- From: Lapierre, Frederic <Frederic.Lapierre@aucs-europe.com>
- Date: Tue, 23 Nov 1999 03:25:20 -0500 (EST)
- To: "'www-talk@w3.org'" <www-talk@w3.org>
All, I have a problem using Content-Encoding What I'm trying to do: I want to serve a document (HTML, DOC PDF...) (via HTTP/1.1) using a compression mechanism if the browser supports it. What I'm doing: I check in the request (Accept-Encoding) if the browser support compression . If it supports I send back a compressed version of the requested file. Content-Type: <type> ;(e.g.text/html, application/msword, application/pdf) Content-Length: <length> Content-Encoding: <encoding> ;(e.g. gzip) The result: When sending back compressed html, gif, jpeg files the mechanism works just fine (i.e. file uncompressed and loaded in the browser) When sending back compressed doc or pdf files the external application (msword or acrobat reader) is launched but fails to load the document or load the doument but display the compressed content (the behavior is browser specific; I don't get the same behavior with Netscape Navigator and MS Internet Explorer). Reading chapter 7.2.1 in RFC 2068: entity-body := Content-Encoding(Content-Type(data)) further 14.12 "...and thus what decoding mechanisms MUST be applied in order to obtain the media-type referenced by the Content-type header field. ..." I don't understand why the mechanism is not working for all media-types. Any idea is welcome. Kind regards, and thanks Frederic Lapierre AUCS Communication Services Post Address: Postbus 2003, 2130GE Hoofddorp, The Netherlands Visit Address: Spicalaan 1-59, 2132JG Hoofddorp,The Netherlands Tel: +31 23 569 7741 Fax: +31 23 569 7788 Mobile: +31 6 222 43355 Email: Frederic.Lapierre@aucs-europe.com http://www.aucs-europe.com
Received on Tuesday, 23 November 1999 04:53:10 UTC