Posts Tagged ‘html5’

Serving html5 videos with Apache

Wednesday, February 24th, 2010

Just a quick tip today – the solution to a bit of a ‘gotcha’!

While writing the jCaps plugin, we came up against an issue where .ogv videos were refusing to load, but bizarrely, only in Firefox. In its place was a grey box where the video should be, with a question mark in the middle.

After a little bit of head-scratching, we found the solution – Apache wasn’t sending the right MIME type with the video, resulting in a Firefox fail. The solution is to put the following AddType directive in your .htaccess or httpd.conf file (mod_mime is required):

AddType video/ogg .ogv

Make sure that you restart Apache if you put it in httpd.conf.

On another note, jCaps is nearing another release – hopefully this week. It will take into account some of Bruce Lawson’s wishlist, and includes a big refactoring of many elements of the plugin, including the API. It should be leaner, meaner, and easier to use! In the meantime, you can fork or download the project at Github.

Tags: , ,
(1) Comment

Atwood’s Law and browser features

Friday, February 19th, 2010

“Any application that can be written in JavaScript, will eventually be written in JavaScript.” – Jeff Atwood

Wednesday’s blog post introducing the jCaps plugin (which aims to provide accessible captions for HTML5 videos) sparked an interesting discussion with John Foliot on Twitter the other night.

John brought to my attention the WAI’s Media TextAssociations specification, which will tell browsers how to associate alternative content with rich media, such as video. Little did I know that while I was developing a “home-made JavaScript solution” (which the spec is specifically designed to counteract), the task force was carrying out a meeting to move the spec forward to the testing and implementation stage.

If there’s a W3C endorsed spec which will handle the issue, then great! My concern, though, is the length of time it will take for the specification to be implemented by browser manufacturers. Even if we’re only months away from a spec, it could be much longer until it is reliably implemented.

So in a similarly self-aggrandising way, I’d like to propose McCollum’s law: “If a browser feature which should exist, doesn’t, someone will write a javascript implementation.”

Witness the excellent jQuery UI library. HTML5 provides a datepicker element, a slider element, and native drag and drop functionality, but until these elements are natively supported across the majority of browsers, we’re stuck with home-made JavaScript.

Finding the right compromise between pragmatism and idealism is essential for modern web development.

What are your thoughts? How do you find a balance? Should jCaps development proceed, or should we wait until standards catch up?

Tags: , , , ,
No Comments