"How We Measure Standards (and why it’s sort of a problem)" by John Jansen medium.com/@thejohnjansen/how- (/ht @tomayac)

I worked with John on the Edge team, and I was always impressed with his integrity. This is a good post about how browsers approach "standards," and how controlling the narrative is often as important as actually fixing interop bugs.

The truth is that the actual stuff that makes browsers interoperable is downright boring. It's so boring. It's stuff like "this website broke because it was queuing requestAnimationFrame and setTimeout callbacks and measuring the duration between the two, and browsers were subtly different in their timings."

No browser vendors talk about this, because it would put you to sleep. Instead they talk about "standards," which are often niche, wonk-y features that no website is actually using.

Afficher le fil de discussion

@nolan Why would you adhere to standards when browser vendors have a history of not following them or doing partial implementations? Which, admittedly, leads deeper into the whole chicken and egg scenario.

@kirot @nolan There have been disagreements over the years on implementations and always will be. Today partial implementations can arise from disagreements (web components) or from the complexity of it (css grid) or even because vendors decide to implement features long before the specification is even remotely stable (flexbox). Things were much worse before HTML5. The real danger today comes from where Google can practically dictate the platform just as Microsoft did years ago.


Things were much BETTER before #HTML5.

The #Web was still mostly a #HyperText distribution medium, we were all focusef to build a #SemanticWeb, standards were simpler you could build a compliant browser without #Google's money, and JavaScript was not the burden it is today.

And to be honest I don't think Mozilla is fighting a battle. They exist so that Google cannot be sued as a monopoly.

@kirot @nolan

@Shamar @kirot @nolan A large part of the HTML5 specification outlines additional elements, their semantics, and semantics for elements from prior versions. HTML5 began because of the div soup many people’s websites were. The rest contains the boring stuff — an extremely detailed instruction on how to build a compliant parser. Prior to HTML5 even the workings of a browser’s parser varied wildly, especially when tokenizing a document.


#HTML5 begun the raise of the #Web as an application platform.

Today we know it's broken, rotten to the core: bugzilla.mozilla.org/show_bug.

Fixing it is simple: we should turn it back to a #HyperText medium.

But even the obvious and cheap mitigations described in the vulnerability report are not going to be introduced.


@kirot @nolan

Inscrivez-vous pour prendre part à la conversation

Le réseau social de l'avenir : Pas d'annonces, pas de surveillance institutionnelle, conception éthique et décentralisation ! Possédez vos données avec Mastodon !