RowReader.float is misleading

[witoldsz]

I do not know who invented this, but look:
C# – Characteristics of the floating-point types

C# type/keyword	Approximate range	Precision	Size	.NET type
float	±1.5 x 10−45 to ±3.4 x 1038	~6-9 digits	4 bytes	System.Single
double	±5.0 × 10−324 to ±1.7 × 10308	~15-17 digits	8 bytes	System.Double

and now look at what F# has:

F# – Basic types

Type	.NET type	Description	Example
float, double	Double	A 64-bit floating point type.	1.0
float32, single	Single	A 32-bit floating point type.	1.0f

---

So:

• C# float maps to F# float32 (or single) and
• F# float maps to C# double. 🤯

This is crazy and can cause lots of troubles.

Now, I can see the RowReader follows this and row.float maps to F# float32 instead of float a.k.a double.

    /// Gets the value of the specified column as a `System.Single` object.
    member this.float(column: string) : float32 =
        this.fieldValue(column)

The problem is, the change could cause even more harm than good at this point, but I would recommend:

• make RowReader's float deprecated (with explanation that the name is ambiguous)
• use the single and maybe an alias float32 to describe 32-bit floating point conversion

With this change, we would avoid the contaminated float and also make the issue more expressive for those who are not aware of the trap. What would you say?