October 2016

357 tweets

Replying to @seanjtaylor and @vagabondjack

Someone should make a metamodel :-)

Replying to @tabatkins

The next step would be the ability to operate on split documents and cache what's needed to only reprocess changed parts.

Replying to @stevefaulkner

No one was ineffectual man, I just needed to do something else!

Replying to @frivoal

Indeed, but within the realm of things that we have power over it's a good inspiration. @amyw3c @marcosc @w3c
This isn't new, but it totally made my day. youtu.be/-qTIGg3I5y8

Replying to @stevefaulkner

Awwwwwwww... I miss you too man! 😘 💋
In case you were thinking of renting with @Enterprise, be advised that they can just decide to bail and leave you completely stranded.

Replying to @null and @amyvdh

Hahaha, that's too sweet! I'll try to wash my hair more often. @stevefaulkner

Replying to @null and @amyvdh

Just call me Strider ;-) @stevefaulkner

Replying to @enterprisecares

My experience was that I booked online with my French driver's licence. Then when I showed up they said I needed a passport

Replying to @enterprisecares

This left me stranded at the train station with my kids, and no other agency there. Had to pay $55 to go to the airport.

Replying to @enterprisecares

No no, let's have this conversation in public. You screwed me over for no reason, no need to make this private.

Replying to @enterprisecares

Thankfully none of your competitors have this sort of asinine policy, so we could rent with them.
"This might be wrong", an artist's view of the early universe. #MassMoCA
↺ 1
WTF Columbia? 😿
↺ 1
Why do American trains honk all the time? 🚂

Replying to @tabatkins

Is there even a vague attempt at a justification for it?
"Information and its Role in Nature", JG. Roederer. The 1st chapters are good, then it goes all teleological/anthropic on you. Please don't.

Replying to @gregwhitworth

Well played. @tabatkins

Replying to @kzwa

But... But... That’s crazy!

Replying to @kzwa

Well, those intersections all seem to have closing barriers, so why honk? It can only create nuisance for passengers and neighbours.

Replying to @gregwhitworth

Heh, indeed! Reminds me of a Parisian friend who always wondered why Japanese tourists took so many pictures. He spent...

Replying to @gregwhitworth

@gregwhitworth... five days in Tokyo, I think he came back with 8000 pics 😊

Replying to @leonor_de_b

Han. J'attends encore le jour où je ne serai plus "Robert" ou "Romain".

Replying to @_aurelienfabre and @_noemes

Mec, un jour il faudrait que quelqu'un fasse une série sur les think tanks américains. Tu peux tenir douze saisons sans problème!
They're talking about @dat_project in Nature :-) nature.com/news/democrati…

Replying to @stevefaulkner

So... it doesn't mention stabbing developers in the face when they do that?

Replying to @domenic

Are those the blocking factors? I can’t think of a single SDO with decent inclusiveness. @slightlylate @zeigenvector @littledan

Replying to @hillbrad

Random idea: could this be extended to ranges? Say I've bundled a popular lib and it's at a given range in my bundle. Given a…

Replying to @hillbrad

…Range RI of sorts craft an HTTP request that only loads the parts of the bundle that are actually needed. The security…

Replying to @hillbrad

…implications seem relatively similar. Given today's major development practices, it could provide for a nice win.
Does anyone know why Google’s software update daemon (ksfetch) started hitting auto.schema.org today?
↺ 1

Replying to @hillbrad

It would, but I recently analysed one of my bundles and my own code represented only 7% of the whole. So ranges would be sweet :)
Three minutes in we already have open future from Kaine and scaremongering from Pence. #VPDebate
Is it me or is Kaine talking about Trump too much? #VPDebate
I think I’m still processing Kaine’s gun owner + 2nd amendment + Virginia Tech + Holocaust survivor + Russian collapse survivor. #VPDebate

Replying to @bulldog6112

Well, no, it’s a completely coherent string of facts, it’s just a lot at once.
Is it just me or would it not be a better idea if Kaine spent the time making himself likeable instead of rehashing Trump? #VPDebate
↺ 2
«Criminal aliens» says, Pence. Here’s how I would go if I were Kaine. #VPDebate youtube.com/watch?v=FJoBWi…
Pence repeats Trump’s claim of ICE endorsement, which is a lie. Why isn’t Kaine stating that? #VPDebate
The «have you prepared that?» line is great if you’re winging it and are up against any cogent argument. Expect more of it. #VPDebate
Is Pence saying «Alepp*a*»? #VPDebate
Wait, Pence is saying the US should be ready to strike Russian military targets in Syria? #VPDebate

Thread of 2 tweets

You’ve got to give it to Pence, he’s really good at pretending he’s never heard Trump speak. #VPDebate
↺ 1
Ten minutes to the clock Pence brings up the foundation — he’s doing this absolutely by the book. #VPDebate
Pence is a pro, if he were at the top of his ticket Republicans would be in the lead. #VPDebate
Ah, yes, the famous Trump Organisation. #VPDebate
In North Korea, watching the #VPDebate… youtube.com/watch?v=PHTVuP…

Replying to @briankardell

Also, it gives your opponent airtime.

Replying to @briankardell

Yeah, we have 2500 years of experience running debates; you’d hope the commission took some ideas from that.
Okay, I’m tuning out. I hope Clinton does better on Sunday. #VPDebate
So @timkaine did little more than than give airtime to Trump's ideas while Pence addressed the rest of the right. Really bad #VPDebate.
With both candidates hailing from New York you'd think one would address the fact that it's October and we still have mosquitoes.

Replying to @zeldman

There’s something like ~110 elements total in HTML, how could 5 possibly have added 121? @adactio

Replying to @ShaneHudson

Even then that would be a stretch on the idea of «adding». And you’d have to count the on* attributes. @zeldman @adactio
"If you're not a citizen of the world, you're a citizen of the twentieth century." --Ben Wermuller, via @torgo
↺ 6

Replying to @eviltrout

You realise that this would be a whole lot more convincing if written up as a blog post in which two made up characters converse?
An evening of free drinks of your choice to the first frood to use Houdini to make `text-transform: --chicago;` work. #CSSFTW

Replying to @tobie

Did I mention it’s not just the drinks but I’m also there keeping you company?

Replying to @null and @amyvdh, @wseltzer

I just love how good the targeting is.

Replying to @mzabriskie

A list with search, add, delete?

Replying to @annbass

Thanks! I was copying from memory after having seen it on Facebook :-) @benwerd @torgo
Users: «I want to type a title, roughly, and find an article.» Startups: «Let’s build a crazy complicated article recommendation system!»
Quoting a tweet by @Gene_Regulation ↗

Replying to @Sniperovitch

Ben d’ailleurs ils lui ont un peu fait payer…

Replying to @Sniperovitch

Nan, mais bon, c’est pas comme s’ils s’éloignaient beaucoup de la réalité historique :) @Languedeuxpute

Replying to @Sniperovitch

Oh, attends, c’est pas comme si on te parlait du truc surprenant avec Javi! @Languedeuxpute
New chapters.io chapter setting up in Ottawa: if you want to change the face of Web standards, go! meetup.com/Ottawa-Chapter…
↺ 4
Anyone in NYC interested in Web standards? Should we kick-start a local chapters.io?
↺ 3

Replying to @TzviyaSiegman

Run. Now. Do not take your belongings with you. Never look back. @w3c

Replying to @TzviyaSiegman

Of course. Or at least, so long as you don’t bring composition events along. @w3c @frivoal

Replying to @frivoal

It’s always useful to get updates on that situation, even if painful. @glazou @TzviyaSiegman @w3c

Replying to @glazou

You’re right, it’s a bit of a weak word :) @frivoal @TzviyaSiegman @w3c
Today in science stock photography: this equation is truly not for the faint of heart.
↺ 6
@peerus You mention indexing based on «semantic marks». I’m curious: do you document what markup you look at in an article.

Replying to @iphigenie

It is too complicated for us mere mortals. Who can even write that neatly on a glass board? Only the gods. @pjcjohnson

Replying to @vambenepe

Yes, it had me completely on edge for, like, all of those first two lines. It’s nailbiting stuff.

Replying to @fantasai

You heard that @briankardell?

Replying to @fantasai

I know, but they do that because they don’t know they can send it somewhere and be listened to. We can change that. @briankardell

Replying to @briankardell and @fantasai

The chapters page could probably be improved, it was banged together in a hurry.

Replying to @briankardell and @fantasai

Maybe writing up some principles (like here discussed), links to existing chapters, etc. would help at this point.
To second what @slightlylate said, @MSEdgeDev is an awesome team of wonderful people. If you want to help the Web, it's a great opportunity.
Quoting a tweet by @somelaniesaid ↗
↺ 2

Replying to @judithd

I think he's in a gender of his own, he's definitely not cis-human.
Finally got around to "The World Without Us", Alan Weisman. Great theme, lots of interesting tidbits, but overall quite journalistic.

Replying to @leonor_de_b

Je commence à penser voter EELV. Pas forcément dans un fol enthousiasme, mais bon.
"Histoire des sciences et des savoirs: de la Renaissance aux Lumières". Une belle perspective historique sur une période foisonnante.
↺ 1

Replying to @fantasai

I gave you access to the GitHub :-) @briankardell
Kids! Back in my day we had sites depend on VBScript; no one got flak for it.
↺ 2

Replying to @Lucid00

Sure thing, but we had fun anyway :-)
I hesitate between loutrage and floutrage. Tough choice. Maybe one can feel both simultaneously?
Quoting a tweet by @willf ↗
↺ 1

Replying to @mrgunn

I am not certain that they all speak to the same openness. But it would be interesting to dig deeper.

Thread of 2 tweets

Replying to @mrgunn

But access does not strike me as core to the new open/closed political divide. More about change and exchange.
↺ 1
So, horseplaining is now a thing?
Quoting a tweet by @horsplain_js ↗
So yeah, I've been dropping some beets.
I'm joining late, what the hell happened to Trump's hair?
C'mon, someone slam him for lying over ICE endorsing him again. #Debates2016

Replying to @slightlylate

But we have a GDP of 1%, and in my opinion it's much lower much lower!
↺ 1
Ah, a Holocaust related question. Shame Trump is on stage and can't retweet some good antisemitic comment right now.
Confession: whenever I hear "Ivanka", I think "ah, vengeance daemon".
"The invisible politics of Bitcoin: governance crisis of a decentralised infrastructure" policyreview.info/articles/analy…
↺ 2

Replying to @Peerus

The latter, whether you take advantage of embedded metadata and if so whether you do RDFa, Microdata, JSON-LD, etc.

Replying to @jwajsberg

But but! Four times one times one! That's hard!

Replying to @briankardell and @wilkieii

What was it about beyond JSON-LD?

Replying to @fgeorges

Not that I know of yet, @briankardell might have something. We could use some help maintaining the info and all.

Replying to @briankardell

Sorry, I was supposed to open an issue? @fgeorges

Replying to @fgeorges and @briankardell

I don’t have much bandwidth to help ATM, but if you want to look into a Paris Chapter, talk to @ParisWeb!

Replying to @yoavweiss

I once was screamed at for asking Mozilla’s position that way in a meeting. People have no taste for music. @jaffathecake @beep

Replying to @null and @marcosc

It’s almost as if you were… challenging me to share some way worse songs I might happen to know… @yoavweiss @jaffathecake @beep
Is anyone else getting spam follows from accounts affiliated with AfD, notably @AfDBerlin? German friends? Most recent: «balleryna».

Replying to @gimsieke

OK. I'm mostly wondering if they have some weird social networking strategy that would be worth warning about.

Replying to @gimsieke

LOL, you’re right, that specific piece of sarcasm was lost on some people in French so it probably does not translate so well :)
Vous savez si c’est possible d’attaquer sa banque pour défaut de sécurité sur son site? Je pense à @BanquePop.
Today in alternative histories: what if Luke Skywalker had not joined the Rebellion?
Quoting a tweet by @gordonbrander ↗
Does Section 508 apply to research articles produced from public funds?

Replying to @anssik and @richtibbett

While it’s very impressive, 200K is still huge. Pitching the workshop on what would reduce that would be useful IMHO. @anssik

Replying to @null and @mattmay

Between PDF and publishers flocking to ReadCube it looks like no one is enforcing that, then :(

Replying to @auchenberg

Isn't this a rehash of the same discussion except with jQuery ~10 years ago?

Replying to @null and @mattmay

Is it measured using one of those weird mathematical quantities that don’t map to an actual number?

Replying to @null and @mattmay

Incidentally, have you looked at ReadCube a11y? I don’t feel comfortable slamming them too publicly because competitors, but man…

Replying to @null and @mattmay

American friends: if you’re suffering from campaign fatigue, take a moment to think of the French whose primaries have only just started.
I wonder what it would take for @BrooklynBrewery to consider making a local version of Marmite/Vegemite.

Replying to @auchenberg

Do you have a specific example in mind? I’m still failing to see how it’s structurally different.

Replying to @null and @mattmay

And it’s going all the way to June… Considering people there follow the US elections, I think they’re going insane. I am.

Replying to @null and @mattmay

It’s… not as different as you might expect.

Replying to @auchenberg

I don’t think that’s specific to React, there’s just a lot related to it because it’s such a common framework.

Replying to @auchenberg

People are trying to figure out how to package reusable+orthogonal bits, including CSS. Lots of innovations and searching.

Replying to @auchenberg

On the Web side we talked about scoped styles for years before there was movement. This will feed back into the platform.

Replying to @jwajsberg and @mattmay

We have several Trumps though, it does make things more confusing.

Replying to @goetsu

Well, it's basically everything that's wrong with PDF brought to you in HTML. @mattmay

Replying to @tbernard1979

On a au moins Baylet, mais ça m'étonnerait que ça soit le seul.

Replying to @auchenberg

The Web treats its competition as appetising and absorbs it :-)

Thread of 2 tweets

Replying to @null and @amyvdh

Ah, I hadn’t thought of Keynote, thanks! I wonder if it has enough doodling that it’d make her happy.

Replying to @null and @amyvdh

I need to figure out where she can best make doodles, but that ought to be easy! Thanks!

Replying to @null and @amyvdh

Well, it’s mostly that she keeps creating books with loose paper and pencil. They go bad fast so I wonder if she’d try this instead!
J’essaie de choisir un candidat pour les #primaires2016 par élimination, mais à la fin il ne reste personne. Que faire?
Is there a betting pool on Michelle Obama 2024?
The whole thread. The whole fucking thread.
Quoting a tweet by @WillMcAvoyACN ↗
↺ 1

Replying to @goetsu

I haven’t tested it, but I wouldn’t be surprised if it lost accessibility in translation from PDF… @mattmay

Replying to @_aurelienfabre and @_noemes

Pfffft, on a encore trop bouffé.

Replying to @MacsandMoreDH

But the type of yeast makes it taste different! @BrooklynBrewery
Issues like github.com/ianbjorndillin… really show how validation can hurt. Is it time to make span take arbitrary content? cc @sideshowbarker

Replying to @maccaw

If you need brutally small and fast you could use EXI (w3.org/TR/exi/) or even w3.org/TR/exi-for-jso….

Replying to @maccaw

Those are used in a bunch of IoT systems (and various constrained places). But it comes at a cost in complexity and debugging.

Replying to @tomdale

Yes, we use JSON-LD+framing (and similar @id/@type constraints) and the benefits of switching to GraphQL don’t seem very clear.

Replying to @tomdale

The major difference is that sometimes we overfetch compared to what GraphQL would do, but not by all that much.

Replying to @tomdale

Well it depends how much of it you do upfront, but yes. Pouch has been very sweet to us.

Replying to @tomdale

Let’s just say it’s so far down my performance optimisation list I’m pretty sure people won’t be arguing about semicolons by then.

Replying to @bbirdiman

Not really, there's plenty of stuff that may be technically possible but hard for a 7yo.

Replying to @thisNatasha

I've always been (sincerely) curious: why not wait much longer before upgrading? I do it when the next version is announced :-)

Replying to @bbirdiman

PowerPoint? I can’t think of a single problem for which PowerPoint would be a viable solution.

Replying to @null and @dauwhe

You make it sound like failing is an opportunity :) @bbirdiman

Replying to @bbirdiman

Yeah, they have superficially similar goals, but one of them is actually decent :)

Replying to @null and @dauwhe

That’s just perfect :)

Replying to @gordonbrander

Oh, sweet! That came out of research we did at @sciencedotai :)

Replying to @gordonbrander

Yes, we used to do much more AI, eg. github.com/scienceai/neoc… or github.com/scienceai/tsne….

Replying to @gordonbrander

Oh, not *my* work to be clear! I just complained about things that don’t run in browsers :)

Replying to @mnot and @briankardell, @chaals

Yes, it fixes politics, just look at Australia! 🐑ish 😁
Told you, #Obama2024 :-)
Quoting a tweet by @pascalriche ↗

Replying to @jwajsberg

Comment ça?
Looking forward to this publisher's take on culinary magazines: "Anarchy in the Soufflé".
Quoting a tweet by @janl ↗
↺ 1
@_noemes enfin!
Quoting a tweet by @mpigliucci ↗

Replying to @jwajsberg and @SamyRabih

Elle est 20 ans plus jeune mais elle a eu une belle carrière déjà. Mais on me dit qu'elle ne veut pas...
A vision of the real Paris m.vanityfair.fr/video/actu/vid…
↺ 1

Replying to @vgholkar

Il faut y retourner!

Replying to @null and @amyvdh, @koalie

But isn't November for NaNoWriMo? And growing a moustache?

Replying to @brucel

I have been meaning to see if Scotland will grant citizenship if you have Scottish ancestry. I'd apply :-)

Replying to @null and @sideshowbarker

<p> won’t close <span> man, software.hixie.ch/utilities/js/l…

Replying to @null and @sideshowbarker

To give broader context: when people build components that may accept arbitrary parents/children, they prefer span.

Replying to @null and @sideshowbarker

The reason is that it’s what has the least effect (compared to a default rendering of block for div). But users don’t see it

Replying to @null and @sideshowbarker

When you instantiate a component (using any of the frameworks) you don’t see that implementation detail. Nor should you.

Replying to @tobie

The only _justification_ that you will hear that has something like vague merit is performance. You can write more pessimal XPath.

Replying to @tobie

The only _reason_ is NIH. CSS Selectors is a strictly inferior language in every respect (including readability given feature parity)

Replying to @OriPekelman and @yannirbah, @tobie, @takeoffconf

Same for me, wonderful conference! Alas I am far from my beloved @OriPekelman these days.

Replying to @null and @sideshowbarker

Yeah that was my plan, wanted to make sure it wasn't stupid first. Thanks!

Replying to @tobie

Chocolate 🍫?

Replying to @null and @sideshowbarker

<p> parsing is such a nuisance, I’m always surprised when it doesn’t get in the way.

Replying to @tabatkins

I’ve hated it so much I’ve considered proposing a <para> element. But it seems, I don’t know. Wrong. Ugly. @sideshowbarker
Relaxing author requirements to allow arbitrary content in `span` discourse.wicg.io/t/relaxing-aut…
↺ 1

Replying to @dginev

Lifting the restriction on span is more or less a one line change in the spec and validator, and zero change to browsers.

Replying to @dginev

Adding an element is a much bigger change, for no obvious value.

Replying to @null and @dauwhe

Yes, there is nothing that will save you against that. I would argue you should get a parsing warning, not a content model one.

Replying to @dginev

But at the semantic level they have essentially no difference already anyway. This is a pragmatic decision to make, not principled.

Replying to @dginev

«Seems logical» and «HTML», man. Those things don’t go together.

Replying to @dginev

More seriously, the logic of maintaining a distributed system at that size and time scales isn’t the same as for a code project.
Stop submitting papers. medium.com/@tpoi/stop-sub…
↺ 1

Replying to @null and @marcosc, @afbarstow, @tobie

We started a how-to with @jungkees! But people didn't seem super interested.

Replying to @econstyleguide

*than
↺ 2

Replying to @cconcolato

I am well aware. Nevertheless, I thought he made a thought-provoking point.
Any good intro to group theory you would recommend?

Replying to @jeffsonstein

Thanks a lot Jeff, but I meant the *other* group theory :) en.wikipedia.org/wiki/Group_the…

Replying to @jeffsonstein

I’m sure it’s a great recommendation though :-D

Replying to @vorahsa

Thanks! I’ll check it out.
2016: scientists turn CO2 into booze at room temp, world saved! 2021: 1st deleterious effects of global cooling felt popularmechanics.com/science/green-…
↺ 3
I love the @esa live feed, everyone is speaking English with all-over-the-EU accents esa.int/Our_Activities…
↺ 1
Parental paid leave — look at 1) gender inequality, and 2) the USA. economist.com/blogs/graphicd…
↺ 1
Does anyone have a good eslint integration for those people who use emacs? Like running live, not post-facto if that’s possible.
An Important Distinction smbc-comics.com/comic/an-impor…
↺ 1

Replying to @jsscclr

Thanks a lot! Hopefully with this I can get my belovèd colleague to lint :)

Replying to @svgeesus

I always make my content have semantic capitalisation, even if CSS makes it *visually* different. I'd be angry to not get original

Replying to @xaiki and @severo_bo

Thanks a lot!
Is there some clever CSS trick to toggle back and forth between italics, notably to unitalicise `em` in italics context?
↺ 1

Replying to @null and @kytwb

Thanks, I saw that, but this is a broadly reusable sheet so that’s really not an option.

Replying to @dontcallmeDOM

Ah, I knew I remembered something! So sad! @tabatkins do you know if it may come back?
77% the français pensent que leurs députés sont corrompus. Voilà. challenges.fr/france/pour-un…

Replying to @jwajsberg

Tu es sûr que ce n’est pas un petit groupe honnête qui fait qu’on ne les pend pas tous indistinctement? :)

Replying to @jwajsberg

Je pense qu’il y a un réel problème de corruption légale. Un lobby peut influer bien plus que de raison.

Replying to @jwajsberg

Oui et non. Échanger des faveurs politiques contre des avantages c’est de la corruption; même quand c’est légal.

Replying to @jwajsberg

La frontière est compliquée. Un ami assistant parlementaire m’a montré une fois ce que son député recevait pour Noël… Hallu.

Replying to @jwajsberg

Il n’est pas rare que ce soit le lobby qui rédige la loi pour éviter du travail au député, etc. C’est pas joli dans l’ensemble.

Replying to @jwajsberg

Ah je ne sais pas, je ne connais pas Cash :)

Replying to @tabatkins

They feel more interested in decentish typography than used to be the case, so maybe there’s hope! @dontcallmeDOM
"The Philosophy of Information", @Floridi. Broad, well-researched, intelligent views on a topic that needs much more work.
I wish we could get Internet providers to advertise by uptime rather than bandwidth.
↺ 2

Replying to @null and @_paulverbeek, @klick_ass

Sadness :(
«Researchers Restore First Recording of Computer Music, Made in Alan Turing’s Lab» hyperallergic.com/329573/alan-tu…
↺ 3
I have to say, @OneTab is the best thing that’s happened on my computer since, I don’t know, a very, very long time.

Replying to @edasfr

C’est mieux parce que plus sélectif, ce que tu restores ne reste pas dans la liste, etc.
Big outage?

Replying to @null and @kplawver

Oh indeed, really bad. Thanks for the info!

Replying to @SamyRabih

Yeah, I'm looking at the outage map and I'm right in it :-)

Replying to @TzviyaSiegman

Lots, big ddos on Dyn.

Replying to @TzviyaSiegman

I guess it makes for a chilled Friday for lots of people!

Replying to @TzviyaSiegman

Yeah, @github issues should so work offline.
School: where you learn all you need to write a good RomCom.

Replying to @CT_Bergstrom

Well from a document engineering perspective both formats are just about as spectacularly backwards :) @svscarpino @SimonDeDeo

Replying to @CT_Bergstrom

Both give precedence to formatting over semantics. They're basically image formats. @svscarpino @SimonDeDeo

Replying to @CT_Bergstrom

To exaggerate a bit, it's like putting a bar chart on a gif and saying it's a great data format. @svscarpino @SimonDeDeo

Replying to @CT_Bergstrom

Nothing is perfect of course but HTML has a good mix of semantics, implementation, parsing, and style. @svscarpino @SimonDeDeo

Replying to @null and @SimonDeDeo

I'm not a hater, there's worse! I just find the notion that it's better than Word frivolous :-) @CT_Bergstrom @svscarpino

Replying to @null and @SimonDeDeo

As old is of course hard but HTML is ~27 and faring far better! @CT_Bergstrom @svscarpino

Replying to @null and @SimonDeDeo

Are there known, reliable outputs of LaTeX that are accessible eg. to blind people? @CT_Bergstrom @svscarpino

Replying to @CT_Bergstrom

As a sighted person, LaTeX fares better there, yes. @SimonDeDeo @svscarpino

Replying to @CT_Bergstrom and @SimonDeDeo, @svscarpino

To be clear, I'm not hating here. We've failed you by not yet making an easy, better, non-tech option!

Replying to @thisNatasha

That's stupid, let's build just one SDO to unify them all.

Replying to @tabatkins

In French it's what you use to lift a car and change the tyre. (You're welcome.)
The problem with fooling yourself is that you tend to be a pretty good match for your own intelligence.
↺ 1
In the (very sad) absence of an <lh> element, what is your preferred idiom to mark up a list with a title/label in HTML?

Replying to @plhw3org

No I mean in *actual* HTML :)

Replying to @null and @nickbarreto

With no wrapping element? ARE YOU INSANE MAN?!?!?

Replying to @stevefaulkner

So I considered section, but is that not annoying for document navigation? It could lead to a lot of sections.

Replying to @stevefaulkner

When I want to «go leisurely» I just take my pants off, man. I don’t, under any circumstance, rely on the outline algorithm.

Replying to @null and @nickbarreto

Yes… but… but… the outline algorithm! It’s crazy!

Replying to @stevefaulkner

It could be lots of headings and they won’t match the ToC. I’m considering <figure> instead.

Replying to @null and @notabene

🦄

Replying to @Nico3333fr

You’ll see me using an hx with no sectioning wrapper the day you pry <section> from the grasp of my cold, dead fingers brother.

Replying to @null and @nickbarreto

If something like a <p> or whatever content follows the list then it’ll «look» like it’s also under the heading, which is wrong

Replying to @stevefaulkner

Replying to @null and @nickbarreto

Not really, it is part of the content.

Replying to @innovimax

OMG all of those are so wrong :)

Replying to @innovimax

Well, the ones listed as «solution».

Replying to @null and @nickbarreto

Yes, which isn’t what I want. That’s why I’m thinking of <figure> but it’s not great.

Replying to @null and @nickbarreto

Exactly. Something about it bothers me (hence the question); but I can’t tell what. Thanks!

Replying to @stevefaulkner

Yeah, something about it bothers me but *shrug*. Sometimes I wish <label> worked on anything.

Replying to @vampaz

I don’t think that will properly defeat outlining to make the <h> not section.

Replying to @laurendw

But a <p> does not have the semantics and an <hx> will label anything that follows the list.

Replying to @laurendw

If semantics weren’t important I’d use a div ;) I don’t buy that class grants semantics, so far I’m leaning towards figure.

Replying to @laurendw

Semantics. Appearance I can always trivially fix with CSS.

Replying to @laurendw

Actually, figure only implies 1) self-contained and 2) captionable. My hesitation is only on the «self-contained» part.

Replying to @laurendw

Right, which is why I’m trying to rely on the built-in semantics rather than my own class specialisation.

Replying to @laurendw

If and when I need to specialise, I then use RDFa. But indeed YMMV :)

Replying to @laurendw

Ha! And then a while later someone like me will have to try and figure out a way to HTML it :)

Replying to @laurendw

I take the exact opposing view. Using HTML+RDFa+JSON-LD is the only way to get durably interoperable semantics.

Replying to @laurendw

When people design ad hoc schemata, they build in huge amounts of implicit semantics that go well beyond the markup and are lost.

Replying to @laurendw

I’m currently processing JATS, it’s painfully obvious there, and that’s one of the better cases I’ve seen.

Replying to @laurendw

No, but it does imply semantic unity. On mulling it over I’m leaning towards it being good in this case.

Replying to @laurendw

Yes it’s a problem: the problem being that MDN is wrong :)

Replying to @laurendw

See the NOTE in w3.org/TR/html51/grou…. Repurposability is ideal (which is where MDN comes from) but not required.

Replying to @laurendw

Yeah, in Scholarly HTML it’s used for images but also tables, code, video, audio, data…

Replying to @laurendw

Replying to @jaffathecake and @notabene

Hah, web components are like seventh tries or so. Persist!

Replying to @LeonieWatson

Ideally, both. Its semantics is that it labels the list.

Replying to @LeonieWatson

Does that work better than using figure and figcaption, the former containing the list?

Replying to @docum3nt

That gives me semantics similar to a div with a class, it’s not a big win. @laurendw

Replying to @briankardell

That still doesn’t give you actually reusable semantics — only behaviour. @LeonieWatson

Replying to @briankardell

You mean by having it generate all the appropriate ARIA and RDFa stuff? @LeonieWatson

Replying to @briankardell

Yeah, I hope some day that’s the default thing you get when parsing HTML. But not quite yet! @LeonieWatson

Replying to @null and @prushforth

I’m still sticking to figure :) @briankardell @LeonieWatson

Replying to @briankardell

Man, that escalated to nihilistic philosophy fast. @prushforth @LeonieWatson

Replying to @bbirdiman

Semantics are hard. The rest is just bits.
Possibly one of the best endorsements of @HillaryClinton so far :)
Quoting a tweet by @peterdaou ↗
↺ 39

Replying to @briankardell

You missed the joke :) @prushforth @LeonieWatson

Replying to @briankardell

THERE IS NO IS WE ARE THE ALL DOOMEDETH! @prushforth @LeonieWatson
Exhibit A
Quoting a tweet by @KSoltisAnderson ↗
Exhibit B
Quoting a tweet by @scottlay ↗
↺ 1

Replying to @goetsu

Not really, there's only ever one heading. Also, some lists are dl. Then it gets crazier...
Yay standards! *cough*
Quoting a tweet by @nst021 ↗
↺ 1

Replying to @nst021

Do you reckon this could be added to the Web Platform Test project? It could at least be instrumental in making browsers align.

Replying to @stevefaulkner

Yeah, not bad at all! Still, I think an <lh> element might be called for. I forget what happened to that discussion :)

Replying to @LeonieWatson

Right, but the more I think about it, the more I reckon if the author wanted to caption the list, it must be «special», no?

Replying to @tomdale and @rbranson

The conclusion that we're stuck is demonstrably wrong though. HTML doesn't have a version either. Slow+careful does it.
@CrossrefSupport Have you changed your API usage policy recently? We’re getting 429 for things that used to be ok.

Replying to @ckoscher

Ok, I totally understand that. But is there a documented protocol for staying below the threshold? @CrossrefSupport

Replying to @ckoscher

Ok. So if I start limiting at 10 per second now I’ll be good? @CrossrefSupport

Replying to @ckoscher and @CrossrefSupport

Thanks for the info! Yeah, we search for all citations in a paper at once, so it can be ~200 per batch easily.

Replying to @ckoscher

No worries, that’s the life of those who rely on free external services! @CrossrefSupport

Replying to @ckoscher

If you want to look at my recent hits, they’d be from 209.148.63.66. But they could also come from CI or production.

Replying to @ckoscher

It’s unlikely we’d have hit you all that much, unless something went seriously haywire :)

Replying to @gsnedders

Sweet :)

Replying to @sebbean

That’s just for humans. You’d be hard-pressed to find a version in the content. @tomdale @rbranson

Replying to @LeonieWatson

Well, this is converting from another format in which lists can either be titled or not; if the author picked titled then…

Replying to @LeonieWatson

…representing it as something captioned seems correct. That clearer?

Replying to @sebbean

There may be other incidental users too.
There I was thinking the most important thing was having cables coming out of your ears.
Quoting a tweet by @gordonbrander ↗

Replying to @TzviyaSiegman

I know. If only we knew people who worked on HTML… @LeonieWatson

Replying to @TzviyaSiegman

Would it work on <dl> too though? JATS has *THREE* (3!) different headings there. @LeonieWatson
«There’s really not that much of a difference between politics and regular marketing.» bloomberg.com/news/articles/…

Replying to @LeonieWatson and @TzviyaSiegman

I would assume embedded in the ul/ol; either first or last.

Replying to @LeonieWatson

It could be, yeah. I barely have the bandwidth this second, though. @TzviyaSiegman, WDYT?

Replying to @bbirdiman

I see no reason to believe that, no. @TzviyaSiegman @LeonieWatson

Replying to @bbirdiman

That the people working on HTML have somehow decided they don’t need this. @TzviyaSiegman @LeonieWatson

Replying to @bbirdiman

No, you’re right, it’s obviously a conspiracy. @TzviyaSiegman @LeonieWatson
I told you that vertical search is the future! 🔮
Quoting a tweet by @mathowie ↗

Replying to @phae and @slightlylate

I've been wondering why they don't do dongles as a subscription forever.

Replying to @bbirdiman

If you don’t want people to paint conspiracy, maybe you shouldn’t talk of «random accidents» to boot? @TzviyaSiegman @dauwhe
If you needed a reason to work on accessibility… youtu.be/XB4cjbYywqg

Replying to @bbirdiman

I’m not coming at you. That would be boring. I’m just pointing at the nonsense of your argument. @TzviyaSiegman @dauwhe

Replying to @bbirdiman

I find your faith in the sanity of standards… disturbing. @TzviyaSiegman @dauwhe

Replying to @538forecastbot

make it stop!

Replying to @null and @dauwhe, @TzviyaSiegman

Ah, thanks, I remember that now. I think it's a pretty weak point. Just make it listcaption to break the impression.

Replying to @null and @SelenIT2

It would, but now try to style the <p> without adding non-semantic markup :) @stevefaulkner
Twitter spam just keeps getting weirder…
Quoting a tweet by @AGI_Maths ↗

Replying to @brianleroux

I think of it more as a new XSP 😎

Replying to @cesifoti

I think "gaffe" is a slight understatement :-)

Replying to @temptoetiam

Surtout quand ils sont entangled. @cepcam

Replying to @maiolsanauja and @vorahsa

I wonder if there is a correlation to proficiency in foreign languages.

Replying to @null and @SelenIT2

No, IDs aren’t semantic and that would require generating lots of CSS! @stevefaulkner
I have a bunch of poorly produced PDFs (academic papers) I’d like to print bound, as a book of sorts. Lulu makes this hard. Any thoughts?

Thread of 2 tweets

Replying to @docum3nt

Whoa, that’s brutal! I ended up using gs — just as ancient, but more command-liney :)

Replying to @AmeliasBrain

All that I remember is that it's super easy to get the remote in a state in which it loses history. The CLI's advice is bad.

Replying to @csarven

Does it work well in the US? Back when I looked they told me it would almost certainly not operate well with local CDMA channels.

Replying to @csarven

Ah, I have fond memories of the years I spent without a mobile. People were so annoyed! 😹 But then kids, need to be reachable, etc.

Replying to @rhiaro

Ah, 2G only would be a problem given how often @TWC's Internet access is down :-/

Replying to @rhiaro

Yes, the American consumer Internet experience is a very introspective one.

Replying to @briankardell

I always thought Hendrix was kind of meh, at least compared to the legend.

Replying to @briankardell and @amyvdh

Best song of all time is Kylie's "Can't Get You Out Of My Head". Y'all just need to recognise.

Replying to @briankardell

Your Twitter is moving but all I hear is NA NA NAH. NA NA, NA-NA NA, NA NA NAH… @amyvdh

Replying to @briankardell

Though if you want to debate that with «Crazy In Love» from Queen Berjoncé, we can talk, honey. @amyvdh

Replying to @AmeliasBrain

It wouldn't be too hard to make a web version of XPointer. That would help.

Replying to @Sniperovitch

Tu exagères peut-être un tout petit peu mon attachement aux standards :) @glazou

Replying to @Sniperovitch and @glazou

En tout cas vous pouvez dire merci à @Elsa_Trujillo_ de s’intéresser à ce qu’il se passe en coulisses!
Also the writing process behind almost every single bit of «startup advice» smbc-comics.com/comic/path-of-… (via @ZachWeiner)
↺ 1

Replying to @AmeliasBrain

I meant that in the sense of a Web standard. Linking to a specific paragraph with CSS in the link.

Replying to @elrond25 and @AmeliasBrain

And you could build a flexible variant of emphasis atop github.com/scienceai/web-….

Replying to @TzviyaSiegman

It hasn’t been advocated to the right people? @AmeliasBrain

Replying to @AmeliasBrain

I’m retired, I’ll leave that up to @shepazu :) @elrond25

Replying to @elrond25 and @AmeliasBrain

You have a group already, just run with it! w3.org/community/csss… @shepazu

Replying to @ShaneHudson

Thanks :) We’re not looking right now, but it ought to happen at some point. Will keep that in mind!

Replying to @ShaneHudson

BTW your «Hire Me» link doesn’t seem to work!