W3C home > Mailing lists > Public > ietf-http-wg-old@w3.org > May to August 1998

finishing testing of HTTP/1.1 (help requested)....

From: Jim Gettys <jg@pa.dec.com>
Date: Tue, 2 Jun 1998 14:34:22 -0700
Message-Id: <9806022134.AA23304@pachyderm.pa.dec.com>
To: http-wg@cuckoo.hpl.hp.com

I've added recent reports from Lester Waters of Microsoft,
Robert Polansky of Raptor (Axent), Mike Belshe of Netscape; my thanks
to them for filing them.  There have been over 20 reports filed to date.

I encourage others to submit reports.  There are still some significant
1.1 implementations out there for which I don't have data.

To make it easier to update reports, if people who have filed wish to 
drop me a note saying things like "I finished testing feature XXX", or 
"I found that my testing of feature XXX was broken or in error", I am 
willing to edit the existing reports to save people the effort of filling
out the very long form from scratch...

Anyone who can knock of features that don't have enough tested
implementations would be highly appreciated... 

I've automated generating the reports so that updating that page
is now easy for me to do.

See: http://www.w3.org/Protocols/HTTP/Forum/Reports/ for the
current state of reports.  It has a rollup of all the data,
and a list of features for which we are not 

Using the on-line data in the reprot should make it easier to figure out 
who has implementations to test against.

There are a few features where I am skeptical proper testing has
been done, which I plan to understand further to make myself confident we've
met IETF rules.

The list of untested implentations on the base spec continues to shrink. 
Some of the leftover features  below are actually incorrect, as there 
are often no special requirements on proxies; we don't really need proxy 
implementations for those features (though we wouldn't mind having them; 
in fact, one might claim that a proxy that just passes the response back 
from the origin server is correct), and that therefore we have 
implementations in hand in any case...  

Features that do not have requirements on proxies include 202 Accepted, 
203 Non-Authoritative Information, 205 Reset Content, 300 Multiple Choices, 
303 See Other, 305 Multiple Choices, 410 Gone, 411 Length Required, 412 
Precondition Failed, 413 Request Entity Too Large, 414 Request-URI Too 
Long, Content-MD5, TE, Trailer (if my quick look at the spec is correct).

Clients         Servers         Proxies         Feature
 1t  3y  7n  2-  3t  2y 12n  0-  1t  1y  6n  0- H 10.2.3        202 Accepted
 1t  2y  8n  2-  2t  2y 13n  0-  1t  1y  6n  0- H 10.2.4        203 Non-Authoritative Information
 1t  2y  8n  2-  2t  1y 14n  0-  1t  1y  6n  0- H 10.2.6        205 Reset Content
 2t  2y  8n  1-  2t  3y 12n  0-  0t  2y  6n  0- H 10.3.1        300 Multiple Choices
 3t  4y  6n  0-  2t  3y 12n  0-  0t  3y  5n  0- H 10.3.4        303 See Other
 2t  4y  7n  0-  1t  3y 12n  1-  1t  2y  5n  0- H 10.3.6        305 Use Proxy
 2t  3y  7n  1-  2t  2y 13n  0-  1t  1y  6n  0- H 10.4.11       410 Gone
 3t  5y  5n  0-  2t  6y  9n  0-  1t  3y  4n  0- H 10.4.12       411 Length Required
 3t  4y  5n  1-  3t  7y  7n  0-  1t  3y  4n  0- H 10.4.13       412 Precondition Failed
 2t  4y  6n  1-  3t  3y 11n  0-  1t  2y  5n  0- H 10.4.14       413 Request Entity Too Large
 2t  3y  7n  1-  2t  2y 13n  0-  1t  1y  6n  0- H 10.4.15       414 Request-URI Too Long
 1t  4y  7n  1-  2t  5y 10n  0-  1t  2y  5n  0- H 10.4.18       417 Expectation Failed
 1t  5y  6n  1-  3t  5y  9n  0-  1t  3y  4n  0- H 13.3.3        Weak entity tags
 2t  2y  9n  0-  1t  7y  9n  0-  1t  2y  5n  0- H 14.15 Content-MD5
 3t  4y  4n  2-  5t  7y  5n  0-  1t  4y  3n  0- H 14.24 If-Match
 3t  3y  6n  1-  4t  7y  6n  0-  1t  3y  4n  0- H 14.26 If-None-Match
 3t  1y  8n  1-  2t  6y  9n  0-  1t  1y  6n  0- H 14.27 If-Range
 3t  3y  7n  0-  1t  2y 14n  0-  1t  1y  6n  0- H 14.39 TE
 3t  2y  8n  0-  0t  2y 15n  0-  0t  2y  6n  0- H 14.40 Trailer
 3t  4y  4n  2-  1t  4y 11n  1-  1t  4y  3n  0- H 14.46 Warning

Authentication is in poorer shape, where we need another pretty complete
implementation to meet requirements. Our thanks to Ronald Tschalär,
who has done a complete client implementation (though hasn't had a proxy
to test the proxy authentication against).

Clients         Servers         Proxies         Feature
 2t  0y 11n  0-  3t  5y  9n  0-  1t  0y  7n  0- A 3.2.1 WWW-Authenticate Digest
 1t  0y 12n  0-  0t  1y 16n  0-  0t  0y  8n  0- A 3.2.1 qop-options auth
 1t  0y 12n  0-  0t  0y 17n  0-  0t  0y  8n  0- A 3.2.1 qop-options auth-int
 2t  0y 11n  0-  2t  5y 10n  0-  1t  0y  7n  0- A 3.2.2 Authorization Digest
 1t  0y 12n  0-  0t  1y 16n  0-  0t  0y  8n  0- A 3.2.2 request qop auth
 1t  0y 12n  0-  0t  0y 17n  0-  0t  0y  8n  0- A 3.2.2 request qop auth-int
 2t  0y 11n  0-  2t  2y 13n  0-  1t  0y  7n  0- A 3.2.3 Authentication-Info Digest
 1t  0y 12n  0-  0t  2y 15n  0-  0t  0y  8n  0- A 3.2.3 response qop auth
 1t  0y 12n  0-  0t  0y 17n  0-  0t  0y  8n  0- A 3.2.3 response qop auth-int
 2t  0y 11n  0-  1t  1y 11n  4-  1t  0y  7n  0- A 4.2   Proxy-Authenticate Digest
 0t  1y 12n  0-  0t  0y 13n  4-  0t  0y  8n  0- A 4.2   Proxy qop-options auth
 0t  1y 12n  0-  0t  0y 13n  4-  0t  0y  8n  0- A 4.2   Proxy qop-options auth-int
 2t  0y 11n  0-  1t  1y 11n  4-  1t  0y  7n  0- A 4.2   Proxy Authorization Digest
 0t  1y 12n  0-  0t  0y 13n  4-  0t  0y  8n  0- A 4.2   Proxy request qop auth
 0t  1y 12n  0-  0t  0y 13n  4-  0t  0y  8n  0- A 4.2   Proxy request qop auth-int
 1t  1y 11n  0-  1t  0y 12n  4-  1t  0y  7n  0- A 4.2   Proxy Authentication-Info Digest
 0t  1y 12n  0-  0t  0y 13n  4-  0t  0y  8n  0- A 4.2   Proxy response qop auth
 0t  1y 12n  0-  0t  0y 13n  4-  0t  0y  8n  0- A 4.2   Proxy response qop auth-int

				- Jim Gettys
Received on Tuesday, 2 June 1998 14:36:10 EDT

This archive was generated by hypermail pre-2.1.9 : Wednesday, 24 September 2003 06:33:18 EDT