Arthur Vickers
Arthur Vickers
The only valid type for strings if we have no additional data would be "nvarchar(max)". It's worse for precession and scale. There is no combination of precision and scale that...
Some investigation: The underlying problem is not dissimilar to the situation with any parameter that doesn't "fit" in the column it is being used against. For example, consider the following,...
@roji Using `nvarchar(max)` is significantly slower. (Number 100, 1000, etc. is the number of items in the parameter array. Always a million rows in the database.) | Method | Mean...
But then we would have to also delete the other entities sharing the table, and re-create those, which would have different side effects.
We can discuss.
See also the issue described here: #33653
@davidroth Thanks for the suggestion, This is an interesting idea, but not something we are planning to implement for 2.0. We may work on it in the future, but that...
@ErikEJ @wraith2 @NickCraver @mgravell @Drawaes: Let's get this party started! @smitpatel, @roji, and I discussed how to get started with Woodstar, and we have created some initial issues for the...
Bottom line is that we need to measure this. If ADO.NET is adding significant overhead, then we either address that or don't ultimately use ADO.NET. Both of these options are...
@Wraith2 I should point out again that the goal of this project is **not** to replace Microsoft.Data.SqlClient. Rather it is to compliment it with a fast driver that most likely...