Just checking @ritchie46, will this still work?

import polars as pl 
from datetime import datetime 

pl.DataFrame({
    'myDate': [datetime(2020,1,1), datetime(2021,1,1), datetime(2022,1,1), datetime(2023,1,1)]
}).filter(pl.col('myDate')>'2021-06-01')

>>>
shape: (2, 1)
┌─────────────────────┐
│ myDate              │
│ ---                 │
│ datetime[μs]        │
╞═════════════════════╡
│ 2022-01-01 00:00:00 │
│ 2023-01-01 00:00:00 │
└─────────────────────┘

Just checking @ritchie46, will this still work?

import polars as pl 
from datetime import datetime 

pl.DataFrame({
    'myDate': [datetime(2020,1,1), datetime(2021,1,1), datetime(2022,1,1), datetime(2023,1,1)]
}).filter(pl.col('myDate')>'2021-06-01')

>>>
shape: (2, 1)
┌─────────────────────┐
│ myDate              │
│ ---                 │
│ datetime[μs]        │
╞═════════════════════╡
│ 2022-01-01 00:00:00 │
│ 2023-01-01 00:00:00 │
└─────────────────────┘

That never worked as expected. It will cast the date to string and do a string compare. You don't want that l.

Those codebases would have been broken to begin with. Polars already printed a warning that this likely was not what you should do.

Polars is not pandas and we choose to not support this as string as dates are ambigious.

I can live with that (though I'll be sad).

If I can argue my case, it is not just Pandas - it is the convention in SQL as well:

SELECT * FROM table WHERE myDateCol  >= '2022-01-01'

This is also the case in PySpark and Snowpark among others.

Moreover, it is not clear to me that this:

df.filter(pl.col('myDate') > '2021-06-01')

is any worse than

df.filter(pl.col('myDate') > datetime(2021, 6, 1))

My preference would be that when comparing a pl.Date/pl.DateTime object with a UTF-8 object, the string be cast to a datetime implicitly, and to only raise an error if that fails.

Conversions of strings to dates have so many edge cases and gotcha's that this conversion should be explicit. Then you can specify the format, the timezone handling, etc.

This explicitness will prevent bugs downstream.