Re: The interpretation of script

On Jan 20, 2011, at 12:50 , Michael Kay wrote:
> Personally, my response to this is the brutal one: conform to the rules in your own product, and make no concessions to people who don't. It's tough, but people will respect you for it, and in the end, the people who don't conform will be left out on a limb and will change.
> 
> But I guess that's diametrically opposed to the html/web culture, so I'm on a losing wicket.

Actually, it's not necessarily opposed to the HTML or Web culture. In the heady days of XHTML and XML-based SVG people actually did try to do that. Going strict was initially seen as a great step forward for mobile implementers who didn't want to reverse-engineer all the existing browsers' parsing idiosyncrasies; it also brought the advantage that it was far easier to implement on constrained devices than some of the semi-sentient algorithms required (or then thought to be required) by HTML.

It turns out that when content breaks, people don't respect you for that. A small, dedicated bunch of geeks might, but they're not the ones making you money. You can tell end users all you want that a well-formedness error in the external subset prevented DOM creation — it works with your competitor's implementation. The people who were buying your browser (something that happened then in that industry, and still does) might have some sympathy for the standards argument (many did) but at the end of the day they needed to ship something to their users. It wasn't long before mobile SVG implementations also used XML parsers that were a little bit more open-minded about well-formedness than one would have liked.

Doing this otherwise would require a web police with the executive power to enforce the rule of standards. We don't have that; we don't have the requisite power. What we have a market and a community that we can influence. I think that we can be more effective if we accept that premise and work with it, rather than try to go against its grain.

-- 
Robin Berjon - http://berjon.com/

Received on Thursday, 20 January 2011 12:12:22 UTC