- From: Gorm Haug Eriksen <gormer@opera.com>
- Date: Thu, 02 Aug 2007 12:10:54 +0200
- To: Joćo Eiras <joao.eiras@gmail.com>
- Cc: "Web API public" <public-webapi@w3.org>
On Thu, 26 Jul 2007 16:28:10 +0200, Joćo Eiras <joao.eiras@gmail.com> wrote: > After a quick glance I have two sugestions. > a) the tcpsocket object could have a peerAddress property which > returns the IP of the peer. This is great to resolve addresses. Could > there be any security issues ? Hi Joćo, The idea was to have an optional security model that applies to regular web *pages*. For web *applications* a more relaxed approach is often needed. I can't see that the peer property introduce any new security issues. > b) the ability to read/write data synchronously. It's certainly not clear enough, but this should work already. The socketdata event is just informal and contains no payload. It only contains information on the number of bytes available for reading. -- Cheers, - Gorm
Received on Thursday, 2 August 2007 10:10:38 UTC