eating cucumbers without permission...

There was a big problem where sometimes pages would mysteriously fail to save. Moving from the “load” event to the “DOMContentLoaded” event fixes the problem, though as usual I’m not sure why. (My guess is that something else hooked to the insanely crowded load event was causing it, but I can’t say for sure.)

Moving to DOMContentLoaded causes a few problems, the worst of which is that Ark can no longer get the file size of the current page from the cache (unless it’s already there because the page has been recently visited).

The page size getter was already problematic, given that it was utterly useless for the application/x-gzip encoded pages which comprise 90% of pages I save, so it’s not too much of a blow. As a partial workaround, I added a dodgy file-size estimator. When I say “dodgy” I mean “dodgy”; all it does is count the number of characters in the document body. Thus, size estimation is utterly, utterly wrong for images, PDFs and the like.

It solves my second-to-last annoyance, though, so I can’t bring myself to care.

Download in the usual place.