[Standards] Re: Inband Images

Richard Dobson richard at dobson-i.net
Wed Mar 28 21:50:18 UTC 2007


> I agree with Richard here. We should just use one method and a
> bytestream with a hash is at the top of my list. Like I mentioned in my
> last post, there's no reason we can't just use a long lived IBB and a
> queue (or for the more adventurous a long lived p2p connection of some
> sort).

Yea although it would be better to use a method a far less verbose than 
the si-pub method, thats so big that the xml could possibly end up being 
bigger than some of the smaller images, something nice and simple just 
specifying the hash would be better, with optional attributes to specify 
such things as an alternate source JID to retrieve it from or things 
like the content type of the image.

Richard




More information about the Standards mailing list