W3C home > Mailing lists > Public > ietf-http-wg@w3.org > January to March 2009

Re: Weak ETag handling

From: Roy T. Fielding <fielding@gbiv.com>
Date: Tue, 3 Feb 2009 11:17:57 -0800
Message-Id: <EE9F07F2-158C-4574-8F9C-FCEDAC141B95@gbiv.com>
Cc: ietf-http-wg@w3.org
To: Lisa Dusseault <lisa.dusseault@gmail.com>

On Feb 3, 2009, at 10:22 AM, Lisa Dusseault wrote:
> I know I've ranted about this before, but just ran across it in  
> code again today:
>         etag.weak = (value[:2] == "W/")
>         if etag.weak:
>             etag.tag = value[2:]
>         else:
>             etag.tag = value
> Stripping the "W/" off a weak etag to get a strong one, because it  
> works with Apache.
> This bothers me to an unreasonable extent.  Is there nothing we can  
> do about this?  Redefine the standard to match practice?  Clarify  
> the standard to condemn this practice?

It is hard to have the standard specify restrictions on nonstandard
behavior (assuming that etag.week is not being used later to modify
its behavior).  At best we could clarify all the reasons that a tag
might be called weak and explain the risks of ignoring that weakness.

Received on Tuesday, 3 February 2009 19:18:34 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 1 March 2016 11:10:48 UTC