It might actually even be factually incorrect since might have been done not from CRON job. So let's replace it with a composition of information on number of added or updated dandisets and either there were other updates (could be just "other updates" if any) which aren't to submodules in the summary line. E.g.
dandisets: 1 added (000544), 2 updated (000023, 00026); other updates
omitting listing of individual dandisets if e.g. more than 3.
And then completing the body of the commit with summaries of the commits per each dandiset, e.g.
000054:
- [backups2datalad] 20 files added
- Instruct annex to add text files to Git
- [DATALAD] new dataset
000023:
- [backups2datalad] 20 files deleted
...
I think we should not bother describing other updates in the details since they would be pertinent to the dandisets superdataset thus easily interrogated using git log.
This would allow much better overview of introduced changes than git log --stat for the subdatasets ATM, which requires installation of subdatasets of interest and use of datalad diff to gather more detail about changes.
It might actually even be factually incorrect since might have been done not from CRON job. So let's replace it with a composition of information on number of added or updated dandisets and either there were other updates (could be just "other updates" if any) which aren't to submodules in the summary line. E.g.
omitting listing of individual dandisets if e.g. more than 3.
And then completing the body of the commit with summaries of the commits per each dandiset, e.g.
I think we should not bother describing other updates in the details since they would be pertinent to the
dandisets
superdataset thus easily interrogated usinggit log
.This would allow much better overview of introduced changes than
git log --stat
for the subdatasets ATM, which requires installation of subdatasets of interest and use ofdatalad diff
to gather more detail about changes.