whosonfirst-data / whosonfirst-data-venue

Venues in Who's On First.
Other
4 stars 0 forks source link

Missing meta-data file on some of the repos #3

Open stuartlynn opened 7 years ago

stuartlynn commented 7 years ago

Some of the venue specific repos don't seem to have a metadata file in the meta folder, just the concordance files.

Is this data stored somewhere else or is there a reason its missing for these reops?

Specifically these ones:

https://github.com/whosonfirst-data/whosonfirst-data-venue-in https://github.com/whosonfirst-data/whosonfirst-data-venue-us-in https://github.com/whosonfirst-data/whosonfirst-data-venue-us-al https://github.com/whosonfirst-data/whosonfirst-data-venue-us-ar https://github.com/whosonfirst-data/whosonfirst-data-venue-us-co https://github.com/whosonfirst-data/whosonfirst-data-venue-us-pa https://github.com/whosonfirst-data/whosonfirst-data-venue-us-de https://github.com/whosonfirst-data/whosonfirst-data-venue-us-hi https://github.com/whosonfirst-data/whosonfirst-data-venue-us-ne https://github.com/whosonfirst-data/whosonfirst-data-venue-us-id https://github.com/whosonfirst-data/whosonfirst-data-venue-us-ia https://github.com/whosonfirst-data/whosonfirst-data-venue-us-az https://github.com/whosonfirst-data/whosonfirst-data-venue-us-tn https://github.com/whosonfirst-data/whosonfirst-data-venue-us-ks https://github.com/whosonfirst-data/whosonfirst-data-venue-us-ky https://github.com/whosonfirst-data/whosonfirst-data-venue-us-la https://github.com/whosonfirst-data/whosonfirst-data-venue-us-ms https://github.com/whosonfirst-data/whosonfirst-data-venue-us-ok https://github.com/whosonfirst-data/whosonfirst-data-venue-us-ri https://github.com/whosonfirst-data/whosonfirst-data-venue-us-mo https://github.com/whosonfirst-data/whosonfirst-data-venue-us-nc https://github.com/whosonfirst-data/whosonfirst-data-venue-us-oh https://github.com/whosonfirst-data/whosonfirst-data-venue-us-wi https://github.com/whosonfirst-data/whosonfirst-data-venue-us-mt https://github.com/whosonfirst-data/whosonfirst-data-venue-us-nv https://github.com/whosonfirst-data/whosonfirst-data-venue-us-nh https://github.com/whosonfirst-data/whosonfirst-data-venue-us-nm https://github.com/whosonfirst-data/whosonfirst-data-venue-us-nd https://github.com/whosonfirst-data/whosonfirst-data-venue-us-fl https://github.com/whosonfirst-data/whosonfirst-data-venue-us-me https://github.com/whosonfirst-data/whosonfirst-data-venue-us-sc https://github.com/whosonfirst-data/whosonfirst-data-venue-us-sd https://github.com/whosonfirst-data/whosonfirst-data-venue-us-ut https://github.com/whosonfirst-data/whosonfirst-data-venue-us-vt https://github.com/whosonfirst-data/whosonfirst-data-venue-us-nj https://github.com/whosonfirst-data/whosonfirst-data-venue-us-wv https://github.com/whosonfirst-data/whosonfirst-data-venue-us-wy

thisisaaronland commented 7 years ago

So the short version is: That's a bug. I will try to get that sorted today (maybe tomorrow).

The longer version is: In the interim the easiest thing to do is use the wof-build-metafiles tool to rebuild the metafiles locally. You'll need to have Go installed to build the tool but otherwise all the dependencies are vendored in its parent package:

https://github.com/whosonfirst/go-whosonfirst-meta#install https://github.com/whosonfirst/go-whosonfirst-meta#wof-build-metafiles

The long version is: There is talk about moving the meta files out of the repos themselves as they've gotten ridiculously large in some cases (us-ca for example) and require people to deal with git lfs weirdness. Nothing will change in the near future but we're thinking of a) publishing bz2-ed versions online b) bundling binary versions of wof-build-metafiles with each repo and letting people generate the metafiles on demand. But nothing has been decided yet.

In the meantime, I will sort out why those repos are missing their metafiles.

stuartlynn commented 7 years ago

Awesome thanks! Will try processing the metadata locally, will be good to know how to do that anyway. Thanks for the quick reply.

The binary version of the metadata might be a good idea. Or even if they were just available on S3 somewhere that would be good too.