THE Open Standards Stack for Data Portability? Really?

by Timothy Appnel

I just saw a link in the del.icio.us hot list that caught my attention called Data Portability, that bills itself as the open standards stack of the ubiquitous sharing and remixing of data.

Nice, but RSS and not Atom Syndication? OPML and not XOXO (The XHTML Outlines Microformat), JSON or YAML? And no Atom Publishing Protocol? "THE" stack? What is the criteria here?

This all seems rather arbitrary in what is "open" that I am suspicious of a hidden agenda at play here.

I'm all for open standards, data portability and anyone who wants to advocate it, but advocacy of "open" should be just that -- open and not so closed, arbitrary or apparently biased. (If it's not biased then someone didn't do their homework.) We have enough agenda disguised in open advocacy clothes already.


2 Comments

Chris Saad
2008-01-05 02:28:13
Timothy if you dig a little deeper into what we have built you will notice there is a reference design that explains how all the standards fit together. The stack is also still open to debate.


The point, though, is that the standards fit together and the debate should move on from 'If' or 'How' but rather when...


Chris

Timothy Appnel
2008-01-05 08:53:19
I didn't think I had to dig any deeper. The point you are making was not at all clear looking at the site that was up when I made my post. I also found the use of "the" rather presumptive. It's since changed and thankfully the tagline I was referring to is gone.


Also to your point other standards not included can also fit together to create a data interoperability stack -- arguably with the less effort and fuss. For instance, the two Atom formats support where made to work together and can carry any payload in its content element that has a MIME type. Look at Google's GData API as example of this potential.


This all makes me wonder what the criteria was for that groups proclamation that this is "the" stack. What vendors are backing the adoption of this stack as a whole for data interoperability? Whitepapers are nice, but ultimately code and implementations will decide.