[fpc-pascal] Why is Random(255) some 529x slower compared to Delphi 7?

Dimitrios Chr. Ioannidis d.ioannidis at artesoft.gr
Thu Dec 8 11:09:51 CET 2011


Hi,

On 8/12/2011 9:48 πμ, Graeme Geldenhuys wrote:
> [like what was told to me numerous times before] They (FPC users) 
> should speak up now, or forever hold your peace. And those that have 
> spoken so far, all seem to be fine with a less statistically strong 
> default Random(), and have the statistically strong one available in 
> the Maths unit. 

IMO, there is two separate subjects in this discussion. The first 
obviously is the implementation and the Delphi compatibility of the 
Random function but for me the hidden and most valuable subject is the 
fpc's lack of documentation. If the algorithm used to implement the 
Random function was documented then it wouldn't be any reason for this 
discussion . The user will be warned, informed and acted accordingly to 
suit his needs.

And plz don't use the UTS ( Use The Source) on me. There is situations 
that you don't have the time and the state of mind to read and study all 
the fpc sources.

In the end IMO don't change the Random implementation just documented it.

regards,
-- 
Dimitrios Chr. Ioannidis




More information about the fpc-pascal mailing list