It’s also more complex on the server.
Replying to @adam42smith
♥ 1
@danbri Do you know if your colleagues from Scholar might consider relying on some better metadata practices than scholar.google.com/intl/en/schola… ?
♥ 1
Replying to @dalapeyre
Indeed, but it’s still relatively small. Then again, dozens of style guides suggest submitting on floppy disks… #DCIPBoston
It’s quite depressing to see <meta> elements being the state of the art for data landing pages (or articles). #DCIPBoston
♥ 2
Replying to @null and @amyvdh
hahahaha :)
#DCIPBoston The term that HTML uses for pragmatic transitional features like @mfenner’s <meta> could be a model? w3.org/TR/html5/obsol…
Replying to @mrgunn
I’ve asked here to see if there may be a way in
♥ 1
Replying to @pgroth
I would consider them problematic: scienceai.github.io/data-citations… #DCIPBoston
Replying to @pgroth
Absolutely, but if you have the data, what is the use case? Don’t you risk bugs for no extra value?
Replying to @pgroth
We can’t yet back that up with data, but experience in other domains certainly suggests so strongly.
Replying to @pgroth
Well, we haven’t published enough articles with schema.org, so we don’t have stats basically :) @danbri
Replying to @pgroth
Oh sure, but that’s a hard problem. Bidirectional links have been a problem since at least SGML :)
I’ve always been ahead of the style curve nytimes.com/2016/02/04/fas…
Replying to @danbri and @mrgunn, @pgroth, @dalapeyre, @mendeley_com
Happy to help put that together.
♥ 1
#DCIPBoston talk of software citations. That’s package.json, right? ;-)
↺ 1
#DCIPBoston Just for the record: Leibniz > Voltaire. :)
♥ 3
Replying to @null and @amyvdh
Likewise! I wish we did that more often!
Replying to @jaffathecake
I need a recording of @jaffathecake saying that that I can play at meetings.
Quoting a tweet by @jaffathecake ↗
Replying to @fielding and @rdeltour, @mnot
Oh I've learnt it, I just use the terminology in use outside the http WG.
♥ 1
Replying to @invisiblecomma
You can also do that with links!
Replying to @invisiblecomma
But in your scheme the user has to know a fair bit more than the url, no? Do you send a 300 for discovery?
Replying to @invisiblecomma
And broken crawlers. So it's not discoverable, not a big deal but it annoys me :-)
Replying to @invisiblecomma
You put all those headers on every response?!
Replying to @invisiblecomma
You send html on no match, right? (I'd test but I'm mobile.) And html has links to alternates?
Replying to @invisiblecomma
Right, so you're really using conneg as protocolar sugar. No problem with that since the system is resilient to its absence.
Replying to @invisiblecomma
I'd use HTML for that, but you probably would've guessed as much :-)
Replying to @fielding
Or even better, just say «conneg», it’s a great word that I’m told means pretty much anything.
♥ 2
Replying to @fielding
I have read the thread, and I make no claims as to whether you like conneg or not.
Replying to @fielding
All I say is that I have seen nothing to convince me there is value in the all-encompassing, non-colloquial meaning of conneg.
Replying to @null and @sgalineau
oh yeah. Oh yeah.
Replying to @tabatkins
That’s what the «nitprick» we were talking about yesterday is for.
Replying to @null and @sgalineau
Yay, thanks :) @dauwhe
Replying to @null and @nitot
It’s funny, you wanted to type «sexy» but it came out as «silly». @sgalineau
♥ 2
Is there good introductory/review reading on Wheeler's pregeometry work?
♥ 1
Replying to @null and @ndw, @fielding
I shall make an effort to Accept-Language: en-httpwg ;-)
Has someone written down a more or less comprehensive list of CSS methodologies (BEM, OOCSS, etc.)?
Replying to @blahah404
Wicked, ta.
C'est marrant, à $50 le reblo la tartiflette perd un peu de son goût populaire.
Replying to @Nature_PlCTURES and @adamatzky
Can you get one of those as a pet? 💕
Replying to @lhirlimann
La roue. Je ne sais pas combien ça pèse, mais je pense moins qu'un kilo.
Replying to @null and @dauwhe
Which one? I haven't tried it yet, I have Zee's. Does it cover that sort of stuff?
Replying to @null and @dauwhe
Yes, I know the book, I was asking which author signed it :-) Man... I'm going to have to buy it now.
Replying to @null and @dauwhe
I wish I could find the pic I RTed the other day about how crazy I go buying books... I'm not sure it's safe for me to talk to you.
Replying to @null and @dauwhe
Hahaha, there are cheaper editions I think :-)
Replying to @null and @dauwhe
Oh, that one has really cheap second hand copies 👧 One of mine is amazon.com/gp/aw/d/019507…
Replying to @LeonieWatson
Hahaha, thanks! I decided to hell with it and I've cooked one for tonight!
Replying to @null and @sgalineau, @dauwhe
Ah, I know that one, it's one of the big biases. It'll come back to me :-)
Is there a good alternative to Apple's built-in parental control? It's nigh unusable.
Replying to @torgo
It's just too painful. I'm wondering if just using Little Snitch wouldn't work better.
@sgalineau *cough* it’s very nice of you to accept PRs, but it’s even nicer if they are followed by npm publish :-D
Replying to @tabatkins
Would you ever not want —follow?
Replying to @LeaVerou and @keithjgrant, @Tevko, @getify
See also twitter.com/SlexAxton/stat… :)
Replying to @null and @sgalineau
Dude, if it’s not on npm, it doesn’t exist.
Replying to @tabatkins
Either way it’s neat, thanks! I’ve immediately aliased it as «lastchange».
♥ 1↺ 1
Replying to @null and @sgalineau
Works like a charm, thanks!
So it looks like Twitter on my phone went crazy and posted hundreds of pics — sorry about that! :(
Replying to @kzwa
I am, thanks. Weird bug!
Replying to @null and @pkrautz
Phone seems to have gone crazy…
Replying to @null and @amyvdh
LOL, you’re so adorable :-*
Replying to @vtlynch
Not really :)
Replying to @rigow
None of that saves you from application-level bugs :)
Replying to @fabi1cazenave
Dur de débugger, mais on il semblerait que l’app se soit juste emballée toute seule…
Replying to @jwajsberg
Bon, je retente ça demain pour voir :) @nhoizey
Replying to @null and @nhoizey
Not half… but quite a few :( Ah well, that’s life!
Replying to @cyberbaloo_ and @glazou
Comme quoi on n’est jamais à l’abri d’un bug à la con :)
Replying to @libgrog
Well the phone ran out of battery so it did stop :)
Replying to @steltenpower
Not a script, just the Twitter app!
Replying to @null and @_paulverbeek
I think it broke itself :)
Replying to @null and @m_strehl
Well, now you know about @PulpLibrarian :)
Replying to @null and @nhoizey
Voilà, les vrais sont restés!
Replying to @cyberbaloo_
Rien! Dur de savoir, on dirait que l’appli est devenue folle. Tél à plat, tout ça… @glazou
Replying to @glazou
Oui, en anglais on parle de «balc update» @pdwn @cyberbaloo_
@tobie It looks like you’re changing everything in w3c.github.io/ambient-light/? But sort of halfway through? We’re using the previous version :)
Replying to @null and @ndw
Reacting to this morning's flow of pics? :-) App just went crazy!
#Montreuil: @RHammadi a donc voté pour #PJLConstit, montrant une fois de plus qu’il est moins à gauche que Marion Maréchal-Le Pen.
↺ 1
Replying to @tobie
No worries, mostly curious. I hacked together something that changes the blueness of the light in dim conditions.
♥ 1
Replying to @blahah404
Replying to @blahah404
Devil’s advocate: getting out of the way can be hard when people just dump Word/PDF at you. @mrgunn @Protohedgehog @McDawg @tpoi
Replying to @blahah404
.docx is a dreadful format to start with, it has virtually no semantics. +/- editable PDF. @mrgunn @Protohedgehog @McDawg @tpoi
Replying to @blahah404
Okay, but that’s not far from saying that the world would be great if we all were nice… @mrgunn @Protohedgehog @McDawg @tpoi
Replying to @blahah404
My point is: there are solutions, but they’re hard. Publishers might just be an asset there. @mrgunn @Protohedgehog @McDawg @tpoi
Replying to @blahah404
Yeah well that’s the @sciencedotai mission right there :) But it’s an intricate problem. @mrgunn @Protohedgehog @McDawg @tpoi
♥ 3
Replying to @blahah404
Oh, I largely agree, but it’s hard to fix publishers without also fixing authors. @mrgunn @Protohedgehog @McDawg @tpoi
♥ 2
Replying to @blahah404
♥ 3
Oh, actually I can’t give you access to the test server just like that… (and I’m not putting that in prod :). @anssik
Replying to @jwajsberg
Ooooh! Want!
Using deep learning in JavaScript for large-scale scientific concept extraction news.ycombinator.com/item?id=110764…
The Clintons’ War on Drugs: When Black Lives Didn’t Matter newrepublic.com/article/129433…
Replying to @jwajsberg
Bah normal, toi tu es de la maison, moi je suis un développeur en détresse!
♥ 27↺ 39
EELV c’est vraiment la fête du slip.
Quoting a tweet by @edasfr ↗
Replying to @null and @dauwhe
Man, now I'm jealous!
Franchement, je vais finir par me présenter en 2017.
♥ 3
Replying to @edasfr
Faudrait que je sois rentré en France… Mais sinon, à ce stade j’ai du mal à croire que je ferais pire.
Replying to @SimonSapin
JAMAIS
Replying to @invisiblecomma
Whoa, what the hell happened there! I'm off work today but I will definitely fix this horror tomorrow, thanks.
Replying to @johnallsopp
I think pop undress would work on me!
Replying to @null and @dauwhe
We have a repository of components called «scholarly-react». That’s exactly what it does :) @glazou @barefootliam
How to make Promises usable gist.github.com/darobin/d8bda9…
↺ 1
Replying to @tabatkins
You’re making Friday JS-trolling a whole lot less enjoyable with your seriousness.
♥ 1
Replying to @xibe
clap-clap-clap-clap-clap-clap-clap-clap-clap-clap-clap-clap-clap-clap-clap-clap-clap-clap-clap-clap-clap-clap-clap-clap-clap-clap-clap
Replying to @invisiblecomma
Should be fixed now — let me know!
Replying to @yoavweiss
Hey that looks cool, what's it?
Replying to @michaelhkay
That looks sweet. Are you interested in early adopter testing? We have some limited but genuine needs.
♥ 1
Replying to @michaelhkay
Right now I'm considering porting some XSLT to JS manually.
Politique
Étape 1: créer des niches fiscales pour soutenir ceci ou cela.
Étape 2: se plaindre que d'autres entreprises s'en servent aussi.
♥ 1
Replying to @Meekostuff
At this stage, mostly running OO math to MathML and MathML to TeX. Pretty simple.
Replying to @tobie
Bah, c'est du SEO, normal. <meta name=keywords content="performance, thune, sexe..."> @PerformanceWeb_
♥ 1
Replying to @stevefaulkner and @powermapper
Really good article, thanks. Does aria-level on th work in any way?
Replying to @powermapper and @stevefaulkner
It's relatively common for scientific content to represent multidimensional data with progressively indented...
Replying to @powermapper and @stevefaulkner
...row headers. It seems natural to convert that to aria-level (which we do), hence the question.
Replying to @powermapper and @stevefaulkner
Gah, that sucks :-(
Replying to @adelaigue
C’est écrit dans le texte :)
Replying to @adelaigue
Disons que l’impression manifestement couleur et plastifiée peut orienter le choix :)
Replying to @stevefaulkner
Yes, that’s what I meant (and what we do). But the visual result we have is on th (and the meaning derives). @powermapper
Replying to @Meekostuff
Oh, absolutely in the browser, but nothing to do with ReSpec. I am no longer a maintainer.
Replying to @gimsieke
Thanks, I was planning to use the former at least (hence the idea to port to JS, since it’s XSLT 2).
Replying to @gimsieke
Re the latter, is it very different from MSFT’s or did you mostly make it more XSLT2-happy?
Replying to @Meekostuff
I do, but it covers neither of those conversions.
Replying to @gimsieke
Hahaha, and then there will be XSLT 4, which is basically JS :)
Replying to @gimsieke
Sweet, ta.
Replying to @gimsieke
More prosaically, I think that a JS lib that did just the core template/apply algorithm would be useful.
Replying to @gimsieke
It could take arbitrary pattern matching and work with arbitrary output. XML is trivial with JSX.
Replying to @gimsieke
Precisely!
Replying to @gimsieke
That said, that tweet was almost a year ago and I see nothing on GitHub :)
♥ 1
Replying to @gimsieke
Oh sure, it wasn’t criticism — just pointing out there’s what there is. @AlainCouthures
Replying to @null and @ndw, @gimsieke
Which means there is so much potential for growth, both individually and as a group :-)
♥ 1
Replying to @AlainCouthures and @gimsieke
Looking forward to it!
Replying to @LeaVerou and @svgeesus, @darenju
Cmd-R works pretty well for me in Atom.
Replying to @hlapp
Replying to @briankardell
It traditionally happens to me on a dance floor. The trick is to drink your way to making it a fashion asset.
♥ 1
Replying to @briankardell
Needs more dancing.
Is there a reason why JATS copyright-year does not take an iso-8601-date attribute? For e.g. <copyright-year>Heisei 28</copyright-year>
♥ 3
@dalapeyre Is there a reason why JATS does not have Ruby elements rtc and rp?
Replying to @null and @ndw, @gimsieke
So you kind of hoped I was kidding. See github.com/scienceai/deja… :)
Replying to @null and @ndw, @gimsieke
It’s used as a JATS -> HTML «style sheet» in github.com/scienceai/deja…. 💖💞💓
Replying to @null and @sgalineau
I know, right? The hat! And they had a big bus and all.
Replying to @gimsieke and @ndw
Well of course, it's a 1h hack. Yet it's already more readily usable in Node and the browser.
Replying to @gimsieke and @ndw
Also, that's not entirely true. The matcher is an arbitrary JS function, as is the template. JATS is only using a tiny bit.
Replying to @gimsieke and @ndw
Finally, it wasn't done to impress but because I need that transformation working now :-)
Replying to @gimsieke and @ndw
There's nothing global about Matcher. It could use XPath, it would be trivial to add given an XPath lib supporting patterns.
Replying to @gimsieke and @ndw
It could also support CSS, or something domain-specific. Walker is independent of it.
Replying to @gimsieke
I must be explaining it wrong because that’s just not the case. I’ll release and document (but not this week).
Replying to @gimsieke
Right, but we could use node.match(css) as a different matcher — that would work today. Priorities, modes: sure, lots!
Replying to @null and @nhoizey
L’image d’incipit. Aussi connue comme «la putain de grosse image en haut, tu sais?» Mais certains trouverons «incipit» plus classe.
♥ 1
Replying to @null and @nhoizey
Enfin méfies-toi, je viens d’inventer l’expression. Mais bon, elle est construite à base de vrais mots élevés au grain.
Replying to @null and @notabene
Trop de la chance, moi aussi j’ai envie de me coucher… @nhoizey
Replying to @pygillier and @nhoizey, @ubiquitic
«C’est l’information la plus importante; d’ailleurs, sur notre site on l’a mise au placard» O:-)
Replying to @ubiquitic
Je saaaais, je fais exprès rien que de pour t’embêter. @pygillier @nhoizey
Replying to @ConquetFI and @ConquetFC
Whoa! Classe!
Replying to @elfpavlik
♥ 1
Replying to @manusporny and @elfpavlik, @rdfjs, @bergi_bergos
We're totally slammed too but I'll put it on the @sciencedotai radar if no one beats us to it.
♥ 1
Want to Change Academic Publishing? Just Say No chronicle.com/article/Want-t…
♥ 13↺ 13
Replying to @babette_babich and @german_sierra
Indeed, but it lacks the review, and the metadata. Whether that matters varies. Hoping to help fix this :-)
♥ 1
Replying to @german_sierra and @babette_babich
It depends, in biomed subject indexing is crucial. Otherwise you just don't find anything in the volume.
♥ 1
Replying to @leonor_de_b
tu voudrais faire quoi?
Replying to @leonor_de_b
Ah oui, nous sommes nombreux comme ça. Malheureusement à mon dernier tour d'horizon je n'avais rien trouvé.
Replying to @leonor_de_b
Il est certain que la gauche politique est morte en France. Peut-être ne reste-t-il que Miami.
Replying to @leonor_de_b
Plus sérieusement, il nous appartient de créer la solution, elle n'est pas au menu.
♥ 1
Replying to @manu_thy and @pressecitron
"Le transformateur absorbe de l'énergie" LOL. Et elle va où?
Replying to @anaisrassat
I should probably spice it up :-)
Replying to @RReverser and @fkling42
If anyone wants to work on that, it shouldn't be too hard, I'm happy to provide pointers.
The New Web Typography robinrendle.com/essays/new-web… A most excellent piece.
♥ 2↺ 1
Replying to @RReverser
It’s been a while since I looked at the spec, but last I checked there was no «namespace» construct. @fkling42
"Designing Beauty: The Art of Cellular Automata", by Adamatzky/Martinez. A relaxing overview of the domain's variety, if you're into that.
Friends in France should also read this. nytimes.com/2016/02/26/opi…
♥ 1
If you've ever wondered what philosophy is for: philpercs.com/2016/02/a-publ…
♥ 2↺ 2
Any good book recommendations for a 6yo too smart for her own good?
Replying to @ShaneHudson
Ah that's a good idea, I really liked that one. I think it could work, thanks!
Replying to @_Florian_R
I thought of that but she scares easily.
Replying to @alanstearns
Nope, are they good? Checking it out, thanks!
Replying to @rdeltour
I'd like some more time before she understands just how fucked up the world is.
♥ 2
Replying to @tabatkins
Augie is next on her reading list. Is Thing Explainer actually good? It seemed a bit... I don't know. But I didn't look closely.
Replying to @rudyrigot
Thanks, getting it!
Replying to @perigrin
The magic treehouse ones don't last long enough, she could read several inside the day. I'll check the others, thanks!
Replying to @ShaneHudson
I've been thinking about HP but I'm pretty sure she'd be really scared. It's quite dark, quickly!
Replying to @null and @amyvdh
I can't wait for HP but I just know she'd be terrified. I'll try those, thanks a lot :-)
How justice actually works in the US fusion.net/story/272058/4…
♥ 2↺ 4
Replying to @tabatkins and @alanstearns
I'll check it out, see if it might work, ta.
Replying to @tabatkins
BTW do you say "Audjie" or "Augggie"?
Replying to @tabatkins
Man, that would have been epic and mean :-)
Replying to @null and @lshlj
Yeah, there was a scheduling clash with my birthday.
Specially designed for @annevk? teeturtle.com/products/not-g…
♥ 1