[Info-vax] Roadmap

[Arne Vajhøj]

On 1/8/2019 3:20 AM, Dave Froble wrote:
>  From the latest Basic help:
> 
> DATA_TYPES
> 
>    REAL
> 
>       Floating-point values are stored using a signed exponent  and  a 
> binary
>       fraction.   BASIC  allows  six  floating-point formats:  single, 
> double,
>       gfloat, sfloat, tfloat, and xfloat.  These  formats  correspond 
> to  the
>       SINGLE, DOUBLE, GFLOAT, SFLOAT, TFLOAT, and XFLOAT keywords.
> 
>       Keyword             Range                                  Precision
> 
>       SINGLE (32-bit)    .29 * 10^-38 to 1.7 * 10^38              6 digits
>       DOUBLE (64-bit)    .29 * 10^-38 to 1.7 * 10^38             16 digits
>       GFLOAT (64-bit)    .56 * 10^-308 to .90 * 10^308           15 digits
>       SFLOAT (32-bit)   1.18 * 10^-38 to 3.40 * 10^38             6 digits
>       TFLOAT (64-bit)   2.23 * 10^-308 to 1.80 * 10^308          15 digits
>       XFLOAT (128-bit)  6.48 * 10^-4966 to 1.19 * 10^4932        33 digits
> 
>       In declarative statements, the  REAL  keyword  specifies 
> floating-point
> 
> First, I've never used SINGLE, don't see any point.
> 
> My understanding is that on VAX D_FLOAT is implemented in HW, at least 
> on some models.
> 
> And G_FLOAT on Alpha is implemented on HW, at least for some  models.
> 
> What is the best format to use on VAX, Alpha, itanic, and soon x86?
> 
> John, I seem to recall that you indicated T_FLOAT would be the best to 
> use on itanic.
> 
> Is X_FLOAT slower?  Basic is stuck on 32 bit addresses, but I didn't 
> think that had anything to do with the size of data types.
> 
> Me thinks that it's time to do a bit of testing.  Determine if other 
> than D_FLOAT can be used in programs for better performance.

I would go for T.

It should be true HW on all Alpha, Itanium and x86-86 *and* be
binary compatible with most other platforms.

It is not supported on VAX, but do you really need to support VAX??

Arne