post MIxS 6 merge -> master -> main notes

[turbomam]

Some of this could be reiteration of existing issues. I'll try to link issues as needed.

• I checked in .idea, which contains some minimal configuration for PyCharm or other JetBrains IDEs. Some would argue that this is unnecessary or even undesirable
• I don't think I've contributed anything to any changelogs
• we might need new, improved documentation of the repo contents. For example "what's .idea? what's doc_templates?"
• doc_templates contains jinja templates for the LinkML auto-generated documentation. @sujaypatil96 and I need to work on the contents and aesthetics some more. We should look for opportunities to integrate with other MIxS or GSC (static?) documentation, like the GH pages maintained by @only1chunts
• I haven't done anything to push the LinkML auto generated documentation to this repo's GH pages
• Should we get rid of the downloads fodler? That's where the MIxS Google Sheets were downloaded for processing by mixs_converter.py
• Should we remove or somehow bury mixs_converter.py etc.? (Since the LinkML that mixs_converter.py generated is now the source of truth, not the Google Sheets?)
  • if so, we should probably remove no-longer-relevant portions of the Makefile. I think I have at least commented them out.
• Let's make sure the distinction between generated and release is clear (to us and our collaborators!)
  • I think I have a new PR for Makefile rules that move artifacts from (https://github.com/GenomicsStandardsConsortium/mixs/tree/main/generated) to release. It also attempts to set prepare the repo for poetry build; poetry publish but I haven't gotten that to work yet. See #378

CC @ramonawalls

[turbomam]

Actually, before we nuke mixs_converter.py, we should get several people's eyes on the logs. I have been git-ignoring the logs... maybe they should get pushed up to GH?

[turbomam]

Representative entries from sheet2linkml.log:

• ERROR:root:Multiple slots with URI
  • from owl generation, because of combinatorial classes?
• ERROR:root:No PK for quantity_value
  • from sql ddl generation. probably a reasonable design that quantity values like 15000 meters aren't getting their own identifiers
  • I think there was some discussion of silencing that kind of error in future versions of LinkML

---

• INFO:root:Default_range not specified. Default set to 'string'
• INFO:root:Skipping excel as it is in exclusion list
• INFO:root:Skipping java as it is in exclusion list
• INFO:root:Skipping markdown as it is in exclusion list

[turbomam]

confirm we're using the right (and consistent) licenses

[ramonawalls]

@turbomam have these questions been resolved? The only one I know is still relevant is

• Let's make sure the distinction between generated and release is clear (to us and our collaborators!)

and I think we are addressing it in #692.