May 2013

214 tweets

"Il pensait avoir des principes, il n'avait que des préjugés." #àRecaser
↺ 2
Avant de tous pleurer sur Dailymotion, vous pouvez me parler un peu de leur politique en matière de respect de la vie privée? #jdcjdr
Go Fork and Multiply, or why @W3C specifications should use an open license: berjon.com/blog/2013/04/w…
↺ 15
Hey @twitter, that's three people just today I never unfollowed but was no longer following. Is there hope for eventual consistency ever?
↺ 3
Les vendeurs et vendeuses dans les magasins de fringues. Pourquoi? Mais pourquoi? #pourquoi
Des recommandations pour un extincteur de maison?
↺ 1

Replying to @stommepoes

I fear my simple presence is a big dongle joke in and of itself.

Replying to @dr_cyno and @cynorrhodon

seulement si ça tombe à moins de 5 phrases de "de tout temps, l'homme".

Replying to @pbeyssac and @recifs

Peut-être un peu l'avocat du diable, mais les amis et l'air ont de fait un coût.

Replying to @graham_steel

actually, I never did any such thing :-)

Replying to @pbeyssac

Vraiment? La qualité de l’air se monnaie relativement directement, ou en tout cas c’est un but avéré. @recifs

Replying to @pbeyssac

Quant à l’amitié, si tu penses aux points de rupture ce sont souvent des ruptures de l’échange. @recifs

Replying to @pbeyssac

Bref, je ne connais pas le contenu du rapport mais le soundbite ne me choque pas. Un écosystème viable a une structure… @recifs

Replying to @pbeyssac

…d’échange stable. Parler de troc plutôt que de monnaie me semble juste être une image. @recifs

Replying to @pbeyssac

Au contraire, ne pas se préoccuper du coût de l’air fut un bref répit anthropocentré. Demande ça aux cyanobactéries :) @recifs
In case you missed it Friday: Go Fork and Multiply, or why @W3C specifications should use an open license: berjon.com/blog/2013/04/w…
↺ 6

Replying to @chaals

Heh, for a second I thought you were talking about the open license ;) Thanks!

Replying to @null and @sideshowbarker, @naoishikura

Whoa, you’re both so classy :)
Today I saw the future, by @BrendanEich brendaneich.com/2013/05/today-… Very sexy stuff (doesn’t talk about Futures though :)
↺ 1

Replying to @briankardell

You mean using WebDriver? Maybe @tobie knows? @tabatkins @marcosc @cowboy @FremyCompany

Replying to @julien_igel

burn!
Some very good notes on how to better organise specifications, by @dontcallmedom lists.w3.org/Archives/Publi… lists.w3.org/Archives/Publi…
↺ 1
Media Queries are a Hack ianstormtaylor.com/media-queries-…
↺ 3

Replying to @briankardell

Nope, pointer? @FremyCompany

Replying to @briankardell

Pretty sweet! I’ve been wondering if Hitch isn’t the thing to standardise though :) @FremyCompany

Replying to @briankardell

Layering! Modularity! All that crazy talk! @FremyCompany

Replying to @anaisrassat

Is this just dark stuff or Euclid in general? If the latter, can you give details on the exoplanet hunting side of things?

Replying to @briankardell

AFAIK it’s supported more deeply but only by Gecko.

Replying to @anaisrassat

Whoa, you’ve become quite the community manager :)

Replying to @anaisrassat

I think the jury’s still out on a precise definition, but it’s like that other thing: you know it when you see it :)

Replying to @briankardell

On the pro side: packaging. Pretty much everything else is annoying though.

Replying to @briankardell

You mean how to start something new in a way that can transition to standards easily? @FremyCompany

Replying to @briankardell

What I’d do is use ReSpec with specStatus=unofficial, and then grow from there. @FremyCompany

Replying to @FremyCompany

Editing a spec without a proper tool is a lot of useless work :-) @briankardell

Replying to @briankardell and @FremyCompany

sure thing, but probably not before tomorrow though.

Replying to @yandexcom

I really like some of the stuff in BEM, have you considered pushing it more, eg in standards? @ianstormtaylor @bem_en

Replying to @chaals

Maybe if I go to YAC we could sit down to chat about it? @bem_en @harisov @veged @yandexcom

Replying to @tobie and @briankardell, @FremyCompany

oh yeah, ReSpec also supports specifications written in MD if, you know, you hate HTML and the web :-)

Replying to @null and @marcosc

Great for the TAG, sucks for Mozilla! Or, err, something like that.

Replying to @dietrich and @Boot2Gecko

Wow, MMS. You have guts. Lots of guts.
↺ 2

Replying to @FremyCompany

Hahaha, I'll have to buy you a beer for being funny like that :-) @tobie

Replying to @chaals and @mursya

Is there a CfP I need to fill out or something? I tried Yandex.Translate on it but I’m only getting so far :)

Replying to @tobie

Heh, you beat me to it — great! @briankardell @FremyCompany

Replying to @FremyCompany

It does a lot of the annoying spec-related stuff for you. Switch to a standard? Easy: tobie.github.io/respec-example… @tobie

Replying to @mursya and @chaals

Ok will do!

Replying to @FremyCompany

It doesn’t work with the «unofficial» style because it doesn’t have the autofeedback section. @tobie

Replying to @FremyCompany

But it does exist, it’s subjectPrefix. See tobie.github.io/respec-example… @tobie

Replying to @FremyCompany

I didn’t «forget», the docs are just lagging :) I’ll fix that in the docs rewrite that’s badly needed.

Replying to @FremyCompany

Thanks man! @briankardell
Some *very early* notes on a new take on Web Intents/Activities that might kill the UX issues. darobin.github.io/web-wish/ Comments welcome!
↺ 2

Replying to @fabricedesre

can you show me use cases you have in mind? I think it's not necessarily limited that way.

Replying to @fabricedesre

Well if it’s out of UA scope, it’s solved. Seriously, it really is. Mucking with tabs was a bad idea.

Replying to @null and @karlpro

To me is good enough; to the GH repo’s issue tracker if you want :)

Replying to @fabricedesre

That's the same for the whole web. Bad designers create bad ux. But you can create good ux with this, in fact by default.

Replying to @null and @karlpro

bring it on baby!

Replying to @julien_igel

Ties for what? And the proper use of CONNEG is no, no, don't, no way :-)

Replying to @julien_igel

But ordering preferences is UA-only, so I’m not sure where that would appear?

Replying to @julien_igel

But I’m not sure where the q value would come from since the sources are disparate :) Maybe it’s unclear, I’ll prototype.

Replying to @julien_igel

Yeah, q values only make sense compared to others. But you don’t know the others.

Replying to @rubin

je vois bien comment hacker ça à la main, mais je ne connais pas de service existant.
An interesting take on making W3C even more open: «Reforming the W3C Advisory Board » tantek.com/2013/128/b1/re… by @t
↺ 6
@hadleybeeman Actually, it’s the antepenultimate — there’s PR after that :)

Replying to @hadleybeeman

Yeah, PR tends to be a formality :)

Replying to @Tris_Acatrinei

C'est de la balle. Je n'utilise plus que ça.

Replying to @briankardell

A protocol? You mean a URL scheme? That would kill the transition strategy.

Replying to @briankardell

Sure :)

Replying to @briankardell

Starting to use the feature before all browsers support it. A scheme breaks existing stuff. <link> is an easy upgrade path.

Replying to @briankardell

Replying to @briankardell

Not necessarily, see lists.w3.org/Archives/Publi…

Replying to @briankardell

Well, for values of server-side that don’t require any code :)

Replying to @domenic

the very first thing that I want to do with those is a rich text editing form control without the sheer insanity of contentEditable
↺ 1

Replying to @null and @_jden

I’m wondering if starting from scratch with a sane model in a new element is something implementers could be sold on. @domenic

Replying to @GPHemsley

The West Wing didn't get cancelled!

Replying to @garannm

Aha! @horse_js, I have unmasked you!
DRM in HTML5 is a victory for the open Web, not a defeat arstechnica.com/business/2013/…
↺ 6
Lots of science headlines on plant communication. Given the rest of the biosphere does it, it would be very surprising if they *didn't*.
↺ 1
If you want to build a ship don't herd people together to collect wood and don't assign them tasks and work, but rather teach them to (1/2)
(2/2) long for the endless immensity of the sea. Antoine de Saint-Exupéry, Wisdom of the Sands

Replying to @null and @karlpro

fascinating!

Replying to @FremyCompany

I thought you'd get electronic sheep.

Replying to @FremyCompany

shame it's a pokemon :-)

Replying to @FremyCompany

didn't Windows 95 already do that though?
Antibiotic shown to prevent stupidity in men arstechnica.com/science/2013/0…
«The Quantum Universe» by Brian Cox & Jeff Forshaw. A good coverage of quantum physics, both correct and making sense (via @dontcallmedom).
I Don't Want Your Fucking App idontwantyourfuckingapp.tumblr.com (via @xcazin)
↺ 18
"Complex Adaptive Systems" by John Miller & Scott Page. Interesting but slow introduction to the topic.
Lee Smolin on evolutionary cosmology edge.org/conversation/t…
Inforworld has a great new logo for HTML5, we should all switch to that infoworld.com/t/cringely/drm…

Replying to @null and @sgalineau

I sense a use case for WebGL.

Replying to @distobj

Apparently!

Replying to @Simon_Lucy

Remove your polarising glasses, it certainly does blink if you look at it right.
Les logiciels de compta, pourquoi vous êtes TOUT EN MAJUSCULES? C’est codé par des gamins de 13 ans?
↺ 2
What’s a good Mac tool to pick subsets of a diff to apply? Git support a plus.

Replying to @heycam

That scales poorly to 120k lines documents, I find :) I think something visual is called for here, to get better context.

Replying to @heycam

Ha! You know what, I actually have it too, just never used it :) Thanks for the reminder, I’ll look!

Replying to @heycam

I’m just a bit afraid of Atlassian because of Jira...

Replying to @heycam

Likewise, and I’d like to keep it that way :)

Replying to @thommay

No, I want to pick diff parts over many commits, not pick commits. What’s vim? :)

Replying to @thommay

Thanks! I’ll check that out. Looking at SourceTree now, but it’s pretty slow.

Replying to @heycam

So, SourceTree seems to have the feature, but trying it out brings my laptop to its knees...

Replying to @null and @madsgraphics

Thanks, I’ll look at DiffMerge. Kaleidoscope seems nice but it’s expensive.
I’ve probably tried every single diff tool for Mac by now. The only one that scales is TextWrangler, trailed by @Kaleidoscopeapp and Meld.
↺ 1

Replying to @null and @karlpro

I looked, but I couldn’t find a way to diff a file with another branch.

Replying to @null and @karlpro

Hmmm, but that requires there to be a merge conflict! I just want to select parts of a very large diff.

Replying to @null and @karlpro

I can get that with local git clone -b other-branch. The problem is finding a tool that can handle a 120kloc file with 25kloc diff.

Replying to @sangwhanmoon

I have, it looks nice, but it simply cannot handle this sort of load. It crawls to a stop. @karlpro

Replying to @sangwhanmoon

120kloc file with 25kloc of differences.

Replying to @sangwhanmoon

No, it’s not about forgetting to commit, it’s a change in the way branches are handled.

Replying to @sangwhanmoon

Thanks for the tip, but it’s not available for Mac :)

Replying to @sangwhanmoon

Have you tried TextWrangler’s diff? I tried it out of desperation, and it’s the best I’ve found. Not great, but works.

Replying to @anssik

and in several pieces :-)

Replying to @mikeal

Replying to @null and @jerekapyaho

Components! Define your own HTML! Remove implementers from the critical path to innovation! @anssik

Replying to @null and @jerekapyaho

But they’re *shadow* tree. @anssik
@marcosc You wanted a pony?
Mailing List of Fire youtube.com/watch?v=HCwgKt… (cc @rodneyrehm)
↺ 1

Replying to @chaals

Remind me to ask Ashley when I get home :) @marcosc
Is using Haibu any good? This is just for the one server, but with a bunch of Node services. It looks ok, but the docs are quite sparse.

Replying to @Tris_Acatrinei

et Amazon, à la rame, c'est long. #jesors

Replying to @null and @francbelge

Il n'y a pas vraiment de régularité pour -acy. Peut-être privacité?

Replying to @null and @francbelge

Joli mot, well played! Ça me donne un nouveau sens au Privacy Working Group :)

Replying to @null and @sgalineau

I haven't looked closely, but isn't it the new name precisely for Toolkitchen?

Replying to @garannm

I also struggle in choosing between length and lenght. So many options!

Replying to @mikiane

"Oublient"? :-)

Replying to @tobie and @slightlylate, @jaffathecake

I've been wondering if we'll need resource packaging or not.
Je loue un super bureau dans un super co-working dans le 11ème pour juillet-août. Qui veut? cc @spintank.
↺ 5
A faulty build of ReSpec (v3.1.56) was shipped for a few minutes. If you are having problems seeing references, that’s it. It’s now fixed.
@beaufour Are you in Denmark this summer? We'll be there from mid-July for two weeks :-) Somewhere on the riviera, not far from Copenhagen.

Replying to @beaufour

sweet! IIRC the two weeks after the 18th. Let's try to hook up at some point!

Replying to @beaufour

ah, shame! Keep me posted in case it changes!

Replying to @philarcher1

and that's precisely why you'll do it anyway! GitHub is calling!
@koalie z? w?
FIP ça serait la meilleure radio du monde si seulement le jazz n'avait pas été inventé.
Is there some trick to get Firefox Mobile to always open the tabs from last time? Can't seem to find it in about:config.

Replying to @econoclaste

est-ce vraiment en cela différent des autres Cronenberg ?

Replying to @null and @koalie

I guess that as someone who has tw, ls, i, or f as mailbox names I shouldn't comment too much.

Replying to @edasfr

en comptant les standards, autour d'un mois par an.

Replying to @DeeCeeMountain and @douglasc_hill

but history doesn't have all my previously opened tabs.

Replying to @mattur

what, it's not mattoor? Damn.
Et si on fermait la Cour des comptes? blog.francetvinfo.fr/classe-eco/201…

Replying to @mounirlamouri

Sweet, thanks! Nah, I’ll see it when it comes with the daily Aurora update :)
I wonder if it’s time to start the negative ads blitz for the @w3c AB election. «Wrong on Good Standing — Wrong for the AB!»

Replying to @DeeCeeMountain and @douglasc_hill

Thanks for your help, but I’ll wait for the actual feature to ship (apparently soon)!
L’idée qu’il faille une décision de l’AN pour que des universités puissent enseigner en anglais me désole profondément. Belle autonomie!
I wonder how many XML documents sitting on the Web would suddenly become well-formed if &nbsp; became a built-in entity.
↺ 1

Replying to @jirkakosek

Not fair, JS supports arbitrary size and precision number types. Just delimit them with " and " .

Replying to @mnot

Oh I’m not proposing we actually do it, just idly wondering. But I did just see that in the IETF’s XML RFC index :)

Replying to @mnot

FWIW it was in xml.resource.org/public/rfc/bib… in case you know someone who’d fix it.

Replying to @mnot

Super, ta. We might actually be using it for github.com/tobie/specref

Replying to @slightlylate

I wish @github would show the version from inside package.json directly in a module’s UI.

Replying to @AAlaphilippe

j'ai envie de te répondre: never!

Replying to @null and @francbelge

ça peut être rassurant, les modèles qui prévoient bien le passé sont souvent surentraînés sur celui-ci.

Replying to @null and @sgalineau

your dedication to the cause shall be justly punished.

Replying to @null and @sgalineau

upgrade to TextMate? *ducks*
HTML — now with even more crocoducks inside! w3.org/html/wg/drafts… (thanks to @stevefaulkner)
↺ 4

Replying to @null and @notabene

il y a un caractère Unicode pour le double point d'exclamation. Essaie de convaincre ton interlocuteur que c'est ça. #astuce
↺ 1
↺ 3

Replying to @null and @karlpro

hahaha, well done!
Il est urgent d’enseigner l’art du formulaire HTML à nos administrations. Les impôts, c’est juste catastrophique.
↺ 2

Replying to @leonor_de_b

Direction? We have elected to...?

Replying to @edasfr

fais gaffe, quand c'est trop lent après ta session timeout :-)

Replying to @lauredlr

C'est triste à dire, mais l'immortalité des positions de cette manifestation mérite-t-elle vraiment mieux?

Replying to @null and @marcosc, @barbaraneves

So stylish!
Et là c’est le moment où tu te rends compte que tu as raté le CfP @ParisWeb… :-(

Replying to @echotraffic

Tout ma sympathie, ça ne doit pas être drôle aujourd’hui. A Montreuil on a moins ce problème :)

Replying to @sebcbien and @madsgraphics, @ParisWeb

voilà c'est ça. Bouhou je suis trop nul.

Replying to @null and @madsgraphics

Ah oui, un bisou! Un bisou!

Replying to @mnot

I get a different WF error :)

Replying to @null and @madsgraphics

Ouais! Ça me consolerait presque d’avoir raté le CfP :)

Replying to @mnot

Still nbsp, but different location (there are several) Line Number 162467, Column 7: IEEE&nbsp;802.3&nbsp;Medium...

Replying to @mnot

It’s declared in an external subset. Not all parsers read those. Declaring it in the internal subset should fix that.

Replying to @mnot

I saw the location shift on reload, but maybe that was due to another fix?

Replying to @mnot

XML is hard ;)

Replying to @mnot

I could certainly do with new shoes.
Sauvons les auto-entrepreneurs! #PiouPiou change.org/petitions/ne-t…

Replying to @johnallsopp

The JSON indeed lacks details… I don’t know about SOAP, but I get XML from jigsaw.w3.org/css-validator/… cc @wz43rtx

Replying to @johnallsopp

From the link I gave you? That’s weird, I get XML (in Firefox). Replacing ucn with json I get JSON, but no details. @wz43rtx

Replying to @johnallsopp and @wz43rtx

I’m not saying that JSON wouldn’t be better — just trying to help you out quickly :) It’s not the same code base.
Spend several minutes being seriously puzzled by a bug report, before noticing it’s not the right Bugzilla instance. *sigh*

Replying to @johnallsopp

wd — A node.js client for webdriver/selenium 2, github.com/admc/wd (cc @briankardell)
↺ 1

Replying to @briankardell

Yeah, I was just wondering if you’d seen it given how popular Java is :)

Replying to @briankardell

I know; I’m not sure. Just noodling it over.

Replying to @briankardell

LOL! I wasn’t complaining man, I just noted the «why Java?» in the thread and thought about alternatives.

Replying to @briankardell

I’ll keep noodling, let’s see if something comes out eventually :)

Replying to @briankardell

Well, doing screenshots would cover a lot of it. But that requires extensions for all browsers...
@github We moved a repo from a user to an org, and now people can’t access things (e.g. issues) despite being on the right Team and all.
@github Repo is github.com/w3c/manifest. Previous user was marcoscaceres. Any help very welcome! We can’t work on making the Web better :)

Replying to @Turblog

j'ai pas regardé de près mais IIRC le bouton twitter fonctionne par injection de JS. Qui a pleins droits. Sur une page du fisc.
↺ 1

Replying to @Turblog and @vincib

Faudrait regarder, tu peux injecter par iframe mais j'en doute. Et après ils nous empêchent de faire du copier-coller...

Replying to @null and @xquery

Just use jQuery man!
L’URSSAF, c’est magique! berjon.com/blog/2013/05/u…
↺ 18

Replying to @jwajsberg

Mais non, ils veillent sur nous! C’est pour notre bien!

Replying to @alpinfra and @2k3s

j'y ai pensé mais la dernière fois que j'ai vu le guichet ils ont bien foiré mon dossier !

Replying to @wilto

fwiw it just looks like a train wreck on Firefox.

Replying to @wilto

sorry, I forgot to mention : Firefox on Android 2.3. (Aurora precisely)
References and Modularity in specifications lists.w3.org/Archives/Publi…
↺ 2

Replying to @mnot

Wunderbar, it works — thanks! Who cares about validity? :)
@jcartillier Jérôôôôme!!!
«There can exist no two systems such that their PHP configurations are identical and interoperable.» -- Robin’s PHP Deployment Conjecture
↺ 13

Replying to @FremyCompany

Whether limited to production systems (weak conjecture) or all systems (strong conjecture) is indeed debatable.
Montreuilloises, Montreuillois, j'ai perdu mon chat, Billie. Ma fille la cherche partout! Cc @ville_montreuil
↺ 3

Replying to @Tris_Acatrinei

Yes, @poulpita à #OVEI c'est une superbe idée :-) @Turblog
Great post by @Hoedic about how open data needs standards with open licenses dataholic.ca/2013/05/24/sho… cc @w3c
↺ 6

Replying to @Hoedic

Out of curiosity, have you looked at W3C Community Groups? Those can already use open licenses.

Replying to @poulpita

Fais gaffe, c’est super cher… @Tris_Acatrinei @Turblog

Replying to @Hoedic

Oh, hey, I’m on that group :) If transitioning from a CG to a WG, I think @w3c would be receptive to talking about an open license.
↺ 1
That’s how I read @W3C standards: xkcd.com/1219/
We’ve broken all turnout records for a @W3C Advisory Board election. Polling closes tonight, AC reps: go vote if you haven’t!
Madame @SylviaPinel selon nos sources à Bercy la fin des problèmes informatiques du RSI serait en 2016 au lieu de 2014 annoncé. Est-ce vrai?
The Open Food System research project groupeseb.com/en-en/content/… #want
Ça, plus une bonne soupe, et je suis prêt à commencer le mois de juin.