- From: Martin Thomson <martin.thomson@gmail.com>
- Date: Tue, 27 May 2014 13:40:13 -0700
- To: Michael Sweet <msweet@apple.com>
- Cc: Nicholas Hurley <hurley@todesschaf.org>, HTTP Working Group <ietf-http-wg@w3.org>
On 27 May 2014 11:52, Michael Sweet <msweet@apple.com> wrote: > And the issue is not just complexity but overhead - Huffman coding alone > requires relatively slow bit manipulations, the header tables add to the > memory overhead of every connection, and proxies get to do header processing > twice... Not a big deal on a desktop machine with a dozen connections, but > embedded devices and proxies have tighter constraints. Let's be very clear about the costs here. The only cost you cannot avoid is the cost of Huffman decoding. At a proxy, processing costs can be turned into memory costs relatively trivially and then amortized across multiple connections. A constrained device pays one way or other, but those costs are trivial in comparison to other costs that device incurs. The argument for a setting incurs complexity that is - at least in my opinion - a wash when compared to a simple Huffman decoder. My original point was focused on the concerns some raised regarding security.
Received on Tuesday, 27 May 2014 20:40:41 UTC