To Do list for 5th - 9th June

[justinclift]

DBHub.io

• [ ] Update our internal data model - *in progress*
  • Working on it here: https://github.com/justinclift/dio
  • [ ] Integrate dio data model with dbhub.io backend
  • [x] Initial update of new data model to use PG & Minio
    • [x] db
    • [x] upload
    • [x] download
    • [x] list databases
    • [ ] remove - *still needs doing*
    • [x] branch
    • [x] create
    • [x] list branches for a database
    • [x] change branch description and allow description removal https://github.com/justinclift/dio/commit/c908924d3505e498c98f292781bbde3bd616682e
    • [x] remove
    • [x] history (log) https://github.com/justinclift/dio/commit/92c1ca530ab1ab0848cfe6b96698ae45bc91f15b https://github.com/justinclift/dio/commit/77a89cad75d991267f1613a28361714b93d711c7
    • [x] revert to prior database version
    • [x] set default branch https://github.com/justinclift/dio/commit/2ef235fd0e1572b110bdf4208dbcfc207279fea4
    • [x] get default branch https://github.com/justinclift/dio/commit/2ef235fd0e1572b110bdf4208dbcfc207279fea4
    • [x] tags https://github.com/justinclift/dio/commit/8b203b7993ff4ce73714a04dbc9e94c584b21f1e
    • [x] create simple
    • [x] create annotated
    • [x] list
    • [x] remove
  • [x] Add licence info
    • [x] Store licence info in the commit/tree entries and backend database
    • [x] Update dio for licence handling
      • dio licence <action>
      • [x] add https://github.com/justinclift/dio/commit/32e953c7688b06e35f6fe9d8c093f74f9c3563ab
      • [x] get https://github.com/justinclift/dio/commit/9c97d1de480484decd6a6a0a52d207d3095e93e9
      • [x] list https://github.com/justinclift/dio/commit/78dc7f7e4dc1afbcd16f2744703880bf626f414c
      • [x] remove https://github.com/justinclift/dio/commit/3bac2bec3c837d1c182df2cfeef36be9f069df43
      • [x] update (ie the friendly name, source URL) https://github.com/justinclift/dio/commit/8bfe2b7793e9814958c10a36a600f6077e6c712e
      • [x] Add --licence option for upload https://github.com/justinclift/dio/commit/9ebb72b5cc3c010fc74e7bcbb065d788482d8f52
      • [x] Show licence info when listing databases https://github.com/justinclift/dio/commit/4d71fb0906643e434d4c09bd23d0130a02fff91d
      • [x] Display licence info in commit history https://github.com/justinclift/dio/commit/73f2907937de684193e85e43ae75f5e721128660 https://github.com/justinclift/dio/commit/c046e29567b39a892697ab9cc7ab773690ebe43e
      • [x] When uploading a new version of a database that already has a licence, if a licence name isn't given on the cli then use the existing licence https://github.com/justinclift/dio/commit/1f9e5d0542fa11a62ca533374ddbcefecfa27f14
      • [x] Update licence for existing database https://github.com/justinclift/dio/commit/1d77124c0b02bf0cb93bc0b41e926832a7fb10fd
      • Importantly, this will always generate a new sha256, as the licence info is part of the hashed data. This should help defend against people claiming a licence has/hasn't changed in regards to any particular data.
      • [-] Should we support multiple licences for a database? eg similar to how DB4S is both GPLv3 and MPL licensed
      • For now, lets not. We can make the licence entry an array if it turns out to be needed
      • [x] When downloading a database with a licence, automatically download the licence as "[dbname]-[licence friendly name]" https://github.com/justinclift/dio/commit/44e14b60abca23ae3e7b8c1182be785828863b9c
      • We should probably make sure "license" and "licence" are both acceptable for all commands too. At least initially before even thinking about translation bits. :wink:
      • Meh, this is more a task for later.
    • [x] Ask subject matter experts if we should include optional support for (say) GPG encryption of commits.
    • This might be a good approach for guaranteeing authenticity of data, over using sha256sum's which just does data integrity.
    • While likely a good idea for authenticity, it's probably not something we need to implement right away. We can leave this for later.
  • [ ] Integrate with DBHub.io database schema - *in progress*
    • [x] Update table structure 2f060bcdf95e2d75f11fd013eb4b4da16715e1ec
    • [ ] Initial changes to server and dio to prove out the new structure - *in progress*
    • [x] user creation
    • [ ] licence
      • [ ] add
      • [ ] get
      • [ ] list
      • [ ] remove
      • [ ] update
    • [ ] db - *in progress*
      • [ ] download
      • [ ] forks
      • [x] list databases
      • [ ] remove
      • [ ] settings page
      • [x] upload
      • [x] view
    • [ ] branch
      • [ ] create
      • [ ] list branches for a database
      • [ ] change branch description and allow description removal
      • [ ] remove
      • [ ] history (log)
      • [ ] revert to prior database version
      • [ ] set default branch
      • [ ] get default branch
    • [ ] tags
      • [ ] create simple
      • [ ] create annotated
      • [ ] list
      • [ ] remove

[glawrence]

Just a thought but why don't you blog with a GitHub repo? I think there are a couple of options but at work we use Jekyll and GitHub. The author writes an item in Markdown and when it is merged into the repo it is published. Simple and with version control!

[justinclift]

Ahhh yeah, that's a decent idea. :smile:

Hugo and Jekyll are both static site generators, fairly optimised for committing into a repo (git or whatever). I've used Jekyll before, and it's not terrible.

My interest in Hugo mainly comes from it doing pretty much the same thing, but being written in Go (which I'm getting into) instead of Ruby (which I really don't like).

Jekyll does have the advantage of GitHub automatically rendering the content from the files in the repo. So any update to the repo automatically goes live. Hugo needs the website to be generated from the files manually (pretty simple) then committed to the repo before they're visible.

I'm not wedded to Hugo :wink:, but would prefer a non-Ruby solution if feasible. I guess I'll need to muck around with things more and see how it goes. I have a feeling this blog task will be pushed out a week or two as the new data model one is taking a while.

[justinclift]

Stuff that still needs doing has been moved to #51.