expert review

[paulirish]

also we're interested in reviews from people like this...

• @ slightlyoff
• @ necolas
• @ tabatkins
• @ chriscoyier
• @ rem
• @ desandro
• @ fyrd
• @ malarkey
• @ jdalton
• @ simurai

also either heillman or nyman..

(intentionally not @mention'd in order to not ping them too early)

[addyosmani]

Not on the list and far from an expert, but here are my thoughts:

Audio - imo, this shouldn't be considered an 'avoid' if we're saying it's safe to use HTML5 video. They both have the same caveats and I've seen quite a few sites use SoundManager or jPlayer in production quite happily. If we're talking about WebAudio/the audio manipulation APIs, that's a completely different story but playback is something I personally consider solved. Consider switching to 'caution' if you feel unhappy with that? :)

DataList - I was under the impression Mike's polyfill did the job quite well. Are we assuming that polyfills with a jQuery requirement have points counted against them for the dependency? If so, perhaps this highlights a need for someone to work on library-agnostic versions of these polyfills. It's not that hard to do :)

P:S:

If we're saying people should be using polyfill X or Y, should there be some indication to preference should there be a few different polyfills available for a feature? Gradings? Not sure quite how granular you would like to get, but it's just an idea.

//cc @nimbupani

[nimbupani]

Thank you! Could you pull request on those markdown files with your updates? :))

Yes, the intention is to recommend one or two not more. Right now, the idea is to recommend what most people feel has worked best for most of their usecases w.r.t that feature. There are no gradings :/

[addyosmani]

Sure thing @nimbupani!

[addyosmani]

@nimbupani here you go: https://github.com/h5bp/html5please/pull/14 :)

I'm currently thinking about how additional guidance on polyfills mentioned might be provided. A simple solution might be just expanding the list across a few lines. e.g

Polyfill A: Works cross-browser. Limitations: requires jQuery Polyfill B: Works in all browsers from IE8 onwards. Limitations: Slower than Polyfill A Polyfill C: Smaller than other solutions. Limitations: limited feature-set.

Again..not sure whether this is something you guys would like to cover. I tend to overthink things :)

[nimbupani]

Going to reach out to a few to get feedback.

[desandro]

Yay, this is an awesome project. A couple times I've scoured caniuse and wondered if something is ready for "prime-time." I feel that this is the mission of HTML5 please, rather than say "can I use X feature on Y devices."

caniuse.com is SOPA blacked out, so maybe today isn't the best day to review the site, as I feel like I only have half the information I'd want. The following are personal suggestions, nothing that MUST be changed, so take 'em or leave 'em.

• I feel that CSS3Pie causes more problems than it solves. I wouldn't recommend it. It's better to take the TAFEE/progressive-enhanced approach.
• References help anyone looking for more info. CSS Animations are still in a flux that's fascinating, where can I learn more?
• Does CSS3-Colors include RGBA / HSL / HSLA or is just the fun keywords?
• All caps is nice text treatment, but for big lists like this, it would improve legibility to have no text-transform on h2.name.
• What do you think about putting elements in brackets? <meter>, <menu>, <canvas> ? I immediately communicates that these are HTML elements and that the others are not.
• That HTML / CSS / JS data needs to be visible. I was like "Um, what is Eventsource?". I see it's in there, but not visible right now.

[paulirish]

great feedback.

killing uppercase. will add those brackets. starting to add references too.and we'll kill css3pie reccs.

what'd you mean with that last bullet?

[desandro]

what'd you mean with that last bullet?

Is this list item a HTML, CSS, or JS feature? Some are clearly HTML elements, others are specifically CSS properties. From viewing the source, I see that data is already in the page in <h4 class="kind"></h4>. I imagine the designers are just still deciding on a visual style for it

[nimbupani]

yes I understand @desandro it is getting hidden in overflow hidden which I need to fix. :)

[Raynos]

I'd personally say that webgl should not be an "avoid".

There has been some really impressive stuff done with it.

The main downsides are no IE support but one can live with that.

The main reason for recommending it's use is that we need more people to work on major flagship web applications using webGL. webGL is currently stable enough to develop PlayStation2 style games for the web.

As a community we really want to encourage webGL so that the web can be considered a worthwhile platform for 3D applications.

Sidenote: There are some big companies working on AAA style games for WebGL. I can't remember the names or the details but the platform is being taken seriously by big game companies.

Sidenote 2: The platform has little maturity or tooling, but we still don't want to disencourage the community from using webGL.

[paulirish]

agreed on webgl. shouldn't be avoid. it's been very stable over the past 12mo for sure, as well.