96656d98549f8f0bb4ad2931330f35110ee8b857 kent Sat Sep 14 14:37:02 2019 -0700 Documenting changes since last version. diff --git src/hca/hcat/mysite/versions.txt src/hca/hcat/mysite/versions.txt index a2d8a20..fbc501d 100644 --- src/hca/hcat/mysite/versions.txt +++ src/hca/hcat/mysite/versions.txt @@ -1,45 +1,59 @@ 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 - Adding uuid field to the project table. +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.