Agree: let's not overextend

Drummond Reed drummond.reed at cordance.net
Mon Nov 28 13:11:50 PST 2005


+1. It's worth noting that the XRI effort at OASIS has followed Johannes's
logic. Although our initial ambitions were quite broad, for the past three
years we have kept scoping back and scoping back and breaking the specs into
discrete, reusable components that can each stand on their own. 

For example, XRI Syntax 2.0, which goes up for a full OASIS vote in
December, can be used without any relationship to XRI resolution or XRI
metadata.

A second good example is XRDs -- this simple resource descriptor format can
be used independently of XRIs altogether.

So the smallest first steps that produces significant value is a great place
to start.

=Drummond 

-----Original Message-----
From: yadis-bounces at lists.danga.com [mailto:yadis-bounces at lists.danga.com]
On Behalf Of Ernst Johannes
Sent: Monday, November 28, 2005 8:59 AM
To: YADIS List
Subject: Agree: let's not overextend

Some wise comments were made on this list by various people that I'd  
like to echo.

The safest way of killing of a promising new initiative is to attempt  
too many things at once, or to create more dependencies than  
absolutely necessary. It's great to hear that many people on this  
list are aware of this danger and want to approach this as a series  
of small, incremental steps, focusing on making that small step  
successful first before going to the next step.

Even if the first step was very small in terms of technical agreement  
reached, it might turn out to be very big because it may affect a  
very large number of URLs -- as it was pointed out, likely far beyond  
the subject of digital identity (for people). Funnily enough, the  
smaller that first step, the more likely is its appeal going to be  
very broad.

It it turns out that we can agree on many small steps -- not too  
unlikely -- then we can take many small steps in rapid succession.  
And if step N+1 suddenly turned out to be not as easy, we can stop  
right there -- without stalling anything else other than step N+1.

I like this approach to standards development ...



Johannes Ernst



More information about the yadis mailing list