Closed gordonwatts closed 9 years ago
Ok, according to DAST this just isn't easily possible. So, we will have to implement some sort of pruning I guess. When there are one or two files I don't think it matters. But if you download half the dataset first, and then the whole thing, that seems like wasted bandwidth. :(
So, the toher thing to do is just only download the new files. This is the route I'm trying to take here.
This causes problems with finding the dataset, and duplicate files. Is this us or is this rucio?