This post originated from an RSS feed registered with Agile Buzz by James Robertson. Original Post: Not exactly Feed Title: Cincom Smalltalk Blog - Smalltalk with Rants Feed URL: http://www.cincomsmalltalk.com/rssBlog/rssBlogView.xml Feed Description: James Robertson comments on Cincom Smalltalk, the Smalltalk development community, and IT trends and issues in general. Latest Agile Buzz Posts Latest Agile Buzz Posts by James Robertson Latest Posts From Cincom Smalltalk Blog - Smalltalk with Rants

Dave Winer talks about Postel's Law:

Postel's Law has two parts. This is something a lot of people don't want you to look at, they only want you to think about the first part, the part which XML says is not a great idea -- be liberal in what you accept. This tends to favor the big guys who have the resources to catch up, and then the chutzpah to throw a big fat hairball into the middle of the market, one that no one else can handle. Maybe Postel didn't live in a world where these big companies could create such big messes, but I have had to deal with them, many times in my career, and they usually end competitive markets. A lot of well-intentioned people in the syndication community don't have the benefit of this experience, and we may have to learn this once again for their benefit. I hope not.

I'm not so sure that being liberal in what you accept favors "the big guys". They are all using sealed, immovable XML frameworks - which means that they have to hack up their own parser to deal with things rejected by the library parser. I've mentioned before that I don't have to do that. See, Smalltalk doesn't have bad ideas like "final" classes. To deal with malformed XML, I simply subclassed the VW XML parser. In the methods where fatal errors were being raised due to bad characters (and such like), I simply have the parser "move on", and flag an error (so that the 3 people who care can report it). This required virtually no effort on my part, and I haven't so much as touched that parser code in 6 months (possibly more; it was awhile ago). The people who say that this is a lot of effort do so from the perspective of the kinds of effort that is required with the mainstream tools. Look outside the mainstream a little and you'll find that the "hard" things just aren't as hard as you thought (not that VW is without flaws... like any other language/library platform, it has its own set of issues).

Bottom line - the "rules" for how to deal with malformed content are going to differ based on context. In a consumer context (most aggregator usage), you'll favor the accept almost everything idea. In a business context, you'll favor the reject anything malformed idea.