Hacker News new | past | comments | ask | show | jobs | submit login
HTML5 Peeks, Pokes and Pointers (diveintohtml5.org)
60 points by pistoriusp on Aug 18, 2010 | hide | past | favorite | 20 comments



FYI: Article has nothing to do with memory pointers or linked list peeks, and everything to do with DOM attributes of HTML5 elements and attributes.


Yes, my mind was blown for a second and then I was slightly disappointed. ("Poke" and "peek" are also commands to write to and read from a given memory address, respectively.)

It's presumably a reference to old time tables of "peeks" and "pokes" showing what each memory address controlled, or somesuch.


Here's the story about the name: I wrote http://diveintomark.org/archives/2010/01/29/tinkerers-sunset earlier this year, decrying the lack of "tinkerability" of the iPad and similar closed devices. "There won't ever be a MacsBug for the iPad. There won't be a ResEdit, or a Copy ][+ sector editor, or an iPad Peeks & Pokes Chart."

One of the common refrains in the comments was that the iPad includes top-of-the-line support for a perfectly open, tinkerable platform: the web. Thus, the "HTML5 Peeks, Pokes and Pointers" chart was born.

Also, I just purchased a real live physical original Beagle Bros "Peeks, Pokes and Pointers" chart to hang up in my mancave^Woffice. Next to my Apple //es. Plural. So there's that.


Here is what it says in the credits (scroll down):

Inspiration: Beagle Bros “Peeks, Pokes and Pointers” chart for Apple ][ (+)

(+) links to this: http://www.lazilong.com/apple_II/bbros/


Fond memories of the Beagle Bros. poster that hung on the wall next to the Apple ][ 's in my high school's computer lab.


Is this a chart of individual tests for "HTML5" feature support, so that you could start using them today (providing graceful degradation)? Why, yes it is.

Does it say so anywhere on the chart? Let me check…


What I want to see is a drop-in JavaScript library that makes many of the new HTML5 elements, attributes, etc. backwards-compatible with all common browsers starting with IE6.

E.g., I want to say <input type="text" placeholder="What is your username?" /> and have a script that detects if the browser supports the placeholder attributes, and if not, simulates it transparently.

I'd happily understand/accept if this script was limited, so long as it had good clean documentation of what it supported (like placeholder).


Modernizr (http://www.modernizr.com/) is as close as you're going to get at the moment. It's pretty solid.


This is actually specifically NOT want I want:

"Modernizr does not add missing functionality to browsers; instead, it detects native availability of features and offers you a way to maintain a fine level of control over your site regardless of a browser’s capabilities."

I understand that not all HTML5 features can be reproduced in JS, but for the ones that can (such as "placeholder"), I don't want to have to think about reimplementing it for older browsers when it could be done automatically.


Everyone's so hyped up about HTML5. And yet I am slowly coming to the realization that the W3C has made a very, very, grave error in not going ahead with XHTML2.


The W3C did go ahead with XHTML2. And everyone else, most notably the browser makers, realized that with such a broken W3C as would create such a monster, they’d have to make a new spec themselves if they wanted it to be reasonable... whence WHATWG and HTML5.

Years later, XHTML2 abandoned by the world, the W3C decided it no longer wanted to be sidelined, and it begged for the HTML5 spec to be brought back under its purview. So that’s where we are now.


What do you find troubling? Even when XHTML 1.1(?) Was the hype around town I couldn't see the value in using it over traditional HTML 4.


Oh come on, even Zeldman's given it up and recognized that the 'XHTML' he'd been marketing was bullshit and that he'd never actually used XHTML.

I think Microsoft did a great mitzvah with the way they mercifully prevented it from ever actually being used — IE would offer to save it as a file if served as application/xhtml+xml, saving people from the fate of having compliant browsers refuse to render their documents over petty draconian bullshit. That period of about five years when everyone was super hyped about their 'XHTML' while actually using HTML was pretty annoying though.

Don't worry though, you can join the old XHTML fashionistas in trolling about accessibility instead.


You might enjoy http://diveintohtml5.org/past.html

Or you might not.


Why is that?


HTML5 basically takes HTML4 and adds 3 tags: audio, video, canvas. These solve very specific issues, sound, movies, and drawing. HTML5 does not address some larger scoped problems related to HTML -- the biggest being that HTML sucks. It has always sucked, and it needs to go. It needs to be replaced with flexible, programmable XML.


There is much more to HTML5 (or to HTML4 for that matter) than three elements (those are elements, not tags).

You can have your XML with XHTML or HTML5 XML serialization. Oh, but that does not solve the "flexible" part, right? Well, you have to choose there — either flexibility or interoparatibility. HTML5 focuses a lot on the latter.

What do you want? To have any arbitrary element in your markup? Ok, you get your <foobar>barbaz</foobar>. And then what? What should it mean to everyone else beside you? What do you gain by that?


"What do you want? To have any arbitrary element in your markup? Ok, you get your <foobar>barbaz</foobar>"

Yeah I want my own elements that do anything I want, and I'm working on a solution to it. And this project is what led me to the conclusion that beyond adding 3 tags HTML5 really doesn't address any of the bigger issues. I think it's really silly how much hype HTML5 is receiving when it's a patch job at best.


Well, the sooner you have your solution ready, the sooner you will understand that you were solving the wrong problem.

By the way some of the new elements in HTML5 (and there are not three but almost thirty new elements) are in fact supposed to save web developers from the need to invent theri own way to express some common semantics in HTML, e.g. section, artcile, nav, header, footer. There are alos a buch on new API, much richer inputs collection for forms, etc.


Isn't that what we used to call a web browser plugin?




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: