[rfc-i] feedback on draft-iab-styleguide-01

Russ Housley housley at vigilsec.com
Mon Apr 7 08:41:01 PDT 2014


>>>>    [ErrNNNN]  RFC Errata, Errata ID NNNN, RFC XXXX,
>>>>               <http:/www.rfc-editor.orgerrata_search.php?eid=NNNN>.
>>> 
>>> Again, that is misleading, as this confuses the RFC with the errata about the RFC.
>> 
>> This section of the Style Guide is about referencing errata.  The URL I offered goes to the errata, which is the thing being referenced.
>> 
>> For example:  http://www.rfc-editor.org/errata_search.php/?eid=1912
>> 
>> This provides the errata, and it includes a link to the RFC, which is RFC 2978 in this case.
> 
> That's the way it currently works, but using those URIs in long-lived documents is a bad idea for at least two reasons:
> 
> - The RFC Editor might later switch away from using PHP
> 
> - They may actually make each erratum statically accessible
> 
> Using URI such as the one you proposed have been considered bad practices in the web community for over 15 years.
> 
> Instead, using a URI that simply names the erratum directly would be far preferable. Something like;
>   http://www.rfc-editor.org/errata-by-number/NNNN
> and
>   http://www.rfc-editor.org/errata-by-rfc/NNNNN

I do like the idea of using a URL that does not include .php, but I think that a reference by the unique errata number is fine.

Russ




More information about the rfc-interest mailing list