[rfc-i] reject the past ( was Re: Fwd: New Version Notification for draft-flanagan-plaintext-00.txt)

Dave Crocker dhc at dcrocker.net
Thu Jun 26 08:43:36 PDT 2014

On 6/24/2014 2:37 PM, George, Wes wrote:
> On 6/24/14, 3:29 PM, "Dave Crocker" <dhc at dcrocker.net> wrote:
>>  The document essentially
>> makes ascii art of questionable utility.
> WG] No, that happened as soon as something better was widely available,
> many, many years ago. This is just admitting that such newfangled
> technology exists and pretty much works. ;-)

This nicely demonstrates a basic problem in these discussions, namely
ignoring demonstrated utility over many years.

It encourages embracing whatever shiny new capability attracts us now,
rather than considering carefully the trade-offs between whatever
benefits have been present in the long-standing capability, versus
whatever is appealing about a new capability.

In the current context, the above effectively is a call for the
elimination of txt documents.

If we are to retain txt documents, they need to be useful.  If they are
to be useful, the art they contain must be useful.

> WG]   ... It should NOT be a primary goal
> that page numbers or indeed pagination be retained. It should be a primary
> goal that there be useful anchors that are visible in multiple display
> formats such that they identify the same location in the document's text
> regardless of the method used to render it. 

+1, and nicely phrased.

>> We need to be much more clear about the reason for doing graphics art.
>> If it is for prettiness, then we shouldn't make the switch.  There is no
>> evidence that what we've done using ASCII art has not been sufficient,
> WG] You're dramatically oversimplifying things. The process required to
> get useful ASCII art that meets the formatting requirements breaks the
> workflow most people use to author modern documents that can support
> vector graphics.

The process of producing useful RFCs breaks the workflow of most folk
who are not already IETF geeks.

We presume some benefits in the RFC model, so I guess I don't see your

> Even when using a tool to do so, generating ASCII art
> diagrams wastes a lot of time creating output that is mediocre at best at
> the task of conveying the visual information it's trying to convey, in
> order to meet limitations that are not and have not been due to the limits
> of technology for ~2 decades. Do our ASCII diagrams convey the point their
> authors intended? Mostly, but that's a pretty low bar.

Interoperability often requires choosing a least common denominator.

What you call 'wastes a lot of time' I tend to call 'forces a discipline
in representing what is essential'.

I was impressed in the very early days of laser printers -- there were
just a few outside of PARC, as I recall -- to see a grad student produce
a summary of his work, for his department review seminar, using many
different fonts.  The document was so pretty, I almost missed the fact
that it was nearly content free.  But of course, that was the point.

Presentation restrictions often force us to do extra work at being clear
and useful.


Dave Crocker
Brandenburg InternetWorking

More information about the rfc-interest mailing list