September 2013

197 tweets

Any suggestions to teach (very) basic astronomy to a four year old? Cc @anaisrassat

Replying to @bortzmeyer

I was thinking more about a book, esp since she likes looking at them alone too. Or an app/site, she likes computers too :-)

Replying to @rem

I wonder that every time I slice some onions. Haven't worked myself up to try it so far.

Replying to @null and @karlpro

Légèreté rêve, poésie - si je ne te connaissais pas, je craindrais de m'endormir :-) @anaisrassat

Replying to @JibberJim

Both of us would have to learn to sing first... @anaisrassat

Replying to @julien_igel

"What else?" She can already spot that, as a saucepan of course. @JibberJim @anaisrassat
Well, saw that one coming from miles away.

Replying to @julien_igel

Perhaps indeed, but you never know, Microsoft knows a thing or two about turning around.

Replying to @michaelhkay

How much have you looked at dynamic code loading (assuming it’s possible)? E.g. some functions are likely pretty rarely used.
Before making a nav sidebar fixed, you want to check that it fits in the screen height. Looking at you @twitter Bootstrap, @mozilla Brick.
↺ 1

Replying to @briankardell

Part of the problem is also that people don’t pre-compile their templates, though they often can. @jaffathecake @tomdale
Great piece on writing good docs: Teach, Don’t Tell stevelosh.com/blog/2013/09/t… (Loved the part about wikis and autodocs.)
↺ 4
Please don’t *ever* implement your own scrolling behaviour. Spreadsheets are bad enough, custom scrolling makes Google Spreadsheets useless.
↺ 1

Replying to @null and @karlpro

Not when other people insist on using these things :)

Replying to @anaisrassat

Thanks! That’s more for Thelma, but it does indeed look beautiful. @JibberJim @julien_igel

Replying to @grtjn

Indeed ;) Some abstracts don’t fit vertically, and I can only scroll them by editing them! @karlpro

Replying to @grtjn and @wcandillon

I’m curious to see your XML Amsterdam abstract! cc @grtjn

Replying to @grtjn

You may simply have a larger screen :) I also think it emulates scrolling differently for each platform (here, Mac). @karlpro

Replying to @null and @karlpro

No way! In for a penny, in for a pound! It’ll teach you to make drive-by comments :) @grtjn

Replying to @anaisrassat

Yeah, pretty pop-up books with little text won’t keep a 4yo very long :)

Replying to @anaisrassat

E.g. she can watch Attenborough’s Planet Earth forever, then rave for hours about submarine volcanoes and schools of dolphins.

Replying to @anaisrassat

However, there’s probably a sweet spot between pop-up books and binary code/classical mechanics equations tees :)

Replying to @wcandillon

you mentioned a prototype?

Replying to @glazou

2 Daniel, 2. @jwajsberg @notabene
@danbri Il est impossible de s'emparer du sujet sans être immédiatement catalogué dans la vile droite des méchants patrons. @nkgl

Replying to @glazou

Ouais enfin je crois que je vais quand même appeler mes potes de chez Samsung pour les prévenir, hein. @jwajsberg @notabene

Replying to @pbailly

SVG’s <switch> element is supported across the board, but useless. Could it be made useful, e.g. for responsive content?

Replying to @heycam

Pretty much what I had in mind, yes.

Replying to @rodneyrehm

Have you ever used it?

Replying to @rodneyrehm

yeah, but have you seen what you can switch on? And where's 2/2?

Replying to @rodneyrehm

strange, that "..." tweet does not show up in my notifications...

Replying to @rodneyrehm

you can only switch on useless conditions, except for language. Adding media queries would be sweet.

Replying to @pierrehaski

Et de Matthieu Conquet aussi! @xporte @mvoinchet

Replying to @stevefaulkner

That’s the thing where including jQuery was breaking ReSpec, right? I should find a general fix. @surkovalexander @jkiss

Replying to @null and @jkiss

I didn't have time to investigate, but it was causing processing to blow up midway. @stevefaulkner @surkovalexander

Replying to @philogb

There have been proposals to do that, but limited support. Happy to chat more over email.

Replying to @null and @jkiss

Not sure, I tried various quick fixes but to no avail. It’s tracked in github.com/darobin/respec…. @stevefaulkner @surkovalexander
@JoshHalliday Small nit: it’s Wendy Seltzer, not «Saltzer», in theguardian.com/technology/201…. Thanks!
Tu vas à @ParisWeb? Fais-nous une proposition de Lightning! Avec @glazou on répond aussi à vos doutes et questions! paris-web.fr/actualites/201…
↺ 1
@ParisWeb Quel(le) plaisantin(e) a rajouté des espaces pleines insécables devant des ponctuations hautes dans paris-web.fr/actualites/201… hein?
Le 02 octobre je serai à Lyon pour causer standards du @w3c à la @blendconference. Le programme est alléchant: blendconference.com
↺ 3
@danbri You leaving the HTML WG sweetie?

Replying to @danbri

Well, we’ve been wondering about you and Microdata; did you get my email from a week or two back?

Replying to @danbri

Talk whenever you want buddy :) @chaals
I have *so* found the new Gangnam Style youtube.com/watch?v=jofNR_… #WTFFriday
↺ 1

Replying to @null and @deezthugs

But man... what does the fox say?

Replying to @null and @koalie

I was more thinking of a furry duet. We need to find a sponsor for @Lachy! @w3c
If you work in standards or open source, read this: Academic assholes and the circle of niceness thesiswhisperer.com/2013/02/13/aca…
↺ 1
Pourquoi nos brevets ne créent-ils pas d'emplois? lemonde.fr/idees/article/…
↺ 1

Replying to @briankardell

Indeed — which is why I wanted to point it out to others :)

Replying to @dietrich

I was planning to, but unfortunately it turns out I won’t be able to make it :(

Replying to @dietrich

yeah, it sucks :-( Thanks, I've already emailed the registration folks.

Replying to @fjhirsch

I've worked on stuff like that, it's actually pretty useful in order to make roads safer.

Replying to @mounirlamouri

it's a real thing @brucel w
Responsive Images meetup starting now at Mozilla’s terribly derelict Paris office. #respimg
↺ 1
Quick a11y question: if I have an arbitrary element and I wish it to have an equivalent of img's alt, is using aria-label appropriate?
↺ 1

Replying to @jirkakosek

I’d like implementations that don’t know the element to use the attribute. I don’t think @title gives us that.
My last-minute suggestion for responsive images, «Recycling Spare Parts: SVG's <switch> element» berjon.com/presentations/… #respimg
↺ 3

Replying to @stevefaulkner

My use case is this berjon.com/presentations/…

Replying to @jirkakosek

I love the example :) I guess I’ll use that for now, thanks!

Replying to @jirkakosek

Thank you, that is very much appreciated :)

Replying to @null and @MarcoInEnglish

Thanks, that’s helpful. It may need different roles since the content can vary.

Replying to @null and @MarcoInEnglish

Yes, but the use case is switching between elements as in berjon.com/presentations/…. So it’s a matter of picking the right role

Replying to @ned

That’s *already* in SVG :) And I’m pretty sure that SVG has been TC for about a decade by now!

Replying to @alastc

I reckon that depends on the content — for video text replacement might not always work so well. @MarcoInEnglish

Replying to @null and @MarcoInEnglish

<switch> picks one child for rendering depending on tests. I’m thinking of using it for responsive design.

Replying to @null and @MarcoInEnglish

So basically, you get different content based on a media query, reusing a piece of SVG in HTML.

Replying to @alastc

Not for everything, but I at least want the label to be available for cases where it makes sense. @MarcoInEnglish

Replying to @alastc

Maybe that means it really is a job for title. @MarcoInEnglish

Replying to @null and @MarcoInEnglish

SVG supports a bunch of stuff; the idea here is to make it more useful.

Replying to @null and @MarcoInEnglish

Actually, you probably want to talk with @shepazu precisely about that. He’s working on SVG A11Y at the moment.

Replying to @alastc

I’m well aware, hence my initial hesitation. Folks on #htmlwg IRC are also helping out. @MarcoInEnglish

Replying to @ned

I’m afraid I don’t know of such a resource, the closest might be w3.org/Graphics/SVG/T…
@rik24d @nitot Vous êtes au bureau?
Demo shim to use the SVG switch element for responsive images: berjon.com/presentations/… #respimg

Replying to @remitbri

yeah it's a five minute hack before lunch, I need to refine it!

Replying to @anthony_ricaud and @rik24d

c'est con pour une fois que je passe! @nitot

Replying to @stevefaulkner

Yeah that makes sense, thanks!

Replying to @null and @nitot

On est dans la salle des fêtes; je passe te dire coucou vite fait si tu me dis où c’est :) @rik24d
After much swearing and ugly hacks, the responsive switch demo/shim now works in Chrome and WebKit berjon.com/presentations/… cc @remitbri

Replying to @cconcolato

I was lazy… Do you have better images?

Replying to @cconcolato

I guess I can just add a gradient or something...

Replying to @cconcolato

Dude, that sounds like work :)

Replying to @cconcolato

I made it clearer using my l33t design skillZ.
If in doubt, just replace every element in the DOM with a clone of itself. #RepaintDammit #respimg

Replying to @null and @sgalineau, @nitot

It positively always works in French youtube.com/watch?v=GSeaDQ…

Replying to @cconcolato

In what?

Replying to @cconcolato

Oh! Sorry… Just resize your viewport :)

Replying to @ShaneHudson

Likewise, it was great to meet you! BTW this is what I was talking about: github.com/w3c/web-platfo…
ExoBrowser — My Quest for a New Browser spolu.github.io/exo_browser/20…
↺ 5

Replying to @Lobstrosity

I know :) It’s not a fatality implementation-wise, but as a temporary solution it makes it suck.

Replying to @briankardell

Yeah, it looks like a potentially good way of trying out new ideas that can’t easily be polyfilled.

Replying to @briankardell

Yes, I’m using the term in a loose way (as always, I don’t really care about the differences :).

Replying to @briankardell

I’m thinking anything that requires UI (e.g. Wishes) or device/OS access.

Replying to @briankardell

Yeah, I thought of that too :)
Pour lutter contre le communautarisme, je propose d’interdire les signes d’appartenance politique en public. Chemise Vichy, t-shirt du Che…
↺ 1

Replying to @alexduloz and @helloanselm, @wilto, @tobie

How come you're all replying to a tweet I wrote a year ago? Going through my Complete Works?

Replying to @wilto

I know where it was lifted from, but the observation still holds! You should get that cough checked out.

Replying to @alexduloz

Whoever succeeds in crossing cats and dinosaurs will take over the Internets. But it's proving non trivial. @wilto @helloanselm

Replying to @wilto

But I'm too young to get the Greatest Hits treatment. I'VE STILL GOT A SONG IN ME. @alexduloz @helloanselm @tobie

Replying to @null and @sgalineau

If they're not fossils, where are their feathers, huh? @glazou

Replying to @ernesto_jimenez

I think the revenue model is that we need more revenue :-) @torgo @w3c @jaimeiniesta
The Feynman Lectures not just freely online but using HTML/SVG feynmanlectures.caltech.edu It's about time we killed LaTeX. Via @mixedberry.
↺ 3
Similar to @TWIT4SHKPNK on Spotify. Ouch.

Replying to @Sniperovitch

Heh. Les saisons 3 à 5 sont optionnelles. Et après ça tu rattrappes ton retard sur quoi? Buffy? West Wing? @Turblog @KatSpi

Replying to @jjgod

Do you reckon that would be a worthwhile project to attempt? Maybe some form of compact MathML?

Replying to @hober

We should make it the official uniform.
A few great example of DocBook element names: berjon.com/blog/2009/12/x… #owpworkflow
↺ 1
«Github For Writers» might be of interest to attendees at #owpworkflow madebyloren.com/github-for-wri…
↺ 2

Replying to @null and @marcosc

Yeah this project really needs to take off!
«Problem: it doesn’t parse the same as XHTML and HTML.» says --@jirkakosek «Solution: simply never parse as XHTML» —me #owpworkflow

Replying to @TAC_NISO

I never said that you might want to use Java!

Replying to @TAC_NISO

I’m not sure which part it was, I did talk of jQuery and JavaScript — but I would never recommend Java :)
@hober Is there anywhere in the spec where we state that elements and attributes with a hyphen in their name are open for all? Can’t find it
@hober I’m getting a sense that it’s something everyone agrees to but isn’t written anywhere...

Replying to @TAC_NISO

No worries, I was just worried for my reputation as a through-and-through Java hater ;)
@recifs Il faut que tu me dises combien de bières il faut pour que SeenThis arrête d’injecter des espaces avant les ponctuations hautes :)
↺ 1

Replying to @glazou

We need to be distinguishable, otherwise the ladies get confused with such a chunk of hunk. @hober
If I understand @hober correctly, a book is basically a web site where you don’t end up on a cat picture. #owpworkflow
↺ 2

Replying to @alanstearns

I’d have to check, but I *think* I have a book or two with no cat pictures. @hober
@hober @ficklepixel So… Staying asleep or is it time to make this interesting?
Good idea from the margins of #owpworkflow: save ReSpec output straight to GitHub, using github.com/philschatz/oct…
GitHub Book: edit books directly over GitHub github.com/philschatz/git…

Replying to @brucel

You told it all wrong, it’s supposed to be with a pirate!

Replying to @brucel

I just know first-hand that Churchill told it with a pirate.

Replying to @brucel

I mean seriously, man, seriously. «Arrrrr, it be drivin’ me nuts!»
I said we weren’t killing XHTML. I didn’t say that it wasn’t dying. #owpworkflow
↺ 9

Replying to @brucel

They’re all «de Gaulle» variations, «gaule» means nob. #truestory
@w3cmemes The honour!

Replying to @rdeltour

The XSLT people are already looking at operating straight on HTML. I really want all the cool XML tools to do the same.

Replying to @rdeltour

It’s not even so hard… Once you have a DOM, which has off-the-shelf parsers, you can already handle most of it.

Replying to @rdeltour

This is the droid you’re looking for: w3.org/html/wg/drafts…
Since people at #owpworkflow keep coming back to this: scroll to the bottom of madebyloren.com/github-for-wri…, there’s a friendly diff example.
↺ 1

Replying to @null and @dauwhe

You would have small changes to some XPath functions due to differences in the data model, but the rest shouldn’t change. @rdeltour

Replying to @null and @dauwhe

It would probably look a lot like jQuery :) @rdeltour

Replying to @stevefaulkner

hmmmm, I'm sure I tried that. Must've been sleeping at the wheel! @surkovalexander @jkiss

Replying to @null and @pkrautz

What I wonder is whether it’s something worth standardising. MathML isn’t very popular, but we need a solution. @jjgod

Replying to @null and @pkrautz

Heh, yeah, I was thinking about 927. That said we need a standard that's accepted and I'd rather not rely on JS for this.

Replying to @null and @pkrautz

Also, there's talk of standardising Markdown... The lack of interoperability there is becoming annoying to many.

Replying to @barefootliam

I haven't played with it enough to know if it's good there. It would be good to have mockups of what good diffs would be.

Replying to @null and @pkrautz

No worries at all :-) You have a good point, if rendering is what scares off Implementers, a new syntax won't help.

Replying to @null and @pkrautz

Well it certainly hasn't been a priority in implementations. It's a limitation on moving science publishing fully to the Web.

Replying to @null and @pkrautz

MathML is not required by HTML5, the spec only defines how you handle it if you support it. Hence less debate.

Replying to @null and @pkrautz

It’s an important technicality if you want to get something into a standard :) Yeah, it is text, I just want to keep the semantics.
@eveningedition Your French edition's article on cookies is sadly almost entirely wrong. Other than that, KUTGW!

Replying to @null and @AymericPontier

on peut commencer plus près, genre lune de Jupiter, il y fera plus chaud :-)
A great read this Friday, «A Day At The Park» kiriakakis.net/comics/mused/a…
↺ 3

Replying to @null and @pkrautz

Well, if you think about the very long term preservation of web documents, it is better to use a well-established common language.

Replying to @null and @pkrautz

That’s something that MML gives you, that’s less the case with ASCIIMath and other microsyntaxes. Also for indexing.

Replying to @null and @pkrautz

Right, which brings us back to square one :)

Replying to @codepo8

OMG, want!

Replying to @null and @nhoizey

si jamais ça t'intéresse, j'ai un script qui fait Kindle -> Evernote. Un peu sale, mais ça marche. @edasfr

Replying to @null and @pkrautz

The square before that one was «How do we get maths on the web to fully work, dammit», as discussed at #owpworkflow @jjgod
A new and much improved CSS module for Ruby w3.org/TR/css3-ruby/
↺ 2

Replying to @kiphampton

What’s SVG?
@robynbergeron I had no idea there would be someone with a name so close to mine :)
La science est trop peu et trop mal enseignée aux enfants. Un projet à soutenir pour changer ça: fr.ulule.com/contes-planeta…
↺ 5

Replying to @evgenymorozov

That's not so different from logrolling in The Calculus of Consent :-)
@RegardsCitoyens Vous pensez qu’il est possible d’avoir une liste de tous les candidats aux municipales d’une façon ou d’une autre?
↺ 2

Replying to @Rorschach_1987 and @LiberTIC, @RegardsCitoyens

Je songeais à un moyen de répertorier ceux qui s'engagent à ne pas cumuler.
@Tris_Acatrinei Hmm, 15-24 c'est pas très loin de 40% de la période de fertilité. Sans nier le problème, c'est pas le meilleur chiffre!

Replying to @EricScherer

peut-être est-il temps de revenir aux séries publiées dans des quotidiens :-)
↺ 1
Tu vas à @ParisWeb? Tu as jusqu’à lundi pour soumettre ton superbe Lightning Talk! paris-web.fr/actualites/201… #AllezAllezAllez
↺ 3

Replying to @TEF_FirefoxOS and @OpenWebDevice

I reckon it's the present of the Web!

Replying to @davrous

Attend il faut que je demande à l’équipe de @ParisWeb si on a ce qu’il faut pour marcher avec Windows :) @deltakosh @SiegfriedEhret
Toi qui va à @ParisWeb et qui n’a pas soumis de lightning talk, sache qu’avec @glazou on peut aussi choisir des volontaires du public.
↺ 1

Replying to @dascritch

Si tu en as un de secours, soumet-le (bordel). @lhirlimann @ParisWeb @glazou

Replying to @dascritch

On s’en fout de ta conf (bordel aqueux), prépare plutôt un LT! @glazou @lhirlimann @ParisWeb
Si vous ne soumettez pas plus de lightning talks pour @ParisWeb, je retire <canvas> de HTML et @glazou jette les gradients de CSS. #menace
↺ 5

Replying to @mariejulien

‘ttention, là, j’ai une syntaxe flexbox/grid qui utilise uniquement les tétragrammes unicodes! @nhoizey @ParisWeb @glazou
Attention, si tu n’envoies pas de LT pour @ParisWeb, @glazou et moi allons twerker sur un mashup de youtube.com/watch?v=jofNR_… et Gangnam Style.

Replying to @desbenoit

Rock’n’roll! @glazou @mcpaccard @tigramlich

Replying to @null and @nhoizey

Oui, avec ça ça te fait direct le layout de wizardgate.org @mariejulien @ParisWeb @glazou

Replying to @null and @sgalineau

On n’est juste pas sûrs que le Palais Brongniart a le light system pour coller @ParisWeb @glazou

Replying to @glazou

Je sais que tu sais où trouver les déguisements! @sgalineau

Replying to @glazou

Fais gaffe on risque de finir à devoir le faire :)

Replying to @glazou

J’étais pas là pour celui-là, mais si tu cherches bien il y a des vidéos d’épisodes précédents :)

Replying to @anthony_ricaud and @rik24d

On verra le degré d’interop que vous aurez sans document de référence :) @ParisWeb @glazou

Replying to @anthony_ricaud and @rik24d

Oh ça ferait une belle quote pour la presse ça tiens. «Mozilla: On s’en fout de l’interop! Fuck les développeurs!» @ParisWeb @glazou

Replying to @null and @nitot

En même temps, d’ici à ce que la police scientifique soupçonne le minitel comme arme du crime… @glazou @rik24d
The Nature of Information, Paul Young. Some really good insights into a tough topic, but tries to cover too much ground and dilutes itself.
"The current return to the authoritarian principle in politics finds its principal sanction in the belief that the new machine technology...
...requires the control of an omnipotent state." Walter Lippmann. Year? 1937 #PlusÇaChange

Replying to @wnas

Connection speed is a really hard problem to nail, I doubt we’ll manage to make anything useful there. @stevefaulkner @tabatkins
@helloanselm What is the problem you’re seeing with CGs? It’s hard to fix if you’re not clearer! cc @w3c

Replying to @n1k0

L’idée est de fournir beaucoup plus que le validateur de base (qui reste gratuit et développé). @nhoizey

Replying to @georgwtf and @gkreimer

Thanks for your outstandingly constructive feedback! Have a nice day! @stevefaulkner @tabatkins

Replying to @georgwtf and @gkreimer

Then feel free to propose alternatives to the RICG. @stevefaulkner @tabatkins

Replying to @stevefaulkner and @komputist

we don't use that component anymore, so please move it just to the spec one, thanks.
Is there a protocol I can use to repeatedly stab in the face people who set viewport user-scalable to 0? What about browsers that honour it?

Replying to @LeaVerou

it's nowhere near as fun as @horse_js though.

Replying to @brucel

wait, if enough of us interact in this thread we can recurse and make it a moment too!
Seriously though, not being able to zoom by default isn't a feature of native, it's a bug. Let's not imitate paper all over again.
↺ 1

Replying to @alainlambert

Il y a encore des smartphones autres qu'Apple qui n'y sont pas déjà?

Replying to @cyberbaloo_

C’est sûr que lancer des Minitels dans le public ça marche moins bien que le chamallow. @nitot @glazou @rik24d