bgilbert at cs.cmu.edu
Thu Apr 14 20:08:27 EDT 2016
On Thu, Apr 14, 2016 at 04:40:33PM +0200, Andreas Schmid via openslide-users wrote:
> I am wondering if it would be possible to adapt the Openslide library to
> support also url’s?
> I was able to partially read a zip file via a seekable http range stream.
> I think reading WSI-Formats over http without a tileserver would be amazing.
In principle it should be possible, but it would be a lot of work. The
entire library revolves around opening and reading files, so a lot of code
would need to be changed. In addition, performance would probably not be
very good, since every read access would need to go to a server potentially
tens of milliseconds away. A good cache could help to some extent, but
OpenSlide's reads can be small, numerous, and random, so the problem is
non-trivial. We've considered HTTP support in the past, but concluded that
putting a tile server next to the data would be more efficient.
It's also possible to implement something like this without modifying
OpenSlide, by reading slide files from a virtual filesystem backed by HTTP.
There should be existing FUSE filesystems that would work.
More information about the openslide-users