On 23/07/15 15:47, Anne van Kesteren wrote: > > * Construct an nsIURI. Might need to take things such as the > document base > > URI and charset into account. > > For many APIs these days you need to parse the URL before you hand it > off. That seems like a somewhat saner setup too, though in theory > fetch() has access to this information through request's client.
I don't know if this is implied when you write "parse the URL", but I'd like to point out that constructing a nsIURI is actually more complicated than parsing a URI. We also have security checks. I haven't checked the specifics for web-facing URIs, but for file:// URIs, these involve actually accessing files. -- David Rajchenbach-Teller, PhD Performance Team, Mozilla
signature.asc
Description: OpenPGP digital signature
_______________________________________________ dev-platform mailing list [email protected] https://lists.mozilla.org/listinfo/dev-platform
