Re: [boost] array formal review

From: Ed Brey (brey_at_[hidden])
Date: 2000-06-26 09:50:22

From: <wb_at_[hidden]>
>May I respecfully add the following points of information to this (very
>important!) naming thread:
>
>1)  The name "c_array" is used in Stroustrup's book (_The C++
>   Programming Language, Special Edition_, pp. 496-7) for a container
>       class very similar to this one.
>
>2)  The name "block" is used in Austern's book (_Generic Programming
>    and the STL_, pp. 60-2) for yet another similar container class.
>    Austern credits Andrew Koenig for suggesting this name.
>
>If we wish to suggest, by the choice of name, that the Boost class is
>similar in purpose to those found in the above, then I would recommend
>we limit ourselves to the above two names as candidates.
>
>On the other hand, if we wish to suggest that Boost's class differs
>quite significantly from the container classes found in these two
>(well-respected) published sources, then we ought eliminate both their
>names from candidacy.

As much as I respect the sources of these name choices, I question
whether the proposed names are engrained enough to be worth such
alignment/conflict concerns.  In a way, the authors are throwing out
names that they seem to think are appropriate just as we are.  It's
quite possible that we are paying more attention to the names and
their harmonization with other future names than the authors.  I don't
have the Austern book, but at least in Stroustrup, there is no
rationalle for the name choice (although this may be due to space
constraints).

Since the precident of other names in this case is relatively light,
IHMO, if boost can document a solid rationalle for the name choice,
that should stand on its own.