3.0 release todo

[ssweber]

• [ ] #198
• [x] #300
• [x] feature-full example for readme
• [x] #313
• [x] #338
• [x] add Table justification to TableHeadings
• [x] Bugfixes

Would your Panda.index changes be all internal?

Any other thoughts about wrapping this huge release up? I’m pretty amazed, we are a full 700+ commits ahead of master 🤩

[PySimpleSQL]

I haven't had a chance to mess with the index stuff at all yet, we are short staffed at work so I've basically been locked in for 16 hours a day. I finally get some guaranteed days off starting next Thursday. The .index changes are all pretty much internal and not user-facing, so it may just have to wait until I get some time. No reason to hold up the release over it

[ssweber]

What I’m working on: I have SQLite/MySQL/Postgres/Sqlserver working almost perfectly with a “multiple_databases” version of orders. Plan to add msaccess to it as well.

I’ve got fixed the docker issues, sql_commands/script, and other smaller issues.

One issue I’ve almost fixed is Boolean handling for sg.Table checkboxes. We infer types, but Mysql/Sqlserver don’t have a Boolean domain (unfortunately mysql has one, but it’s only an alias for tinyint). So I’m coming up with some options that devs can declare a “type_alias” either via column comments, eg “@Boolean”, or after the DataSet has been created in python.

Additionally, our handling of domain is fragile, we are missing some that are available on our current sqldrivers, and there’s no way for a user to easily add domain types. So I’m going to proceed to do a lite-weight version of sqlalchemys approach, and convert domains to “type” in Column info, with classes: String, Integer, Decimal, Boolean, DataTime, and Blob.