I'm behind this 100%. (Though I want to make sure that imread supports 3D tiffs, as Christoph Gohlke's tifffile does.) Do we want to add a paragraph about this to the paper? It seems like a fairly trivial detail, but it also seems perverse having information there that we know will quickly be obsolete... On Thu, May 15, 2014 at 9:10 PM, Stéfan van der Walt <stefan@sun.ac.za>wrote:

Hi all,

Juan recently again brought to my attention the dismal state of the scikit-image I/O infrastructure. This needs to be fixed for 0.11 (0.10 is almost out the door--this weekend, I hope!).

We are currently relying on other libraries to do the job, and while it allows you to do "from skimage import io; io.imread(anything)" on almost any platform, it's just not working--we need reproducible backends.

My current thinking is thus: make Luis's "imread" [0] a dependency, and remove all plugins from scikit-image. Users can, of course, still load data using mpl, imread, astropy or whatever suits them. In the documentation, we may provide recipes for handling files like ITK, gdal etc. if that proves useful.

What do you think?

Stéfan

[0] https://github.com/luispedro/imread

-- You received this message because you are subscribed to the Google Groups "scikit-image" group. To unsubscribe from this group and stop receiving emails from it, send an email to scikit-image+unsubscribe@googlegroups.com. For more options, visit https://groups.google.com/d/optout.