RE: [boost] Re: runtime unit-modifiablephysical-quantitiesusingpqs1-00-01

From: Paul A. Bristow (boost_at_[hidden])
Date: 2004-01-08 06:22:30

I've returned from holiday to read all the considerable discussion on unit &
dimensions.

I hope it is useful to reiterate that these have been recognized as two
interlinked problems for a decade now and there have been numerous attempts at a
solution.  I am sure that many of them were partially sucessful and may well
have proved very useful to their authors and others.  Work-in-progress is freely
available on the Boost files and CVS Sandbox areas and elsewhere, and may be
invaluable for many people, whose success or failure is very valuable feedback.

I would be very surprised if any solution to these issues will receive instant
approval in a Boost review because:

1  There has already been so much disagreement on the objectives, application
areas and boundaries.

2  There must surely be a reasonable body of user experience for this sort of
item to be suitable for the Boost library.

3  There are bound to be compromises between compile speed and what can be
checked. This is very clear comparing the two current proposals.

4  Since a decade has already passed, Boosters are unlikely to rush into any
solution.

5  Acceptance by the very various (and often picky!) users of the 'syntax' is
vital, and yet difficult to judge. The physical world group is large and vocal,
but there are many others.  Few real users are represented in the discussions so
far.

6  Boost libraries often prefer the general and abstract over the specific and
instantly useful.  This is justified as 'taking a long term view'. It may be
that some ideas are ahead of compilers and/or hardware, but may be worth waiting
for.

So I feel it would be useful for authors to improve documentation and for users
to see how the proposals work out in practice for them, and to report back.

Paul

Paul A Bristow, Prizet Farmhouse, Kendal, Cumbria, LA8 8AB  UK
[phone]   Mobile [phone]
mailto:pbristow_at_[hidden]

| -----Original Message-----
| From: boost-bounces_at_[hidden]
| [mailto:boost-bounces_at_[hidden]]On Behalf Of Deane Yang
| Sent: Saturday, January 03, 2004 10:08 PM
| To: boost_at_[hidden]
| Subject: [boost] Re: runtime
| unit-modifiablephysical-quantitiesusingpqs1-00-01
|
| But but but....I think the point I am trying to make (unsuccessfully) is
| that a properly designed units library should not actually define any
| units itself. The units should be created only by the user of the
| library. Then there should be libraries derived from the units library,
| like the physical dimensions library.
|
| I see nothing wrong with focusing on a particular use case when
| designing the library, but if you want a truly useful C++ library,
| you want to design its guts so that they can be used by the library
| user to define whatever units the user happens to need. I think it
| is an error for the library writer to try to predict what units
| the library user needs.

I agree that this is a highly desirable abstraction.