The format should fit the task and I don't disagree that you can't have internal relational datastores as well.
At the same time, I had this fight with the Adium folks as well -- who went off and invented their own XML-based schema for storing chat transcripts when people using that app are actually watching a WebKit window scroll HTML content. Why did they do that? According to them, XML is easier to parse. Why is a "whatever" argument. Because in the case of text-based data storage mechanisms, XHTML will have reuse potential that you're not going to get with some one-off format. If you want to use a real database backend, great. Go for it. But if you're going to be storing simple data in a text format that will inevitably be more valuable when viewed in a browser, I think microformats become incredibly powerful and useful, especially when you asynchronous access to data that ultimately needs to be displayed and manipulated by humans. Chris On 9/14/06, Ian McKellar <[EMAIL PROTECTED]> wrote:
On 9/14/06, Chris Messina <[EMAIL PROTECTED]> wrote: > It's a wonder that engineers ever built a craft to go to the moon. I've heard on good authority that was a hoax... > > Here's a usecase I've been throwing around recently: > > Take Ma.gnolia, which support hAtom and xFolk. Right there, you have a > datastore. Literally, in the HTML, that's rich, parseable data. No. You have an web page that's produced by running code against a datastore (probably sql). The whole point of a social bookmarking service is that you can make inferences based on the information. You can work out who is bookmarking the same pages as you, you can work out how tags correlate, etc. This requires you to store your information in some kind of data store. Now when you're publishing or sharing it makes sense to choose to present your information in some kind of standard way so you can have interoperability and Ma.gnolia have wisely and admirably decided to do this. (snip crack-rock "use cases") I believe in standards or conventions for communications between applications, but the idea that you should repurpose publishing formats as the internal representation of data for applications is foolish. Tantek, is microformatted content inside the Technorati search engine system internally represented as HTML? Ian (who is actually working on an interesting microformats + indexing + atom + opensearch side-project, which ironically will use microformats as its internal data format) _______________________________________________ microformats-discuss mailing list [email protected] http://microformats.org/mailman/listinfo/microformats-discuss
-- Chris Messina Citizen Provocateur & Open Source Ambassador-at-Large Work: http://citizenagency.com Blog: http://factoryjoe.com/blog Cell: 412 225-1051 Skype: factoryjoe This email is: [ ] bloggable [X] ask first [ ] private _______________________________________________ microformats-discuss mailing list [email protected] http://microformats.org/mailman/listinfo/microformats-discuss
