versions.txt

0.1 - Got Django 2.1 admin running on hgwdev and proof of concept that
      the multiple selects on many-to-many fields work there.  Made
      basic Contributor, Lab, Project, Grant tables and a few of the
      biologically relevant tables in first draft.  Showed to Clay.
0.2 - Went through Clay's wish list and got through about 3/4 of the
      stuff we have control over.  There's lots of tables for things
      like organs, assays, and project state.  The controlled vocabs
      are managed by a separate table for the vocab in question that
      the wranglers can extend.  People can put comments on most things.
0.3 - Figured out how to set up file uploads so that people can download
      a file after it's uploaded.  Used this for holding the excel
      spreadsheets and other key files for wranglers.  Got through
      the other 1/4 of Clay's wish list.  Polished up the project
      details page a bit to condense it again with all the recent
      additions.  Changing 'vocab' to 'wrangler' as a prefix for the
      tables most people should ignore, and figured out how to do 
      this without changing the table names.
0.4 - Adding a bunch of date fields in project, and a specific comment
      field for the submission.  Refining state names and adding more so can 
      show project as it goes through analysis a bit.  Breaking now huge
      wrangling section into three:  wrangling, submission steps, post-submission.
          tech note:
      Django migration gets easily confused around multi-to-multi tables 
      that are reflected so as to appear on both tables being related.  
      I've had to do pretty painful patches two or three times now.  :-(   
      I'll probably figure out the pattern though.  
0.5 - Changes done largely in response to feedback from the wranglers and Max.
      These include
         - Making the comments simple text fields rather than many-to-many relations 
	   with a hcat_comment table.
	 - Making a wrangler table, to add more info without burdening contributor.
	   This should also make it simpler to filter by wrangler. 
	 - Made some fields in the lab optional.
	 - You can filter projects by wrangler1.
	 - Added chat url to overall section
	 - Added a new submission step - assigned to curator
	 - The contributor role is now a key into a vocab table.  There's roles defined
	   for program officer (Jonah), pipeliner (Tim), analyst (Jing Zhu and Max),
	   lab contact (not enough), PI (too many), intern (just enough) with parenthetical
	   remarks addressed to our test database.
	 - The submission steps section is changed to be date first, and a few have been
	   renamed, notably curation.
      In anticipation of working with interns I made an Intern object similar to the 
      Wrangler one.  I also added SoftwareDeveloper while I was in there.
0.6 - Changes largely done to integrate in existing datasets and info from Parth's 
      project tracker, a nice tool at https://tracker.data.humancellatlas.org,
      that tracks bundles and matrix cell counts.   Table is hcat_tracker. Visible changes include
         - 22 new projects (we only had 2 of the already submitted ones before)
	 - New "Tracking after submission" object in the admin.  This has most of
	   the nonredundant fields from Parth's tracker that weren't already in the
	   "Projects" part.  The list display shows mostly bundle counts and if it
	   has a matrix computed an actual number of cells!
	 - The "post-submit" section of the project details page is gone, and replaced
	   with a section now at the bottom called "post-submission tracking - bundles and cells"
	   which is just an inline non-editable view of the tracker info.
	 - Note that the import of both project descriptions and validated spreadsheets
	   was done manually and is incomplete.  I think I'd like wrangler help to
	   do about 15 of these from when the ingest ui site seems to have lost track of
	   the spreadsheets.
0.7 - Put in a json API.  
           <tuataraUrl>/api  - Gives you list of api entry points, 
           <tuataraUrl>/api/project  - Gives you list of projects
           <tuataraUrl>/api/contributor  - Gives you list of contributors
           <tuataraUrl>/api/organ  - Gives you list of organ
           <tuataraUrl>/api/disease  - Gives you list of diseases
           <tuataraUrl>/api/tracker  - Gives you list of project post-submission tracking
      Where I'm hoping tuataraUrl will be https://tuatara.data.humancellatlas.org,  but
      at the moment it's http://hgwdev.gi.ucsc.edu:8123/
	   Also Changes suggested by Will and Chris after putting in some of their
      projects in particular:
          - there's a "shared_google_sheet" url that they may use sometimes in place of
	    the spreadsheet to/from lab fields.
	  - the staging area is text instead of URL since S3 didn't work in URL
	  - "chat url" is now called "git ticket url."  If we can figure out a way to
	    start a git ticket when a project is started it'd save them some clicks.
      Chris didn't like seeing "wrangler diseases" and the like, since the diseases really
      weren't about the wranglers, so am prefixing the less public/more preliminary tables
      with "x" instead of "wrangler"
0.8 - Changes made primarily from wrangler feedback.  In general this consisted of simplifying
      things by removing tables or fields people didn't really want,  renaming things where the
      meaning wasn't clear, not sharing contributor contact info beyond name.  This version does
      not address any wrangler specific views but it should contain most of the database changes.
      - Removed responder fields from project since wranglers felt they weren't needed as 
        separate from the contacts fields.
      - Removed assay type fields and table.
      - Renamed state_reached to status and cur_state to wrangling_status.
      - Added WranglingStatus model and field as distinct from overall ProjectStatus.
      - Moving wrangler sections to bottom of the project page.
      - Changed "wrangler1" to "primary wrangler" and "wrangler2" to "secondary wrangler"
      - Made API only return name, projects, and labs from contributors since other info especially 
        contributors might want wranglers to have but not the web at large.
      - Made pubmed id and doi fields optional for publications
      - It's not visible in code but in database I adjusted some tables to view only for wranglers
        at their request including species.
      - Buried Grant and Wrangler behind an x in the admin list since these are also seldom used.
      - Renaming AssayTech to CdnaLibraryPrep
      - Making title field have more space
      - Rearranging top part of Project details page in admin.
      - See issues 1-5 for even more detail.
0.9 - Added a landing page that explains Tuatara and includes links to the admin interface
      and a decent description of the API.  Also:
      - Fixed a bug that made the project search box crash
      - Minor configuration adjustments to make it fit into https://tuatara.data.humancellatlas.org
        ES-2 unit.
      - Removed lab affiliation from api/contributors since not really needed