2010 has been a fantastic year for HTML5* (* and we mean HTML5 in the broadest sense, including CSS3, the HTML5 satellite specs and all the other technologies that have been bundled under the HTML5 banner). All mobile OS’s are now shipping or about to ship HTML5 browsers with the exception of Windows Mobile. And to everyone’s great surprise, IE9 is shaping up to be a fairly decent HTML5 browser too. But there’s still a lot of work still to do. As the new year approaches, we’re taking a stab at a HTML5 wish list for 2011. Some of these are web standards wishes, and the rest are browser wishes, so we just mixed them all in together and put them in a rough order of priority. So without further ado, here’s what we’d like to see from the W3C, WhatWG and the browser makers this year:
10. Sustained effort to move more -webkit effects into CSS3
Although WebKit has become the modern mobile device standard, we’d still like to rely a little less on -webkit extensions to get our job done. Although many -webkit extensions have made their way onto standards track over the last year there are still notable exceptions that need to get there. WebKit masks and background clipping to text are just a couple of the CSS3 effects we’d really really like to see in W3C working drafts in 2011. They’re incredibly useful effects that we shouldn’t need SVG or Canvas to accomplish.
9. CSS3: A Richer Effects Toolbox
Blur effects (including motion blurs and filters) have been on the CSS wishlist since 1998. We can now do shadow blurs within WebKit, Mozilla and Opera, but we should have at least block-level blurs and filters within CSS. If we’re going to be able to fully replace Flash with browser technology, this is one to work on. In addition, more general support of CSS3 3D Animations (cough Android) would be excellent. Firefox and WebKit both have great gradient support but now that we have that basic support, we’re hankering for more. Gradients should be transitionable (according to standards track drafts), but today’s -webkit-gradients are implemented as images which can’t be transitioned. Gradients as border images are also an area for work since today they don’t stay within their border areas. (And we know we can duplicate any of these effects with Canvas, but the answer to every CSS feature request shouldn’t be “do it with Canvas”. Not to mention that Canvas still needs performance work.)
8. High Performance position:fixed for mobile
One of the major reasons to use Sencha Touch is that it gives you fixed position UI elements. But in our opinion, you shouldn’t need an application framework just to get this. It’s starting to show up in mobile browsers, but too slowly and too slow :-) This is a no-brainer for implementation in 2011.
7. GPU Acceleration
Even basic animation can over-stress your average mobile CPU, so it’s crucial to offload as much processing as possible to your GPU. Apple has been working on GPU offloading since at least 2006, so there’s a lot of work for the other folks to do to catch up. Look for even more work being done to offload general computation to the GPU in 2011. We’d like to see world-class GPU offloading from all the major browsers and mobile device makers this year.
6. Deeper Device Access
iOS 4.2 gave us accelerometer access from ordinary webpages on Apple devices, so next on the wish list is camera access. This is already in standards process with a solid specification, and we get this on Nokia but only as part of their web widget packaging. Once we have camera access, a whole class of applications that use cameras for everything from text recognition, to bar-code scanning to QR codes to face recognition will be able to go to the mobile web. We were promised this for Android at Google I/O – and we were hoping for it to arrive in 2.3, but now it looks like we’ll have to wait for 3.0 to get it. After camera, near-field communication access would enable some pretty neat applications – but even native apps have restricted access to these today.
5. Better debugging tools for mobile browsers
This could be on every year’s list. Desktop browsers are actually now pretty good for debugging, but mobile browsers remain a black box. We did some work this year to make Android debugging a little more manageable, but we’d like to see much better debugging support baked into the mobile browsers at a fundamental level and giving us better views into memory consumption and performance.
4. Web sockets stabilization
Are we there yet with Web Sockets? Is it workable or does it still need better security? Does it have to be rewritten from scratch or just polished around the edges? Web Sockets were there and then they weren’t and then they were back (sort of) in 2010. It’s not the end of the world if Web Sockets get delayed, we’ll just keep on using Comet or other methods, but they would certainly be a large gift to web developers if they arrived across all the major browsers. We’re crossing our fingers for this in 2011.
3. IE9 With Complete CSS3 Support
Internet Explorer 9 made a huge jump into HTML5 this year (particularly with Microsoft’s own Bob Muglia saying that Silverlight would essentially be deprecated in favor of HTML5 for desktop development). Our major quibble with IE9 is that it has resolutely avoided implementing the CSS3 effects that Apple has pioneered within WebKit, including animations and transitions. We’d really like to see these in the final IE9 – please don’t make us wait for IE10!
2. A HTML5 codec armistice
HTML5 video — awesome. But the devil is in the details. Right now developers are looking at generating three encodings: a H.264, an Ogg Theora and a WebM version to support all the codec camps. We know there are royalty, technology and strategic issues here, but seriously, isn’t it time for one high-quality, royalty free codec that everyone can ship and rely on. (And is there really that much strategic headroom left in codec technology? Perhaps so?)
1. A Reboot for WebSQL standardization
WebSQL standardization came to a grinding halt this year after Mozilla took issue with reverse engineering a spec for SQL in the browser from the SQLite implementation that Safari, Chrome and Opera is now shipping. IndexedDB now seems to be receiving the most attention as the next generation in-browser database, with both Microsoft and Mozilla backing IndexedDB as the storage mechanism of choice for large amounts of data. Mozilla says that IndexedDB is what web developers want. However, I doubt they talked to any Enterprise developers when making the decision to champion a non-relational data model as the ONLY standard for the browser. The fact is that Enterprise data-sets have lots of inter-relationships and Enterprise web apps want to work with relational data, not hierarchical b-trees. So, our last wish for HTML5 in 2011 is for a major database player (IBM? Oracle?) to collaborate with Google to develop a spec and a clean-room reimplementation of the SQL-92 subset that SQLite supports; and for WebKit to fork SQLite to provide a stable, standards-driven, open source code-base for relational data in the browser.
And all the rest…
So those are our major wishes for 2011. We had some other wishes (more sensor access, another run at standards-based privacy, noise for backgounds, Android browser improvements, text-on-a-path and more advanced text skewing, a complete implementation of the border-image spec, a concerted effort to kill the term “smartphone”…) but alas, the top 23 wishes isn’t quite as catchy as the “Top 10″. Did we miss anything major? We’d love to hear your feedback in the comments!
1402
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
