- From: Eric J. Bowman <eric@bisonsystems.net>
- Date: Tue, 29 Apr 2014 19:16:40 -0600
- To: Matthew Kerwin <matthew@kerwin.net.au>
- Cc: Roberto Peon <grmocg@gmail.com>, HTTP Working Group <ietf-http-wg@w3.org>
Matthew Kerwin wrote: > > I still think you're putting too much emphasis on C-E, particularly in > cases where it doesn't (or shouldn't) apply. What you want is > compressed data (fewer bytes) on the wire. HTTP has limited options > in this regard, and the Right Way™ (T-E) is out of the picture; but > using C-E is at best a hack, but more likely a flat out > misunderstanding of its purpose. And the current text madates, in an > internet standard, that everyone support the hack; and by extension > promotes the misunderstanding. > +1, the ends don't justify the means. The "Other Right Way" is still media types, specifically appending +gzip to HTML/CSS/JS identifiers; if gzip is explicit in Content-Type it doesn't need to be implicit in C-E. Achieves the performance win without all the headaches, I think. -Eric
Received on Wednesday, 30 April 2014 01:16:53 UTC