Thanks for your effort in bringing the Omnisci database to DBSeeder.
Can I ask you, why are you using text encoding none for varchar datatype? This is going to prevent a lot of operations like group by, distinct and so on in the schema.
Unfortunately, I’m not an expert on OmniSciDB - my main strengths are more with MySQL, Oracle or PostgreSQL. I support the following 5 data types at the moment:
|DBSeeder Type||OmniSciDB Type|
|BLOB||TEXT ENCODING NONE|
|CLOB||TEXT ENCODING NONE|
|VARCHAR||TEXT ENCODING NONE|
I’m happy to change this in the next release for OmniSciDB - what would be your recommendations?
Yeah, I’m quite an exper of Oracle (but now on my SQL).
The differences between ENCODING NONE and ENCODING DICT, is that with dictionary encoding you can use the field in a group by or distinct, chile with text encoding none, you can use just for filtering data in a where clause.
Botyh fields can store around 32k (like a varchar2 in oracle), and then only limitation you have with encoding is the number of distinct values that is only 2 Billion.
I have to check the goal of DBSeeder, and then I can make some reccomendation