As we diving into the new and brave world of tens of thousands of subdatasets for some studies, I thought it would have been neat if similarly to size, web ui (as of http://datasets.datalad.org instance) "aggregated" number of subdatasets (immediate/recursive) for each dataset. In principle, in the bright future where metadata is "automagically" aggregated for each and every dataset such information could be picked up from metadata. But meanwhile just using how it is done by web ui's helper (well ls --json triggered via git hook for post-update) - we could gain that additional information in existing setup and not being dependent either metadata was aggregated. E.g. ATM I have no clue how many subdatasets we have already on http://datasets.datalad.org/
As we diving into the new and brave world of tens of thousands of subdatasets for some studies, I thought it would have been neat if similarly to size, web ui (as of http://datasets.datalad.org instance) "aggregated" number of subdatasets (immediate/recursive) for each dataset. In principle, in the bright future where metadata is "automagically" aggregated for each and every dataset such information could be picked up from metadata. But meanwhile just using how it is done by web ui's helper (well
ls --json
triggered via git hook for post-update) - we could gain that additional information in existing setup and not being dependent either metadata was aggregated. E.g. ATM I have no clue how many subdatasets we have already on http://datasets.datalad.org/