[m-rev.] for review: delete RTTI predicates from std_util

[Zoltan Somogyi]

On 16-Mar-2006, Ian MacLarty <maclarty at cs.mu.OZ.AU> wrote:
> I still don't like
> the idea of breaking user code - even in a "minor" way.

We are trying to minimize the breakage, but we don't want to get into
trap Microsoft got into with MS-DOS, being too afraid to make changes
that improve things but also break backward compatibility. Our user base
is also much more sophisticated. If the NEWS file says where the functionality
has gone (which I am working on now), they will be ok.

> It could
> cause quite a bit of frustration, especially if the user doesn't know
> where to look for the right version.  Obsoleting them first is just
> good manners in my opinion.

The documentation had comments about them being obsolete already, e.g:

	% Type_info and type_ctor_info are the old names for type_desc and
	% type_ctor_desc. They should not be used by new software.

Pretty much all the rest of the deleted predicates took arguments of type
type_desc.type_desc instead of std_util.type_info, and the stability
of std_util.m wasn't high, so it is not like there wasn't any warning.
The replacement modules (type_desc, construct, and deconstruct) were
created four years ago. Users' adjustment period is over.

Zoltan.
--------------------------------------------------------------------------
mercury-reviews mailing list
post:  mercury-reviews at cs.mu.oz.au
administrative address: owner-mercury-reviews at cs.mu.oz.au
unsubscribe: Address: mercury-reviews-request at cs.mu.oz.au Message: unsubscribe
subscribe:   Address: mercury-reviews-request at cs.mu.oz.au Message: subscribe
--------------------------------------------------------------------------