Open olexandr-konovalov opened 6 years ago
The new curlInterface package should provide a way to fix this.
I think the no it wasn't, but still, this second point is also in issue #3 :-)curlInterface
comment here was maybe intended for issue #3?
By the way, I would recommend to put that data into this repository anyway; you can then add a .release
script which removes it if for the release tarball. But this way, you have a guaranteed backup of it. Also, when downloading the files, that could be done via the GitHub server instead of a university server.
@fingolfin yes, 32 MB unpacked in data/243
, 18 MB tar,.gz or 16 MB tar.bz - which is better?
git doesn't handle binaries very well, so I'd always put the unpacked data into the repo; .git compresses all blobs anyway, but this way it can e.g. handle diffs in the data should you ever have to tweak it.
First, these need to be recalculated in the same way as #7. Second, the ability to retrieve them remotely should be fixed.