Currently, the following things need to be done manually before serving a loom file:
metadata extraction (still has fallback code)
row expansion (still has fallback code)
column expansion (still has fallback code)
tile pre-rendering (no fallback)
These all happen through the command line. This is confusing for many (I already had to help Ulrika) and will halt adoption.
What we could do is make a special page where you could click on a button to send a simple PUT request to the server to run these expansions (this would be a forked process to not block the server).
Few notes:
in addition to existing metadata, we probably want to send data about whether metadata/rows/tiles are expanded or not. Especially the heatmap will be useful, so we can hide that view if the tiles haven't been expanded (preventing confusion)
all code for expansion is ready to go, and already has protection to safeguard against running multiple times (simply put: if the subfolder for row/columns/tiles exists, expansion is aborted unless explicitly told to overwrite it).
python-side, all we need is a way to send a request to expand a specific file
client side, the page is very simple (four buttons)
for the public site we probably want this to be hidden and behind authorisation
This is an enhancement that probably can wait until release.
Currently, the following things need to be done manually before serving a loom file:
column expansion (still has fallback code)These all happen through the command line. This is confusing for many (I already had to help Ulrika) and will halt adoption.
What we could do is make a special page where you could click on a button to send a simple PUT request to the server to run these expansions (this would be a forked process to not block the server).
Few notes:
This is an enhancement that probably can wait until release.