SQL The problem of long tables

**What's wrong?**
Unexpected shrinking of a table when loading from a database. Longer than > 1 000 000

I know that there is a limit on the maximum table length = 1 000 000. But I think we can be a little friendlier.
We have different behavior depending on the database.

### MS SQL

If table length > 1 000 000 then nothing happens.
I think we can display an error message if the data could not be loaded.

### Postge SQL

If the length of the table is > 1 000 000, then it greatly decreases when loading.
For example, when loading a sequence of numbers from 0 to 1,000,100. We get a table of length 10.400 with a minimum value of 39.220 and a maximum value of 979.019. This is a very strange result.

I propose to discuss solutions:

1. Display an error message
2. [Slicing ranges](https://pandas.pydata.org/docs/user_guide/indexing.html#slicing-ranges) Similar to pandas based on size

**How can we reproduce the problem?**
[jupyter notebook](https://github.com/Mikhaylov-yv/Open_source/blob/main/df_to_db.ipynb)

**What's your environment?**

- Operating system: win11
- Orange version: '3.35.0.dev0+3cc58b4'
- How you installed Orange: from source code

Start with the linked df_to_db.ipynb and reproduce the loading behavior against MS SQL and PostgreSQL using a table longer than 1,000,000 rows. Trace where the database result is limited or reduced; done should mean the data is not silently truncated and the chosen behavior—an error message or documented range slicing—is covered.

Written by the indexing model from the issue text.