W3C home > Mailing lists > Public > www-validator@w3.org > May 2013

Re: Validator ask for bzip2, but barfs when it gets what it asked for

From: Jukka K. Korpela <jkorpela@cs.tut.fi>
Date: Mon, 27 May 2013 19:03:28 +0300
Message-ID: <51A383D0.9030103@cs.tut.fi>
To: Klaus Alexander Seistrup <klaus@seistrup.dk>
CC: W3 Validator <www-validator@w3.org>
2013-05-26 17:20, Klaus Alexander Seistrup wrote:

> I was checking my identity page for errors, but Validator barfed:
>
>   »Sorry! This document cannot be checked.
>
>    A fatal error occurred when attempting to decode response body
>    from http://klaus.seistrup.dk/. Either we do not support the
>    content encoding specified (“bzip2”), or an error occurred while
>    decoding it.

Contrary to what you write in the Subject line, the validator does not 
ask for bzip2. It just announces that it can handle some encodings, with 
x-bzip2 among them.

>    The error was: Don't know how to decode Content-Encoding ‘bzip2’«
>
> Here's the request headers from Validator:
[...]
>    Accept-Encoding: gzip, x-gzip, deflate, x-bzip2.

The names “bzip2” and “x-bzip2” are not the same, although they are 
probably used to denote the same encoding. By normal Internet 
principles, any unregistered encoding should have a name that starts 
with “x-”, but perhaps the validator could be a bit more liberal and 
accept “bzip2” too – it can hardly mean anything else.

On the other hand, the server is in error if it sends “bzip2” when the 
request does not list it.

I’m not sure what happens here otherwise – both Firefox and Chrome get a 
response that has “Content-Encoding:gzip” and “Vary: Accept-Encoding”, 
and they don’t specify x-bzip2 or bzip2 in their Accept-Encoding request 
header.

Yucca
Received on Monday, 27 May 2013 16:04:04 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 1 March 2016 14:18:08 UTC