White on Orange XML Icon? Dave Winer makes some good points about the White on Orange XML Icon, in response to Google's Jason Shellen' proposal. A good discussion, that I can relate to, as it was only a week or two ago that I found out, quite by accident, what the purpose of the White-on-orange Coffee Cup Icon was supposed to do. IMHO, that is the right idea, if we generalize it somewhat.
Here's my reasoning. Yes, indeed, the XML button is obscure and techie. But so are URIs, Email addresses, and File paths. On the one hand, and probably unfortunately, the world at large has learned (or is learning) how to deal with those things. But one reason is that while they exist to some extent as plumbing ('leaky abstractions' perhaps) they all have a very elegant UI veneer over them which allows people to live their life quite happily and get work done without understanding them.
You click on a link and you go to the page. Oh, but you copy a url out of a magazine article and you also go to the page. And also, if that link is to a blog or blog item, that also works just like you would expect. One reason that this works so elegantly is that the link lives in a browser page and launches a browser page.
Now we need a new action, click to subscribe. Also a reasonably easy concept to understand. However here's the rub, the tool you use to access your subscription may not be the browser. It might be a standalone blog reader.
By analogy, click on a link to send an email. Also a reasonably easy concept to understand. However your mail client may not be the browser. This is solved with a mailto link. Or, click on a link to a newsfeed. Solver with a news:// URI. This is a direction to explore.
So in summary: We need to work towards a 'click to subcribe' button. I strongly believe this should be a white on orange icon as we have successfully brought that at least somewhat into the vernacular. But instead of navigating to an obscure page of XML, it should trigger an extensible mechanism which will pass the feed to the appropriate feed reader, transparently.
Update: There are no new ideas (I should have known it.) I stumbled upon quite a discussion of this very idea from Greg Reinacker, and I am sure elsewhere. I want to respond to Greg's 'deal killer' problem:
Problem 1: [severity: deal-breaker] In order to serve up a file with a specific MIME type, you need to make some changes in your web server configuration. There are a LOT of people out there (shared hosting, anyone?) who don't have this capability. We have to cater to the masses, people - we're trying to drive adoption of this technology.
Yes, this approach wont happen overnight, however, I tend to think more long term on this particular point. We are at the very inception, in the process of popularization of blogging or open source writing or whatever you wish to call it.
This is just the time to push for and accept a bit of disruption to get to a place that will allow the kind of universal adoption that we, in our wildest moments might dream of. And by the way, imagine if Google or Yahoo or Microsoft chose to implement this approach.