On Thu, Feb 2, 2012 at 7:07 AM, Tony Yu <tsyu80@gmail.com> wrote:

I should have a little bit of free time to work on the docs this weekend. Any specific goals?

I think we should try and complete the following sections: - Introduction (installation -- link to existing docs, overview and how we fit in on top of scipy, our goals, what we provide) - Reading and displaying images (io module, display, fancy=True, using matplotlib, etc.) - Data-types and the pipeline (dtype conversion, how to ensure your output is in the right format, all functions accept any input but returns preferred dtype output--user should call img_as_*) - Getting help (reference guide, examples gallery, maybe image processing specific websites similar to metaoptimize.com/qa ) And some tutorials are always welcome--although I think the gallery now largely fulfills that role. Did I miss anything important? We also need a man-page for skivi (the image viewer). I've uploaded a template: https://github.com/scikits-image/scikits-image/blob/debian/debian/skivi.1

One thing I'd like to see is a package docstring listing all subpackages (+ 1-sentence descriptions of those subpackages; e.g. type `np?` in ipython); and then each subpackage would have it's own docstring listing of functions (e.g. type `numpy.linalg?` in ipython). I guess this is separate from the user guide, but it's still really valuable for learning the package.

Yes, good point!

While we're talking about release 0.5, I think we should try to clean out a number of pull requests. Some of them should be quick:

PR 99: Rewrite dtype._convert function

Merged (by me).

PR 105: ENH: Allow choice between 4 and 8 neighbor mode.

Merged (by Tony).

PR 95: debian/control updates to reflect name space transition

Rejected (mainly adding the author's name, but very few changes). I updated the control file to be up to date in the meantime.

PR 107: Add rescale_intensity function with test and example.

I'm keen to merge this; we're just waiting for the rest of the list to comment on the matplotlib 1.0 dependency. Cheers Stéfan