hypertopic / steatite Goto Github PK

Whereas EXIF metadata extraction is already implemented for item API (see #8 and #1),
doing it the same way (on-the-fly) would not be efficient enough to be used in the corpus and attribute API.

Extracting EXIF metadata from large number of pictures at the same time requires probably to use a cache.

The corpus API should provide thumbnail and resource (as in item).

So far, there is only one fake corpus.

On our server, with more than 4000 thumbnails in a page, it gets hardly usable.

This bug was reported four times on our current demonstration server. We are still not sure why the ports do not get closed or reused properly.

This might be caused by our new proxy server (Cherokee), and should be tested with Apache instead.

This issue can be reproduced 100% of the time.

In items and attributes API, longitude and latitude are extracted from pictures.

In items and attributes API, Dublin Core creation date (created) could be extracted on the fly from picture metadata.

As in feature #1, exiftool could be used for that.