posted by Adam S on Tue 7th Nov 2006 12:58 UTC

"The Future of XML, Part II"

An RSS feed in Firefox 2Lastly, I searched for workarounds for this behavior in FF2, and found one that amounts to little more than a dirty hack. Firefox determines a feed by sniffing the first 512 characters of a file and searching for the string "<rss" or "<feed". If either is found, the XML is presented as a feed with the feed stylesheet. The unofficial workaround is referred to as "byte-stuffing," and boils down to adding at least 512 characters (spaces included) to your feeds before the opening or tag. In my case, I've used an XML comment on my site's feed, because the default Firefox stylesheet can be confusing. It's hard to even tell where each entry ends, it can be deceiving: like this feed (pictured at right), for example, if you're using Firefox 2. Who can understand what's going on there?

Knowing how feeds are generally handled - which is vastly different from browser to browser, I posted my piece to the USENET group mozilla.dev.apps.firefox. It has become the longest thread in quite some time in this group. There, we're had some very interesting conversation. The Mozilla developers argue that RSS and Atom is intended to be read by a feed reader, and therefore, presenting a consistent and easy to use interface to users is the smartest path and is an intentional design decision. The opponents argue that Mozilla has overridden developers, deprecated an existing valid XML construct, and in the meantime, prevented any Firefox user from ever seeing styled XML (at least, RSS and Atom) as intended. In short, it now cannot be read by a human with intended style, as the style is not accessible - not via a preference, not via about:config, not via a menu. There is no way a feed can be read without hacks even if the client wants to. Firefox thinks it knows better.

For the record, similar discussions have occurred in different venues regarding IE7. As seen here, it's been the same conversation, although over on Channel 9, most seem to agree this behavior is not acceptable.

Either way, the USENET discussion appears to be at a stalemate. My final arguments included the notion that as technology becomes popular, the designers often tend to cater to a wider userbase and alienate the very users who helped grow the application to begin with. There are several other respondents who had similar viewpoints. Ultimately, right or wrong, the developer standpoint seems to be that RSS and Atom are not intended for human consumption and should be styled in a manner that makes it easier for a user to subscribe to a feed automagically. All of this is fine, except it breaks the system for anyone who has come to use and rely upon XML as it has been used for the last several years and is still valid according to W3C standards. Sites that have worked for years that are extremely styled will no longer work without a hack.

Which brings us to the underlying question - without regard to this debate - the question is, when common use begins to define a course of action, but there is already a "de facto" standard - that is, accepted as the common manner of use - how can we/should we go about changing it? Furthermore, who gets to do it?

It appears the Mozilla Foundation is going to push XML subscription forward kicking and screaming, but in the process, they are going to anger at least some group of developers and webmasters. In the process, they could potentially thrust a new technology to a a whole new group of users - a net positive, I'm sure we'd all agree. However, it's equally likely that RSS will never catch on with average users, but the same content providers will remain angry about this type of decision.

Table of contents
  1. "The Future of XML, Part I"
  2. "The Future of XML, Part II"
  3. "The Future of XML, Part III"
e p (4)    24 Comment(s)

Technology White Papers

See More