I think I'll let svgboilerplate.com lapse — if someone wants to take over please LMK.
August 2012
183 tweets
Pubrules are dead?
Replying to @null and @karlpro
huhu!
Mulling over Lazy Blobs: lists.w3.org/Archives/Publi… (and thread)
I'm having weird problems with launchctl starting multiple master nginx processes. Anyone seen this before?
Replying to @rubin
Nice try sir, almost believable.
Replying to @null and @yulz__
Nope, I'm still on Snow Leopard and don't believe I've upgraded anything. Yes, they fight over port 80 and fill the error log :)
Replying to @null and @yulz__
Unloading the launchctl config and starting nginx by hand works a charm.
Replying to @null and @yulz__
It appeared after I switched nginx from port 80 to 9898 to put stagecoach on 80, but reverting that changed nothing.
Replying to @null and @yulz__
No, just RunAtLoad, KeepAlive, and WorkingDirectory. The simplest stuff.
Replying to @null and @yulz__
I'll give that a shot — it's really weird. Thanks for trying to help :)
Replying to @null and @yulz__
Either way it's just for my local machine and I only reboot every other month, so I can probably live with starting nginx by hand :)
Anyone else getting 500 errors from the W3C hg? cc @richtibbett
↺ 1
@w3c_systeam Pubrules seems to be very flaky (yesterday too). Most times it times out loading just the beginning of the page.
Replying to @AAlaphilippe
vous galérez en Charente? Je viens de monter dans le tgv à la Rochelle :-) Bises !
When I grow up I want to be Birgitte Nyborg.
Replying to @oliviertesquet
ils ont littéralement été créés pour des alcooliques !
Replying to @matreq
Même pas de la saison un, mais elle est quand même TROP FORTE!
Is it me, or does it seem like there isn't much happening around Couch Apps?
@brucel So now I'm in j.mp/MoB3VF%0A I hope there'll be nice champagne otherwise it's all your fault ;)
Set up neat backup-over-FTP system on server. Set up powerful iptables rules to protect server. Guess what you notice months later? :)
Discover and talk to HTTP-based services on your home network using mDNS, DNS-SD, etc. from JS w3.org/TR/discovery-a… #W3C #DAP
♥ 1↺ 1
I love how Smashing Hub lists "Hyperlinks" as one of the top design trends of 2012. j.mp/NnzmbI%0A
Support for anything remotely to do with ReSpec version 1 will be completely discontinued by TPAC 2012. You've been warned: upgrade! :)
Sorting out AppCache is *hard*. Why do people care about multi page apps again?
The sheer amount of hacks people use to get AppCache to work is simply staggering. And we've only looked at two apps so far...
↺ 2
@4aa13 Actually, no one likes it as a solution...
Replying to @kaihendry
Yeah, I know you can do cool stuff with it — it's just that it's full of traps :)
Things aren't *really* funny until you start mixing AppCache with @srcset. Ooh, ooh, can I please specify that one?
Replying to @Ls01
en même temps ils nous avaient déjà fait le coup il y a trois ans...
Replying to @Ls01
quand tu dis médias français tu penses à Voici ?
Replying to @Ls01 and @Romain_Pigenel
j'y crois moyen à tordre le bras de BO. Il gagne, il a 4 ans pour te le faire payer très cher.
Si la Hadopi avait été futée, elle n'aurait pas tenté de surveiller internet. Des contrôles dans le train seraient bien plus efficaces.
Replying to @Ls01 and @Romain_Pigenel
le risque existentiel je doute qu'ils y croient. Ils ont la théorie des jeux en Iran aussi! Ça pue l'électoralisme :-)
@w3c_systeam Can someone reboot the dvcs server please? It's 500ing again. Thanks!
Replying to @Ls01
Si j'étais riche, j'achèterais les droits sur son histoire :)
Replying to @k_narre
Sur les sujets qu'elle veut, mais uniquement pour ses fidèles et sans vouloir s'imposer aux autres.
↺ 1
Replying to @k_narre
Pourquoi seraient-elles plus contraintes que nous? Après, on pourrait y conditionner l'aide publique.
I wrote my first app with @angularjs — it's very much amazing. I highly recommend JS geeks give it a spin.
♥ 2↺ 1
Replying to @null and @xquery
It handles data binding very smoothly, you don't get lost in layers of weird syntax as with Knockout.
Replying to @null and @xquery
It manages models in a far more JavaScriptish manner (you just set attributes) than Backbone, with smart deferreds.
Replying to @null and @xquery
And Inversion of Control rocks my socks.
Replying to @null and @xquery
I also think that it would be very easily to generate code exposing a CouchDB, at least for the basics of a CouchApp, with it.
The email address robin@robineko.com is now officially retired.
Replying to @lhirlimann
That works, but the primary is firstname@lastname.com :)
Replying to @null and @xquery
The same old firstname@lastname.com!
Very interesting ideas: "A proposal for preloading resources bound by media queries" j.mp/TECd4C%0A
@heycam When you use [ArrayClass] do you magically get a [Constructor] from Array too?
@heycam In 4.3.7. [NamedConstructor] paragraph 4 starts with "The [Constructor] extended…" — I think you mean [NamedConstructor].
Replying to @null and @koalie
I don't know, but it seems to work now!
Replying to @JeniT
Right, in the maybe five times that I checked it out and wrote it off as a bad idea that was one of the red flags.
Replying to @JeniT
But they make a good case that component-extended HTML is the future and that's an idea that I fully agree with.
Replying to @JeniT
Also, there are alternative syntaxes that you can use if validity is important (something I never much agreed with ;)
Replying to @JeniT
The question is: why require that of everyone?
Replying to @JeniT
I honestly actually don't! I don't think that document-based validity is a useful constraint on applications.
Replying to @JeniT
Applications are bundles — they are by nature meaningless if you remove some of their resources.
Replying to @JeniT
The problematic impact depends in part on how widespread it is, yes (though that's independent of whether it's Google). Libs converge
Replying to @JeniT
More importantly, given declarative component bindings you can discover what is extended such that no one is tied at all.
Replying to @JeniT
HTML's extensibility should not be limited to W3C/WHAT, and yet namespaces are out. So what else is there?
Replying to @JeniT
Partly yes: language ergonomics matter a whole lot! But also, using data-* only solves the collision issue for standards, not devs.
Replying to @JeniT
data-* is great for a few small parameters but it's not an extensibility model. HTML must be extensible by all.
Replying to @JeniT
it's the XML vision done differently, who knows maybe even right :-) Cross-pollination and all that!
Replying to @heycam
OK, thanks!
Replying to @JeniT
If you know of a way of getting ng: I'm all ears. Then ng- and data-ng- are just the same. So given a component model with which...
Replying to @JeniT
...to declare things, yes, it's the best model I can think of. Alternatives welcome!
Replying to @jirkakosek
And adding Web Components to the mix you get a front door to DE. That's what NG is experimenting towards, and nicely too. @JeniT
Replying to @jirkakosek
What I'd like is for Components to be identifiable in such a way that they can be later supported directly by the UA... @JeniT
Replying to @jirkakosek
…in such a way that the UA does not need to load the code if it's a recognised component. This gives DE + evolution. @JeniT
Operation Let's Build a Goddamn Tesla Museum theoatmeal.com/blog/tesla_mus…
I donated to the goddamn Tesla Museum! indiegogo.com/teslamuseum?c=…
↺ 1
Hey, if there's somewhere I can sign up for Ecuadorian citizenship, count me in.
Replying to @jirkakosek
Unless you were given a declarative way of spawning these elements to life within the context of a page. @JeniT
Replying to @null and @ndw
That's actually unfair, Angular works with ng: namespaces if you want it to! @jirkakosek @JeniT
Replying to @null and @ndw
Am I ever not? Either way, I think you'll be wanting to buy me at least one by then ;-)
Replying to @null and @xquery
Not yet, I'm waiting to confirm new Mission() before booking anything. But I'll be there no matter what! @ndw
Replying to @tobie
yes, they're marked up correctly but there's a bug in the CSS somewhere.
Replying to @Sniperovitch
t'étais passé où ?
Replying to @Sniperovitch
ben ouais mais j'étais sûr que je te followais déjà !
Replying to @Sniperovitch
ben ouais je ne comprenais pas pourquoi je voyais des retweets de toi mais pas l'original !
Replying to @Sniperovitch
je ne suis pas sûr qu'on puisse faire confiance à nos followers communs pour RT le bon :-)
Replying to @Sniperovitch
ah si en fait quand j'ai dit ça je pensais justement à @Turblog :-)
"Methods of storing data in DNA have led to densities competitive with the most elaborate forms of digital storage" j.mp/S0THIy%0A
↺ 2
DNA can store 5.5 petabytes per square millimetre, and that's without compression.
↺ 5
Replying to @Sniperovitch
Je peux l'embêter, il a rien sur moi :) @Turblog
Amazing wild lion pictures from a robotic buggy blog.burrard-lucas.com/2012/02/beetle…
♥ 5↺ 12
Replying to @null and @graouts
I did some write-back to the GitHub API, which saved data.
Replying to @null and @graouts
It doesn't do magic model persistence, but hooking it up should be trivial since models are just JS objects. Wiring to CouchDB...
Replying to @null and @graouts
…would be almost automatic. I know some people use it with persistence.js. Depends on what you want to do!
browserver: turn your browser into a Web server — browserver.org
♥ 1
Replying to @k_narre
Ah parce qu'il faut que ça serve à quelque chose?
Replying to @k_narre
Voire ça innove!
Replying to @adamretter
Yes, use the GH version, load it from w3.org/Tools/respec/r…. v1 and v2 are no longer supported :)
Replying to @adamretter
Yes, I have to broadly update the docs — planned for milestone 3.5 :)
Replying to @adamretter
BTW you're probably better off loading it off the Web than locally/from a copy, unless you have specific reasons to.
Replying to @adamretter
Thanks :) Show me your spec!
Replying to @adamretter
Aren't I always? You should put it in a gh-pages branch!
Replying to @adamretter
Yeah, as I said, load from j.mp/OAyRuK%0A and not j.mp/OAySPa%0A — the former is gzipped, it makes a diff!
Replying to @adamretter
It looks good! Only issue: I'm guessing that exquery-restxq@w3.org doesn't exist? But not a big deal for "unofficial".
Replying to @adamretter
Nowhere for specStatus=unofficial, so not a big deal. If you had a gh-pages branch I could show you ;-p
Replying to @adamretter
It should just be git checkout -b gh-pages; git push origin gh-pages :)
Replying to @adamretter
Just pushed a release with your pull request (ta!), should be out any minute now as 3.1.20
Replying to @adamretter
It can take a while to show up — I'm seeing it at j.mp/SDyhOY%0A
Replying to @adamretter
Add whatever you want! But I'm mostly away next week so it won't show up until I make a release!
Replying to @adamretter
Of course I do — and you haven't even seen the upcoming, unannounced superpowers ;-)
Replying to @adamretter
But, just to show you where the email address goes, you can see it for another status: j.mp/SDyvpl%0A
Replying to @adamretter
A pleasure! You too :)
Replying to @adamretter
Thanks, but I went another route: I made gh-pages the default branch on GH, and the equivalent to master in git flow.
Replying to @adamretter
Based on that, I actually completely forgot that I don't have a real master branch — it all flows automatically on its own :)
Vous savez s'il y a un cinéma qui passe des films pour enfants en VO sur Paris?
Replying to @MarkVanParijs and @MarcVanParijs
On ne sait jamais, je vais tenter merci :)
Replying to @MarkVanParijs and @MarcVanParijs
Ouhla, les miennes sont en permanence fermée ;-) Je regarde plutôt leurs emails!
Replying to @MarkVanParijs and @MarcVanParijs
Oui je sais, mais ils font ça le soir — les mêmes films passent en VF la journée :) En tout cas, merci!
Replying to @wcandillon
sweet! I don't have the time right now but I'll check it out more as soon as I can!
Replying to @tobie and @janl, @dietrich, @SickingJ, @NetRoY, @mikeal
btw I'd be interested in knowing if you think it sucks for its ergonomics, or because...
Replying to @tobie and @janl, @dietrich, @SickingJ, @NetRoY, @mikeal
you find it an inherent bad match for its requisite functionality?
Replying to @Sniperovitch
si tu trouves un truc bien qui fait ça, ça m'intéresse beaucoup !
Replying to @4wk_ and @colinbraly
span is a word in itself. Yes!
Replying to @Sniperovitch
ah oui ça irait peut-être plus vite de l'écrire directement :-)
Replying to @NetRoY
the APIs are meant to enable as many approaches as possible so that people can experiment with different higher level interfaces.
Replying to @NetRoY
if the spec is high level and poorly done, you're screwed; if it's low level you can build anything on it. Then v2 can optimise...
Replying to @NetRoY
for common usage. That said, some things can certainly be improved. Just not the low-levelness.
Replying to @NetRoY
Notably, I've argued for Node style callbacks for a while, mostly to indifference.
Replying to @NetRoY
yes, splitting your feedback in ergonomics and core issues would be great. Cc @SickingJ
Le célèbre maki corse.
Replying to @janl
if you don’t think IDB is good for every day hacking you're right. So go build the awesomest lib atop it; or show how it prevents it.
Replying to @null and @sgalineau
I've still no clue why I'd want to work for any large company - what's that make me?
Replying to @_richtr and @richtibbett
yo dawg I heard you like grammatical mistakes...
Replying to @null and @sgalineau
is that how they caught you?
If you're a Web dev, that you want to relocate to Napa, and drink lots of wine: us.nakedwines.com/IHateComicSans
Si PHP est si génial et va si bien, pourquoi faut-il tant d'articles pour en prendre la défense? #jdç...
Any suggestions for a *lightweight* JS bar graph lib (pretty, accessible)? I've been using Tufte, but it's quite heavy for just bars.
Replying to @null and @karlpro
Well, Tufte brings in Raphaël which is ~80K, plus about 10K for itself. Lightweight is "substantially less" :)
Replying to @sressler
Flot looks nice, but it's not all that small either (and it's canvas-only). Thanks though, I'll check it out more.
Replying to @parb
Heh, yes, it's a very nice toolkit for advanced stuff, but building even for just BarGraphs comes out pretty big ;) Thanks!
Oh yes, because "pwah" is such a better name than "widgets". I still thing Web Runtime Application Package (WRAP) was better :)
♥ 1↺ 1
Replying to @Turblog
CSS ou JS ? Si tu fouilles les archives de JavaScript Weekly et HTML5 Weekly y'a pas mal de choses je pense.
↺ 1
OpenStand (open-stand.org) is good politics; that it doesn't go far enough just shows there's still work to be done.
↺ 1
That said, there's room for writing the NextStand, describing where we would ideally be.
I am stepping down as chair of DAP and CoreMob. It was great to work on those projects, I will miss the people. New adventure coming though!
↺ 1
Replying to @k_narre
Hmmm, la sincérité compressée sur 140cc ça peut donner cette impression — mais ce n'est qu'une impression!
Replying to @poulpita
I think that by TPAC you will know what it is :)
Replying to @k_narre
Chambre plutôt les vrais annonces, comme j.mp/O2gGj5%0A :) Même si ce n'est pas non plus de la LDB.
Replying to @k_narre
Enfin, on n'a pas tous une photo du Ministre du Redressement Productif dans son portefeuille hein #LangueDeBois :)
Replying to @k_narre
Ben maintenant qu'il a révélé qu'il est à la solde du grand capital (ce dont je n'ai jamais douté)… ;-)
Replying to @anssik
Thanks to you for being an awesome editor :) I won't participate much, but I'll be around. See you at TPAC?
Replying to @dontcallmeDOM
Thanks :) Yes, I look forward to that too :-D
Replying to @_richtr and @richtibbett
Thank man — we'll be in touch I hope!
Replying to @k_narre
Il est pas favorable aux entreprises — juste à celles du CAC40. Mais ça c'était déjà dans son programme.
Replying to @jorabin
Thanks!
"Violating WebArch makes you feel like a standards-badass" --@marcosc
♥ 1
Replying to @k_narre
Je n'ai jamais compris qui la CGPME était sensée représenter — pas de PME que je connaisse!
Replying to @anssik
Yay for beer in Lyon!
Replying to @vgholkar
I sure hope that we'll meet again — come to TPAC!
If you're getting weird permission errors with node-gyp, "mkdir ~/.node-gyp; chown -R nobody:nobody ~/.node-gyp" seems to do the trick.
Replying to @dougturner
thanks, and thanks for your contributions! I hope we can get together for a beer one of these days :-)
Replying to @girlie_mac
thank you! See you maybe at TPAC?
Replying to @dougturner
deal. I'll get the next one, and then we can see how far we can count ;-)
Replying to @tobie
build a PhoneGap app that's just the wrapper you need?
Replying to @tobie
I think people are just lazy and test in the browser :)
Replying to @tobie
Yeah, I know, I'm just saying I'm pretty sure people who just want to try something out do it in the browser.
Replying to @fabricedesre
it's to motivate you to make it more stable, and live updatable :-)
Replying to @masinter
I guess with the advent of Perl some twenty years ago?
Replying to @brucel
dude, don't add node_modules to your path, just sudo npm install -g to make it globally available!
Replying to @brucel
also, why would you want to manually install something to node_modules? Other than that, great stuff, will try it!
Replying to @fleurpellerin
Ne craignez-vous pas qu'en soutenant la recherche par la paperasse nous n'obtenions une innovation de grattes-papier?