> I adore SQLite but I think the W3C made the right call there - web standards need more than one compatible implementation.

could you elaborate on that? I get the impression that now, eight years later, we're much worse off. I don't really see the upside of not going for what is, AFAIK, a relatively vanilla implementation of SQL.

A web standard needs to have more than one implementation. If you allow one single implementation to define the standard, you get something like Flash.

I don't have a particularly strong practical argument here: for me this is more of a fundamental principle for how the web should work.

Flash was a closed source plugin that was added to the browser as a binary blob. It could be updated by the user, and because it frequently had security bugs users were frequently urged to update. Those updates also often introduced new features, which content providers started taking advantage of once a significant number of people updated, putting pressure on the rest to update too.

SQLite is open source that would have been compiled into the browser by the browser vendor. A typical user would not have the means to update it separately from the rest of the browser.

So why couldn't W3C pick a particular release of SQLite, list a specific subset of its features, and declare that this, as described in the documentation for that specific version, is the standard?

> web standards need more than one compatible implementation

Seems like a similar situation to WebGL. I think they would have looked past implementation count if the feature was really needed.