Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Layout 2013 and Layout 2020

servo.org

Post reply on HN
Print view View original

71–80 of 114 posts

Re: Layout 2013 and Layout 2020

#71

Post by Dalewyn »

madeofpalk wrote: post #69
Dalewyn wrote: post #68

Earlier quoted context omitted.

No, backwards compatibility is the responsibility of the browser, a la quirks mode[1][2] in ye olde days. HTML6 or CSS4 could have a fresh, clean start and there's nothing preventing that. [1]: https://en.wikipedia.org/wiki/Quirks_mode [2]: https://www.quirksmode.org/css/quirksmode.html

What would be the point of that if no browser would support it? See also, https://wiki.csswg.org/faq#versioning-css-fixing-design-mist...

I'm not sure what you're getting at?

The browser landscape today is composed of practically just one browser: Chrome. It is easier than ever to get browsers (read: Chrome) to implement different rendering schemes based on identifiers or descriptors provided by web devs.

Supporting the latest and greatest is a great marketing ploy, browsers (read: Chrome) are incentivized to support them. Browsers that don't will be forced to support them lest they become even further irrelevant.

Quirks Mode happened, Quirks Mode V2 can happen.

Re: Layout 2013 and Layout 2020

#72

Post by zarzavat »

illiarian wrote: post #48
zarzavat wrote: post #41

Earlier quoted context omitted.

I completely agree. The amount of FUD Mozilla spread about Web MIDI was truly distasteful. People say that Google is the enemy, and perhaps they are. But at least Google does not write off entire groups of users (like musicians) because of a swivel-eyed security paranoia. If I wanted a paternalistic entity telling me what I can and can’t do with my device, I’d use an iPad.

> The amount of FUD Mozilla spread about Web MIDI was truly distasteful. As in: everything they said is true, and the moment they launched it they found it's used for fingerprinting (and Google doesn't even hide it behind a permission prompt)

Anything can be used for fingerprinting. Your GPU can be used for fingerprinting. Your fonts can be used for fingerprinting. MIDI is so far down the list of fingerprinting threats.

If Mozilla are really serious about fingerprinting then they need to remove right now and make every website render in Times New Roman.

Fingerprinting cannot be solved by disabling browser features in a standard browser. It can be mitigated by using content blockers such that the fingerprinting code never runs, or by using a specialist browser like the Tor browser.

Re: Layout 2013 and Layout 2020

#73

Post by flohofwoe »

ksec wrote: post #31
linuxandrew wrote: post #7

Servo is now under the Linux Foundation umbrella. It's refreshing to finally see some more players in the browser engine space than the tri-opoly of Google (Blink), Apple (WebKit) and Gecko (Mozilla). LibWeb (Ladybird Browser) is another browser engine, but also in very early days of development. Gecko is depressingly tied to Firefox and not easily embedded. Blink has somehow become the de facto browser engine thanks…

> Even Qt (QtWebEngine) is now based on Blink instead of WebKit. This is news to me. When did that happen ?

IIRC QtWebEngine was introduced in Qt 5.4 and QtWebKit was removed in 5.6.

(and it sucked, because suddenly a Qt application with a web view was several times larger)

Re: Layout 2013 and Layout 2020

#74

Post by Rexogamer »

kelnos wrote: post #54
waboremo wrote: post #34

Earlier quoted context omitted.

Prior to app store creation, Apple and Google were the same regarding the web as a legitimate platform. After it was released, only really Google was pushing for it. Hence the creation of Progressive Web Apps, thanks to Alex Russell et al at Google.

PWAs failed in the marketplace and are essentially dead. Google pushed them because they were already on track for web dominance, and the shift toward native apps was a threat to that. These days Google still wants to control the web (and more or less does), but they also have a healthy app ecosystem, albeit only for Android. I'm sure it still bugs them to no end that they have zero control over the app experience on…

"PWAs are essentially dead"?!? They haven't replaced native apps for sure but I think it's a pretty major stretch to call them done and dusted

Re: Layout 2013 and Layout 2020

#75

Post by flohofwoe »

rektide wrote: post #21
linuxandrew wrote: post #7

Servo is now under the Linux Foundation umbrella. It's refreshing to finally see some more players in the browser engine space than the tri-opoly of Google (Blink), Apple (WebKit) and Gecko (Mozilla). LibWeb (Ladybird Browser) is another browser engine, but also in very early days of development. Gecko is depressingly tied to Firefox and not easily embedded. Blink has somehow become the de facto browser engine thanks…

Honestly why would anyone do anything other than Blink at this point? For one, there's so many examples to crib from about how to do it. But more generally, I just feel like every other player in the space has adopted an adversarial stance. For sure, sometimes Safari or less typically Firefox do lead, but it's rarely by much & usually the set of capabilities overall is far far less. I don't want a monoculture either,…

Just one point, but an embedded Chromium instance is around 10x bigger than an embedded WebKit instance.

Re: Layout 2013 and Layout 2020

#76

Post by jeroenhd »

andrewmcwatters wrote: post #5

None of this would be so difficult if the CSS specifications were well defined. But they’re not, so we all have to more or less guess what the intrinsic algorithms for drawing fundamental aspects of the web are supposed to be. Such a failure of the CSS specs.

I've watched videos about implementing CSS fixes in Ladybird (https://www.youtube.com/@awesomekling) and whenever the web specs get pulled up, I find them pretty clear most of the time. There's a lot of context required if you want to read the details about one specific thing, but all the necessary context is usually linked and readily available if you need it.

That complexity does lead to bugs (and bugs in Chrome can become part of the de-facto standard if the Chrome devs don't fix the bug fast enough) but the spec itself seems quite complete in most areas. In the few areas that are underspecified, simply seeing what other browsers do usually fixes the problem as there's usually an overlap in behaviour between at least two out of three remaining browser render engines.

The basic algorithms and rendering steps are all laid out pretty well in the spec. Even Quirks Mode has a standard (https://quirks.spec.whatwg.org/), though that's far from complete as every browser has its own compatibility quirks because of browser detection and branded CSS properties.

I know CSS documentation used to be awful, but the current version of the WHATWG spec is quite readable in my opinion and doesn't leave as much room for confusion anymore.

Re: Layout 2013 and Layout 2020

#77

Post by jeroenhd »

charcircuit wrote: post #67
solarkraft wrote: post #38

Earlier quoted context omitted.

Well, a fast and reliable browser delivers value. However there are still quite a lot of much, mich easier wins Mozilla has either ignored for 20 years or actively walked back on.

Sure, but will improving layout performance have the biggest impact for users, or was the current one good enough?

The main reason many people abandoned Firefox was because everything about Chrome was faster and more responsive. Render engine performance matters more than implementing APIs that one or two websites may ever use for anything other than browser fingerprinting. Servo was abandoned by Mozilla but the improvements to the browser engine that did make it into Firefox have sped up the browser significantly. They were the reason for a whole bunch of "check out Firefox it's fast again" posts all across the internet.

I care more about my browser being fast than I do about it supporting WebMIDI or WebSerial and I think most users agree with that. Performance and efficiency are also the reason (as far as I can tell) that macOS users stick with Safari.

Re: Layout 2013 and Layout 2020

#78

Post by illiarian »

zarzavat wrote: post #72
illiarian wrote: post #48

Earlier quoted context omitted.

> The amount of FUD Mozilla spread about Web MIDI was truly distasteful. As in: everything they said is true, and the moment they launched it they found it's used for fingerprinting (and Google doesn't even hide it behind a permission prompt)

Anything can be used for fingerprinting. Your GPU can be used for fingerprinting. Your fonts can be used for fingerprinting. MIDI is so far down the list of fingerprinting threats. If Mozilla are really serious about fingerprinting then they need to remove right now and make every website render in Times New Roman. Fingerprinting cannot be solved by disabling browser features in a standard browser. It can be mitigate…

> Anything can be used for fingerprinting.

Yes and that's a major issue

> If Mozilla are really serious about fingerprinting then they need to remove right now

Ad absurdum is not as great an argument as you think it is

> Fingerprinting cannot be solved by disabling browser features in a standard browser.

It also shouldn't be facilitated by just blindly turning them on without propert mitigation. And proper mitigation is complex

> It can be mitigated by using content blockers

So now you're shifting the responsibility onto the user. Even though it's been shown time and again that users can't really understand all the complexities of modern systems, their capabilities and the far-reaching results of what these systems can and do.

Re: Layout 2013 and Layout 2020

#79

Post by fanf2 »

linuxandrew wrote: post #7

Servo is now under the Linux Foundation umbrella. It's refreshing to finally see some more players in the browser engine space than the tri-opoly of Google (Blink), Apple (WebKit) and Gecko (Mozilla). LibWeb (Ladybird Browser) is another browser engine, but also in very early days of development. Gecko is depressingly tied to Firefox and not easily embedded. Blink has somehow become the de facto browser engine thanks…

There is also the Flow browser by Ekioh https://www.ekioh.com/flow-browser/ which is closed source and marketed for use in set-top boxes and smart TVs.

Re: Layout 2013 and Layout 2020

#80

Post by tannhaeuser »

Dalewyn wrote: post #68
MBCook wrote: post #19

Earlier quoted context omitted.

Isn’t CSS, like so many things, limited by backwards compatibility? They can’t clean slate it and call it CSS 4, no one would bother. Everything they do has to work with the old stuff, or be very explicit that you’re changing the rules (like adding “display: flex”). I suspect there is also a fair amount of “this is what everyone did so try not to break it” in there too, much like HTML. I’m not sure anyone could make…

No, backwards compatibility is the responsibility of the browser, a la quirks mode[1][2] in ye olde days. HTML6 or CSS4 could have a fresh, clean start and there's nothing preventing that. [1]: https://en.wikipedia.org/wiki/Quirks_mode [2]: https://www.quirksmode.org/css/quirksmode.html

What's preventing that is the large installed base of sites on the 'net. Quite possibly, the majority of web content has already been created (1), and it's getting worse as old sites are shut down.

Also, Safari is a thing (and personally I prefer FF for webdev, though Moz is certainly working hard to turn me away).

(1) those worth reading anyway, rather than generated content

Post reply on HN