[rfc-i] length of RFCs (was: URIs in RFC references)

"Martin J. Dürst" duerst at it.aoyama.ac.jp
Wed Mar 26 19:52:17 PDT 2014


On 2014/03/27 04:56, Ted Lemon wrote:
> On Mar 26, 2014, at 2:40 PM, John R Levine <johnl at taugh.com> wrote:
>> If we're doing kludges, it would also be reasonable to put the whole RFC into the page up to some size limit like 100K, and truncate it with a link for the rest beyond that.  I see that there are 873 RFCs larger than 100K, and about 6000 smaller.
>
> Right, and this will be exactly the wrong behavior in many cases, of no benefit in most cases, and will happen infrequently enough that it will be seen as an error and reported to the tools team every time it happens!   :)

Yes. People on slow connections are used to things being slow. They are 
also used to interrupt things when they are too slow. That's not a 
feature, but it's an (unfortunate) reality, and cutting off data doesn't 
improve that, because some people on slow connections may just want to 
read everything, even if they have to wait a bit. Even on slow 
connections, reading will still be slower than downloading.

On the other hand, people on fast connections will be truly surprised 
and annoyed if they have to click again somewhere in the middle of the 
document.

If the length of big RFCs is a fundamental problem, then I suggest that 
we communicate that to authors, and have them split things into pieces 
(e.g. like it was recently done for HTTP).

Regards,   Martin.


More information about the rfc-interest mailing list