Jon Udell, who I know
to be an avid
(well… he’s pretty much
avid about everything he
deals with, which is one
of the main reasons
we read him) suggests
in a recent post
a new addition to his
list of attributable truthisms:
Scott Dart, program manager for the Vista Photo Gallery: “The truth is in the file.”
The post references Microsoft’s new
photo gallery facility in Vista
Jon goes on to say:
And when I look at this it strikes me that here, finally, is a microformat that has a shot at reaching critical mass.
Before deconstructing this optimism, it might be useful to invoke the microformats home page to remind
ourselves what a microformat is supposed to be:
Designed for humans first and machines second, microformats are a set of simple,
open data formats built upon existing and widely adopted standards.
Jon invokes the following embedded fragment from a Vista-truthy file:
This is a lot of code to unpack in order to conclude that the photo is a flower. Especially for a human.
But Microsoft deserves kudos for adopting standards established by others
(W3C, DCMI, Adobe), though it adds some of its own (non-interoperable) terms as well.
Examining the namespace declarations is interesting. In order of appearance:
- http://www.w3.org/1999/02/22-rdf-syntax-ns# Resolves to the RDF schema declaration of RDF
- http://purl.org/dc/elements/1.1/ Resolves to an RDF schema declaration of Dublin Core metadata
- http://ns.microsoft.com/photo/1.0 Does not resolve: server not found
- http://ns.adobe.com/xap/1.0/ Does not resolve: page not found
So, while I agree with the design philosophy that Jon commends in the Vista application, and Microsoft deserves kudos for adopting and declaring existing standards, this particular fragment leads to several unsettling concerns:
- The code fragment isn’t really all that simple. In fairness, the overhead looks a lot more reasonable in the case of a richer set of descriptors, and image metadata is rich. Still, it’s a bit of a stretch to rule this within scope of a microformat. Not to say it is unuseful, but simple and human readable? Probably too much to expect from metadata.
- While the code does
indeed refer to existing open standards, some of their declarations are not Web accessible and it is not evident at a distance whether they were intended to be or not. This is one of the arguments for separating identifiers from resolution addresses. When I see a URL, I expect it to point to a web-accessible location. If it doesn't, it seems a fair conclusion that something is broken. Note that elsewhere in this fragment there are identifiers that do not promise resolution, and hence do not disappoint expectations:
- There are a lot of dependencies here, requiring confidence that a somewhat complicated and fragile technical scaffolding is going to be maintained equally well by all parties. One may argue that these namespace declarations are simply identifiers, and needn’t be resolved by most applications, but this assumption leads to unfortunate and unnecessary uncertainties. This is certainly not a Vista flaw... it is a consideration in all our distributed applications.
- Finally, I am skeptical
that we really want multiple sources of potentially redundant metadata written into our
image files. For example,
presumably stands for a user-assigned quality rating… stars, probably. Again, good that it is in the file (this is the sort of data I lost in my unhappy experience with the Lightroom beta). But is iPhoto going to interpret Microsoft’s ratings, and vice versa? I doubt it. Back to the standards room, folks?
As a photographer of some avidity myself, I would be thrilled to imagine that image metadata might be the domain of the first microformat killer ap. As someone who lives in two OS domains, and expects to for the foreseeable future, I have a particularly strong desire to see this stuff work in a cross-platform way. I think we’re still a long way away.
Image: Ceiling detail in the ruins of Kaiser Wilhelm Church in Berlin, which stands as a memorial of remembrance of WWII in the center of the city.