Re: [hybi] WebSockets : Question about masqued frames !

Hi,

It's mainly because of proxy traversal (please read previous discussions 
about cache poisoning & co)

Anthony Catel

Le 09/06/2011 17:13, Jean-Christophe Bos a écrit :
>
> Hello everyone,
>
> I wanted to ask you (from France!) about the latest WebSocket's drafts 
> that I also fully implemented in my own HTTP server (since hixie 76).
>
> On the masked frames, you mentioned earlier, that the mask must be 
> selected by an high entropy but it seemed illusive.
>
> However, you came back to say that just choose a random mask but 
> always in a logical of non-predictability.
>
> So, what is the real utility of this mask and that it should not be 
> predictable except to prevent a simple human readable dump of the 
> connection?
>
> Why have won so much space on the data payload length and lose 
> unnecessarily 32b for the masks contained in each frame?
>
> WebSockets is indeed a protocol over HTTP over TCP, thus ensuring a 
> good packets order and lossless.
>
> So why not simply imagine a mask whose evolutionary of the Salt was 
> fixed at the start (why not from the handshake key) and whose 
> encryption evolve based on the contents of the frames?
>
> I would be really pleased that you explain on that!
>
> Thank you very much because I truly believe that WebSockets are a 
> priority in the future of dynamic web apps :-)
>
> Sincerely,
>
> **
>
> *-*
>
> *Jean-Christophe Bos*,
>
> CEO, Tenactys Group
>
> 0 820 620 118
> 06 80 27 93 84
>
> jcb@tenactys-group.fr <mailto:jcb@tenactys-group.fr>
>
>
> _______________________________________________
> hybi mailing list
> hybi@ietf.org
> https://www.ietf.org/mailman/listinfo/hybi