This is a very basic and preliminary road map as we head towards a version 1.0 release of hkit. Contributions are always welcomed.
Version 1.0
These are the primary objectives that need to be fulfilled to reach hkit version 1.
- Solid hCard implementation (pass all tests: http://microformats.org/tests/hcard/)
- Support for XFN, minimum rel="me"
- Support for multiple class instances within a script
- User documentation for creating format profiles
- Remote document caching
- Representative hCard support
Immediately beyond 1.0
- Native JSON output option
- hReview implementation (pass all tests: http://microformats.org/tests/hreview/)
- hCalendar implementation (when tests become available)
- Support for parsing for multiple compound microformats at once (e.g. hCard within an hReview)
It would be interesting if it were possible to do interop/develop test suites for hKit, hPricot and mofo... so that we begin to have solid, multi-language libraries that all work the same and use approximately the same nomenclature for methods and so on.
Also, representative hcard/hcard-uid would be very useful.
I have some XFN test data (for both valid and invalid relationships) at:
http://deliciouslymeta.com/projects/xfn/test_data.html
Chris Wanstrath uses this for mofo. It does not inlcude any hCard, however - probably a good next step. Also, I should add that data to the DiSo? repository.
The top issue I have seen with use of hKit, especially for social network portability, is correctly identifying the representative hCard for a page, given a page with many hCards (e.g. Twitter profiles, Pownce profiles).
I suggest that a 0.6 release include the following: representative hCard support
Such a revision would immediately benefit users of hKit, such as those that are implementing importing and subscribing to hCard profiles with hKit.
tantek +1
Agreed, representative hCard support is now a requirement for version 1, and hopefully will be in svn much sooner than that. (Working on it.)
I've written an hresume profile that works with hkit. Let me know if you want it :)