Re: [boost] Re: physical-quantities (was: runtime unit-modifiable ...)

From: Matthias Schabel (boost_at_[hidden])
Date: 2004-01-12 22:15:07

> Well, the end-users will rarely define new unit tags, and even if,
> it's once per program. I think it's a small price to pay for better
> compile times. I think it should be seriously considered.

I certainly am aware that compile time is an important issue.   
Unfortunately,
some sides seem to feel that 7 SI dimensions are all they'll ever need  
and
others have a laundry list of various quantities (I'm closer to being  
in that
camp, myself), so I'm not convinced that the need to define new units  
will
be so infrequent.

> Another idea in the same direction is to use sha1 digest of a  
> sufficiently
> unique string as the sorting tag. 160 bits is just 5 unsigned longs.

It sounds cool, but that sort of stuff is completely outside the domain  
of my
expertise.  Since the unit code is independent of the implementation of  
the
dimensional analysis and the dimensional analysis implementation is  
largely
independent of the underlying implementation of the list (other than it  
being
a MPL container), I'm sure something like this is possible - probably  
not out
of me, though...

------------------------------------------------------------------------ 
---------------------------
Matthias Schabel, Ph.D.                           	
Utah Center for Advanced Imaging Research
729 Arapeen Drive
Salt Lake City, UT  84108
801-587-9413 (work)
801-585-3592 (fax)
801-706-5760 (cell)
801-484-0811 (home)
mschabel at ucair med utah edu