Intellectual honesty and html5(0xdeadbeef.com)
0xdeadbeef.com
Intellectual honesty and html5
http://www.0xdeadbeef.com/weblog/2010/06/intellectual-honesty-and-html5/
9 comments
The author isn't criticizing Apple for adopting standards too fast. I don't know why you got that impression. Instead, the author is arguing the following.
The term "HTML5" is supposed to be a standard, but the term has become a dumping ground for various parties. They introduce non-standard (or, at least, not-yet-standard) extensions in their browsers and call it HTML5. It's like Microsoft calling in-browser DirectX transformations HTML5.
I don't think a technology may actually be called HTML5 until it's standardized by WHATWG or by W3C, otherwise the term will become meaningless, and it looks like both Apple and Google are quickly on their way towards making that a reality. Once the term has come to mean "arbitrary cool browser-specific extensions" we are only one small step away from the age-old "embrace and extend" practice.
Coincidentally, Safari's recent Webkit 3D CSS transformations support (used everywhere in their demos) is very similar to DirectX transformations. I'm not saying that there's anything wrong with that, but as far as I know Webkit 3D transformations aren't (yet?) standardized by WHATWG or W3C.
The term "HTML5" is supposed to be a standard, but the term has become a dumping ground for various parties. They introduce non-standard (or, at least, not-yet-standard) extensions in their browsers and call it HTML5. It's like Microsoft calling in-browser DirectX transformations HTML5.
I don't think a technology may actually be called HTML5 until it's standardized by WHATWG or by W3C, otherwise the term will become meaningless, and it looks like both Apple and Google are quickly on their way towards making that a reality. Once the term has come to mean "arbitrary cool browser-specific extensions" we are only one small step away from the age-old "embrace and extend" practice.
Coincidentally, Safari's recent Webkit 3D CSS transformations support (used everywhere in their demos) is very similar to DirectX transformations. I'm not saying that there's anything wrong with that, but as far as I know Webkit 3D transformations aren't (yet?) standardized by WHATWG or W3C.
No, his negativity wasn't nearly that focused. See the argument around his introduction of "fuck you". Aside from which, I'm partly making a meta comment on the discussion over this page. The last couple of posts concerning it have been catch-alls for poorly-argued, poorly-reasoned disdain. The html5 issue alone does not even remotely justify the negative response this page has gotten. That's a rationalization. Else, why have the dozens of articles posted here previously discussing htm5, with the same loose definition, not garnered the same emotional response?
The thesis I argued against is simply the best triangulation of the multiple, weakly expressed arguments made to justify the negative reactions.
And the claim that these "extensions" are the same as MS's from years past is, to put it mildly, nonsense. This has been pointed out elsewhere, so I won't rehash it. The aggressive refusal by some folks to not recognize this looks like bias to me.
The thesis I argued against is simply the best triangulation of the multiple, weakly expressed arguments made to justify the negative reactions.
And the claim that these "extensions" are the same as MS's from years past is, to put it mildly, nonsense. This has been pointed out elsewhere, so I won't rehash it. The aggressive refusal by some folks to not recognize this looks like bias to me.
Yes, but these 'browser extensions' aren't extensions in the same sense that IE's were. -mox, -ms, and -webkit are vendor prefixes, and they're part of the process for a vendor to get something put into the standard.
See this great comment from yesterday: http://news.ycombinator.com/item?id=1403749
See this great comment from yesterday: http://news.ycombinator.com/item?id=1403749
That is correct. But until they're actually put into the standard they shouldn't be labeled HTML5, unless you want the term HTML5 to become meaningless.
They're draft standards, just like the rest of HTML5. It's not a standard yet either.
If their page is entitled "web standards", it's probably a bad idea to restrict the page only to users of Safari. Especially when it shares a rendering engine with Chrome and a number of other browsers that could probably run the demos.
For the record, when Google created Chrome Experiments (which used HTML5-ish technologies), they let any browser attempt to run the experiments, but the performance (in some cases) suffered. Apple is doing exactly the opposite.
For the record, when Google created Chrome Experiments (which used HTML5-ish technologies), they let any browser attempt to run the experiments, but the performance (in some cases) suffered. Apple is doing exactly the opposite.
If you go to the developer page (here http://developer.apple.com/safaridemos/ ) you can attempt to run the demos in any browser you fancy.
On the consumer accessible page Apple are requiring Safari because they can guarantee that the demos will work and they can guarantee they will have reasonable performance. And that's what's important. That's an advocacy page, a PR thing, not "look at this cool code we wrote". Presentation is everything.
In other words, different goals - different implementations.
On the consumer accessible page Apple are requiring Safari because they can guarantee that the demos will work and they can guarantee they will have reasonable performance. And that's what's important. That's an advocacy page, a PR thing, not "look at this cool code we wrote". Presentation is everything.
In other words, different goals - different implementations.
and that's bullcrap
the really important part of html5 at the end of the day is for CONSUMER, not developers.
Here Apple doing exactly what dumb people did when they started the browsers war 10 years ago "this site is best viewed with X", which later degenerated to "if you don't have browser X, go f*ck off" (you know wars ... people pick a side).
A browser is not here to guarantee an experience, it's here to guarantee you to browse content, and wether this browser support whatever version of HTML, this browser should not be blocked in any arbitrary way because the HTML should gracefully degrade, yeah even with Lynx on a command line.
http://www.anybrowser.org/campaign/abfaq.html
that's what people don't get about HTML5 and the article make a good illustration of it.
the really important part of html5 at the end of the day is for CONSUMER, not developers.
Here Apple doing exactly what dumb people did when they started the browsers war 10 years ago "this site is best viewed with X", which later degenerated to "if you don't have browser X, go f*ck off" (you know wars ... people pick a side).
A browser is not here to guarantee an experience, it's here to guarantee you to browse content, and wether this browser support whatever version of HTML, this browser should not be blocked in any arbitrary way because the HTML should gracefully degrade, yeah even with Lynx on a command line.
http://www.anybrowser.org/campaign/abfaq.html
that's what people don't get about HTML5 and the article make a good illustration of it.
Odd, when I go there in Firefox I can get to the pages but if I click View Demo on any of them (well, I tried the first 3) it still yells at me and tells me that I need Safari.
Chrome Dev 6, all of the demo, except number 5, works.
So they'll let you try to run them if you have a webkit based browser but that's it? Can anyone confirm that any non-webkit based browsers work?
Just tried three gecko-based browsers; none work.
> On the consumer accessible page Apple are requiring Safari because they can guarantee that the demos will work and they can guarantee they will have reasonable performance.
I'm sorry, how is that supporting standard again? I'm not illuminated yet.
I'm sorry, how is that supporting standard again? I'm not illuminated yet.
> could probably run the demos.
I tried the demos in Firefox and Chrome. Not all of the demos worked in either of them, so much for "could probably". It's not about Safari trying to make itself look better, it's about Apple not letting html5 look bad. If someone not-so-knowledgeable surfed to the demos with browser X and not everything worked then they may proclaim that html5 is crappy and doesn't work. Apple is trying to show-off html5 and that would be the opposite of their goal.
I tried the demos in Firefox and Chrome. Not all of the demos worked in either of them, so much for "could probably". It's not about Safari trying to make itself look better, it's about Apple not letting html5 look bad. If someone not-so-knowledgeable surfed to the demos with browser X and not everything worked then they may proclaim that html5 is crappy and doesn't work. Apple is trying to show-off html5 and that would be the opposite of their goal.
They restrict it to Safari because the dire truth is that so far, Safari is that one browser that does the STANDARDIZED HTML5 portions best. Chrome hits the second place not far behind, not surprisingly since it also uses the WebKit engine. Firefox however, by many peoples' experience - and, again, using HTML5 demos that do not use vendor-specific CSS magic, APIs and whatnot - simply fails at running most of these demos.
add.: to the downvoters... really, just load the page up in Chrome, check the demos, then load it up in Opera (with User Agent spoofing) and check the demos there, THEN try loading it up in Firefox, and you'll see for yourself whether it's a case of these HTML5 demos being "proprietary", "Safari only" or "broken", or if it's a case of Firefox just not cutting it yet. (Spoiler: it's a case of Firefox just not cutting it yet; almost all of the demos work identically in Chrome and Opera.)
add.: to the downvoters... really, just load the page up in Chrome, check the demos, then load it up in Opera (with User Agent spoofing) and check the demos there, THEN try loading it up in Firefox, and you'll see for yourself whether it's a case of these HTML5 demos being "proprietary", "Safari only" or "broken", or if it's a case of Firefox just not cutting it yet. (Spoiler: it's a case of Firefox just not cutting it yet; almost all of the demos work identically in Chrome and Opera.)
And the reason it doesn't work in Chrome's WebKit is because Google is implementing CSS3 transforms differently, and they're not done yet.
For me (thinking as Apple) it's this simple:
- Let's create some demos of our new HTML5 support!
- Do we want to be responsible for users' bad experience if other browser vendors make mistakes or implement things differently than we did?
- No.
That's it. Nothing nefarious about it.
- Let's create some demos of our new HTML5 support!
- Do we want to be responsible for users' bad experience if other browser vendors make mistakes or implement things differently than we did?
- No.
That's it. Nothing nefarious about it.
> For me (thinking as Apple) it's this simple:
Did you read the article? Those demos are not even part of html 5 yet (the ones that don't work on chrome and firefox at least). So calling them HTML 5 was not honest and a shameless promotion of Safari. Also, most of the demo do work great on both Chrome and Firefox, except the demos that are not HTML 5.
Did you read the article? Those demos are not even part of html 5 yet (the ones that don't work on chrome and firefox at least). So calling them HTML 5 was not honest and a shameless promotion of Safari. Also, most of the demo do work great on both Chrome and Firefox, except the demos that are not HTML 5.
Refreshingly well written ("relaxed" is the word that came to mind).
He highlights a paragraph which apparently is the important bit (and it is) but this, for me, is the really insightful statement (in the last section):
HTML5 is in a dangerous place since everyone wants to own it, but everyone is in a different place in terms of support or even what it means.
So true. This is the risk with HTML5 - it is becoming a buzzword for companies such as Apple to jump on and market with. Which isn't a bad thing; unless it is at the detriment of the standard.
He highlights a paragraph which apparently is the important bit (and it is) but this, for me, is the really insightful statement (in the last section):
HTML5 is in a dangerous place since everyone wants to own it, but everyone is in a different place in terms of support or even what it means.
So true. This is the risk with HTML5 - it is becoming a buzzword for companies such as Apple to jump on and market with. Which isn't a bad thing; unless it is at the detriment of the standard.
I agree, a refreshing analysis.
You're right in pointing out that line, but I think his highlighted paragraph (esp. the first few lines) provide some context for understanding/simplifying the conclusion.
The most important aspect of HTML5 isn’t the new stuff like video and canvas (which Safari and Firefox have both been shipping for years) it’s actually the honest-to-god promise of interoperability.
The most important aspect of HTML5 isn’t the new stuff like video and canvas (which Safari and Firefox have both been shipping for years) it’s actually the honest-to-god promise of interoperability.
Apple wants to show that you can do kick ass stuff on iPads, iPhones and Macs without any of that Flash crap (sort of paraphrasing Apple here, not necessarily my opinion).
That’s probably their intention, not showcasing interoperability or openness. Those two words do their little buzzword duty and that’s that. (I’m really not all that irate about that. It’s a cheap shot at Flash but many – probably all – of the things Apple uses in their demo actually will be a interoperable and open standard in the near future and all those evil browser prefixes will be dropped.)
That’s probably their intention, not showcasing interoperability or openness. Those two words do their little buzzword duty and that’s that. (I’m really not all that irate about that. It’s a cheap shot at Flash but many – probably all – of the things Apple uses in their demo actually will be a interoperable and open standard in the near future and all those evil browser prefixes will be dropped.)
That would be a good theory if their page wasn't titled "HTML5 and web standards." There's not much room for interpretation in that.
I suspect that they simply let their desire to make the demos really impressive distract them from the original point of the demos. They could, I imagine them thinking, make some cool demos that would work in Chrome and Firefox and Safari, or they could make some even cooler demos that showed off things only available in Safari, and have a nice way to promote Safari to boot.
The problem is, once they did that, it went from a showcase of HTML5 standards to a showcase of proprietary Safari features that might be standards eventually. Oops. I'm rarely critical of Apple, but this is one of those times that makes me wonder if anyone there had their brains turned on.
I suspect that they simply let their desire to make the demos really impressive distract them from the original point of the demos. They could, I imagine them thinking, make some cool demos that would work in Chrome and Firefox and Safari, or they could make some even cooler demos that showed off things only available in Safari, and have a nice way to promote Safari to boot.
The problem is, once they did that, it went from a showcase of HTML5 standards to a showcase of proprietary Safari features that might be standards eventually. Oops. I'm rarely critical of Apple, but this is one of those times that makes me wonder if anyone there had their brains turned on.
HTML5 and nice-sounding-buzzword.
That’s something you can do. I don’t think it’s nice of them to throw around buzzwords while being so completely irony resistant, it’s just that I don’t think it’s a reason to be greatly angered.
That’s something you can do. I don’t think it’s nice of them to throw around buzzwords while being so completely irony resistant, it’s just that I don’t think it’s a reason to be greatly angered.
I'm not greatly angered and I doubt anyone else is. It just strikes me as stupid, that's all.
My interpretation is Apple wants to influence what the final standard is going to look like. This is their vision for it which is largely based off the current W3C working draft. What's wrong with that? That's how the process is supposed to work. Apple has a few features not presently in the working draft that they want to see in the final standard. Here's the demo and you can get the WebKit source code to see how it works. The W3C, which Apple is a member of, may ratify it as part of the standard or they may not. I don't think there's any indication it will effect WebKit's ability to follow the final standard.
It's even worse than he says: try using Safari to visit the VR demo at http://www.apple.com/html5/showcase/vr/ and you might be greeted with this:
This demo requires a browser that supports CSS 3D
transforms.
To view this demo, you’ll need Safari on Mac OS X Snow
Leopard, Safari on iPhone OS, or the latest WebKit
Nightly Build.
This post is lovely.
I don't think it's realistic to get people to stop saying "HTML5" and start referring to all the little semi-standard pieces individually, so I'm really curious what this "intellectually honest" Mozilla messaging is going to look like.
I don't think it's realistic to get people to stop saying "HTML5" and start referring to all the little semi-standard pieces individually, so I'm really curious what this "intellectually honest" Mozilla messaging is going to look like.
And history repeats itself. "HTML5" is the new "AJAX".
And it couldn't come soon enough.
I see words like these as very useful inflection points; on one side of the curve, a developer that uses them without revealing a more nuanced understanding is a developer I should avoid.
And on the other side of the curve, it's a way for non-technical people to refer to a complex set of technologies they otherwise probably couldn't talk about (or spend money on).
I see words like these as very useful inflection points; on one side of the curve, a developer that uses them without revealing a more nuanced understanding is a developer I should avoid.
And on the other side of the curve, it's a way for non-technical people to refer to a complex set of technologies they otherwise probably couldn't talk about (or spend money on).
If "HTML5" is a term used for different things for different people, what's the term used for "shiny CSS toys" like the Apple HTML5 page? CSS3? Because that's actually what I'm interested in. All the canvas and video support is fine and dandy, but the shiny stuff like built-in rounded corners, border images, gradients, transitions, and animations are killer features.
I have a feeling that we're going to have to define a whole bunch of terminology to move ahead. People are confusing the HTML5 "markup additions" to the HTML5 "javascript upgrade" to Apple's HTML5 "hey look Ma, no Flash and lots of shiny!" (And to be fair to Apple, even Chrome doesn't fully support Apple's CSS3 shiny 3D toys, so that's why they had a Safari Only "HTML5" demo.)
I have a feeling that we're going to have to define a whole bunch of terminology to move ahead. People are confusing the HTML5 "markup additions" to the HTML5 "javascript upgrade" to Apple's HTML5 "hey look Ma, no Flash and lots of shiny!" (And to be fair to Apple, even Chrome doesn't fully support Apple's CSS3 shiny 3D toys, so that's why they had a Safari Only "HTML5" demo.)
History is clearly repeating itself with the proprietary "HTML5 extensions".
EDIT: To the downvoters, I'm sorry that you don't like the truth.
EDIT: To the downvoters, I'm sorry that you don't like the truth.
To be fair, all or almost all of the features Apple (and others) are showcasing do have open specifications at various stages of standardization.
The bad thing about branding them collectively as some sort of "HTML5 standard" is when people start calling other browsers non-compliant with standards that aren't even stable targets yet. Or targeting one browser's draft implementation, without even checking to see which other browsers have comparable support.
This is a major problem for us on the Mobile Firefox team. Since a huge portion of mobile web pages are now targeting WebKit browsers only, it's hard for competing browsers to enter this space. Mobile Safari is the new IE (in the sense that we need to reverse-engineer its non-standard features like meta[name="viewport"] in order to compete with it).
The bad thing about branding them collectively as some sort of "HTML5 standard" is when people start calling other browsers non-compliant with standards that aren't even stable targets yet. Or targeting one browser's draft implementation, without even checking to see which other browsers have comparable support.
This is a major problem for us on the Mobile Firefox team. Since a huge portion of mobile web pages are now targeting WebKit browsers only, it's hard for competing browsers to enter this space. Mobile Safari is the new IE (in the sense that we need to reverse-engineer its non-standard features like meta[name="viewport"] in order to compete with it).
Well the web needed a means of setting the viewport scale, and the one Apple came up with is as good as any.
Are you really insistent that you come op with your own "-moz-viewport" crap just because there isn't a draft standard? The vendor prefix bullshit of the last few years is going to be around for at least another decade, with every CSS3 property being set three goddamn times because the vendors will have shipped support for the prefixed version for several major releases before they think about blessing it. I wish y'all would only do that stuff in beta releases to keep it out of the wild.
Setting the viewport doesn't require anything like trampolining through VBScript, loading an ActiveX control, or calling a DirectX function via VML (all things I've had to do recently). MobileSafari is not the new IE.
Have fun being another legacy browser on a slow release cycle that can't ship a decent mobile browser.
Are you really insistent that you come op with your own "-moz-viewport" crap just because there isn't a draft standard? The vendor prefix bullshit of the last few years is going to be around for at least another decade, with every CSS3 property being set three goddamn times because the vendors will have shipped support for the prefixed version for several major releases before they think about blessing it. I wish y'all would only do that stuff in beta releases to keep it out of the wild.
Setting the viewport doesn't require anything like trampolining through VBScript, loading an ActiveX control, or calling a DirectX function via VML (all things I've had to do recently). MobileSafari is not the new IE.
Have fun being another legacy browser on a slow release cycle that can't ship a decent mobile browser.
> Are you really insistent that you come op with your own "-moz-viewport" crap just because there isn't a draft standard?
No, we're not insistent. We implemented WebKit's "viewport" meta tag. We can be pragmatic too.
I just wanted to point out one of the many of Safari features that Apple is promoting without writing any specs or standards, and the effect this has on the browser market. Others include touch events, link[rel="apple-X"], -webkit-text-size-adjust...
I agree the proliferation of vendor prefixes in the wild has bad effects, and we are feeling the pain of it too.
----
> Have fun being another legacy browser on a slow release cycle that can't ship a decent mobile browser.
Thanks for the kind words. :) We know we're way behind the game on mobile, and the only thing we can do to fix it is ship great software. I think when you get a chance to see what you can do with add-ons for mobile Firefox, you'll be tempted to switch.
We're on a rough six-month release schedule: Fennec 1.0 shipped in January, Fennec 1.1 will ship this month, and we are targeting this October or November for Fennec 2.0. Look for our first 2.0 alpha release for Maemo and Android in a month or two.
[P.S. I see you're in Seattle too! Want to get together for coffee sometime? Email me if you do.]
No, we're not insistent. We implemented WebKit's "viewport" meta tag. We can be pragmatic too.
I just wanted to point out one of the many of Safari features that Apple is promoting without writing any specs or standards, and the effect this has on the browser market. Others include touch events, link[rel="apple-X"], -webkit-text-size-adjust...
I agree the proliferation of vendor prefixes in the wild has bad effects, and we are feeling the pain of it too.
----
> Have fun being another legacy browser on a slow release cycle that can't ship a decent mobile browser.
Thanks for the kind words. :) We know we're way behind the game on mobile, and the only thing we can do to fix it is ship great software. I think when you get a chance to see what you can do with add-ons for mobile Firefox, you'll be tempted to switch.
We're on a rough six-month release schedule: Fennec 1.0 shipped in January, Fennec 1.1 will ship this month, and we are targeting this October or November for Fennec 2.0. Look for our first 2.0 alpha release for Maemo and Android in a month or two.
[P.S. I see you're in Seattle too! Want to get together for coffee sometime? Email me if you do.]
> MobileSafari is not the new IE.
You said exactly why it is the new IE:
> The vendor prefix bullshit of the last few years is going to be around for at least another decade, with every CSS3 property being set three goddamn times because the vendors will have shipped support for the prefixed version for several major releases before they think about blessing it. I wish y'all would only do that stuff in beta releases to keep it out of the wild.
You said exactly why it is the new IE:
> The vendor prefix bullshit of the last few years is going to be around for at least another decade, with every CSS3 property being set three goddamn times because the vendors will have shipped support for the prefixed version for several major releases before they think about blessing it. I wish y'all would only do that stuff in beta releases to keep it out of the wild.
It seems like Mozilla might benefit from developing and promoting tools like Modernizr, to encourage the standard use of feature detection instead of filtering by user agent. Some kind of complimentary repository for web developers to contribute and discuss equivalent rich web app functionality implemented for various browser capabilities, with sample code ensuring compatibility with Firefox, IE and Webkit browsers using Flash fall-backs when needed, would be the bee's knees. The benefit to no one owning the HTML5 standard means that Mozilla can also define it however they want. Given their relative browser market shares, Mozilla's take on the matter carries more weight than Apple's. It's sad to see Mozilla acting threatened by Apple on this, rather than throwing its own weight around, presenting their own case to web developers, and taking advantage of this transition to shiny HTML5 goodness to replace the crusty old user agent string method of compatibility checking with the shiny new hotness of discreet feature detection.
The bad thing about branding them collectively as some sort of "HTML5 standard" is when people start calling other browsers non-compliant with standards that aren't even stable targets yet. Or targeting one browser's draft implementation, without even checking to see which other browsers have comparable support.
This is like "Draft-N" Wireless too.
This is like "Draft-N" Wireless too.
At least there was only one draft of N.
Is this a joke? There were 13 approved drafts of 802.11n. http://en.wikipedia.org/wiki/802.11n#Timeline
Rather, all the draft-N devices I'm aware of interoperated. If you bought a draft-N network card, it didn't stop working the day after they published a new version, as far as I know.
Apple is at the forefront of standards adoptions, Safari's got some stuff that other browsers don't, and they want to show it off! What on earth is wrong with this? Firefox has better standards support? So what? Are we claiming Apple has a moral imperative to create demos of Firefox-only standards support?! Or, is it the even crazier claim that only the "best" product is allowed to promote itself? If you're selling Porsches and you say they're fast, is this offensive because Ferraris are faster? Good God.
I've also never before seen anyone so self-righteous about the use of the term html5; there's nothing about Apple's page which deserves special enmity for its "misuse" of this ambiguous term. In fact, I'm pretty sure in the future we'll see plenty of articles talking about html5, just as we have in the past, e.g., scribd converting from flash.
If you can't see how a company might find value in illustrating how quickly it implements emerging standards, I suggest you aren't thinking about it very hard. This page isn't about why web standards are important for cross-platform compatibility. Criticizing it for not demonstrating something which is not is purpose is awfully silly.
The outrage stirred up by the last couple of posts on this subject strikes me as follow-on to other perceived "evils" by Apple, and seems to be heavily biased rather than treating this issue on its own ethically distinct terms. But then, that's what people do.