Dave Winer has made an offer to work with the Atom folks to create a single, unified standard drawing from both formats. Will the initiative succeed? I have my doubts.

There are three types of issues facing such a proposition: technical issues, goals, and personalities. I'll start off with an issue that relates to all three, and get to some other issues tomorrow.

For Dave Winer, backwards compatibility is of paramount importance. While I agree that backwards compatibility is of great value, I believe there are times where it must be sacrificed. You don't want to throw the baby out with the bath water, but you do want to throw out the bath water. You don't put clean clothes on a muddy baby.

The discussion that ensued after Dave posted his offer included a post by Dave that I found surprising. He commented that if he had to choose between elegant (dropping backwards compatibility and coming up with a better standard), modular (adding a new, well spec'd element to replace an old poorly spec'd one, but continuing support for the old one) and kloogy [sic] (a method which involved including junk data in a field and telling the tools to ignore it in order to preserve backward compatibility), he would choose the kludge. To me, that's like adding mud to the muddy baby. Henceforth, all tools would have to know to ignore the standardized bogus data, which would add bloat to each feed that used it.

The "modular" solution wouldn't be much better--sort of like putting clean clothes on the muddy baby. The baby would still be muddy, and the clothes would get muddy too. Henceforth, all tools would have to support two ways of doing the same thing, diluting the value of the new way.

If the weaknesses of RSS were like a little dirt under the fingernails, it wouldn't matter so much, but for many people, they are worse than that. In such a case, the elegant solution seems best. Wash the baby, and give him clean clothes. Don't make everyone maintain backwards compatibility with mud.

In all fairness, having multiple standards would require tools to be compatible with multiple standards if they wanted to support everyone. But moving to a clean standard paves the way for phasing out the shortcomings of earlier formats. Would it be painful? Sure. But so was switching from typewriters to computers, from teletype to fax, from BBS to internet. It behooves anyone who develops a new standard to do so carefully, making every effort to come up with a standard that will endure. But when the weight of the old becomes too great, there comes a time to cast it off and leap forward again.