WACZ futurism: mimetype and Pronom ID
Opened this issue · 4 comments
Good morning!
Just an idea. As early adapters of the WACZ format we were thinking that it could be nice in the future to have a specific way we should identify WACZ (before any processing). Requesting a new mimetype to IANA seems a bit out of scope (or not?) but thinking of the data package inheritance that happens in WACZ and based on the extra arguments we we can pass to a mimetype
See : https://developer.mozilla.org/en-US/docs/Web/HTTP/Basics_of_HTTP/MIME_types and see https://github.com/frictionlessdata/specs/tree/master/data-package where datapackages' media type is
application/vnd.datapackage+json seems to be for the JSON? but not sure about the gzipped version of the full package? (I may be lacking coffee here)
application/vnd.datapackage+json;parameter=value
or
application/zip;parameter=value?
application/vnd.datapackage+zip or +gzip?
parameter = content?
value = webarchive?
Too liberal?
Anyhow. Just ideas. On our side we can start by using application/vnd.datapackage+gzip
Pronom goes a bit further by registering file content characteristics. Still important for Digital preservation to have at least some discussion, maybe someone from that realm could give us a hint
Thanks
Since effort is now underway to make WACZ into more of a standard I think working towards an IETF media type might actually be a good idea. Registering a media type doesn't require the standard be developed at IETF.
Pronom also came up today in another conversation, not entirely sure what is needed for that, but happy to explore if there is interest. Looks like submission form is here: https://www.nationalarchives.gov.uk/PRONOM/submit.htm
We're adding wacz detection (maybe parsing?) over on Apache Tika now. As a temporary placeholder at least, is application/wacz appropriate ?