RE: [boost] associative_vector: request for formal review

From: Beman Dawes (beman_at_[hidden])
Date: 2000-05-30 20:45:31

At 12:08 PM 5/30/00 -0700, Gary Powell wrote:

>Looks neat.
>
>I request that the use of 
>#include <boost/...>
>be changed to
>#include "boost/..."
>
>I was recently informed that <> was reserved for standard libraries
only.
>Since boost isn't quite there yet, lets use ""

We discussed this when boost was first getting started.  The uses are
very similar.  For <...> the standard says:

>>>searches a sequence of implementation-defined places for a header
identified uniquely by the specified sequence between the < and >
delimiters, and causes the replacement of that directive by the
entire contents of the header. How the places are specified or the
header identified is implementation-defined.<<<

For "..." the standard says:

>>>causes the replacement of that directive by the entire contents of
the source file identified by the specified sequence between the "
delimiters. The named source file is searched for in an
implementation-defined manner. If this search is not supported, or if
the search fails, the directive is reprocessed as if it read

# include <h-char-sequence> new-line

with the identical contained sequence (including > characters, if
any) from the original directive.<<<

Notice the distinction between "header" and "source file".  So people
who say to use "..." for user supplied files are correct, at least in
a pedantic sense, since "header" refers to standard library headers,
not user source files.

But the distinction is so fine that very, very, few people use <...>
for standard library headers and "..." for other included files.  At
boost we have been following the usual practice of using <...> for
user files.  Maybe we should revisit that decision, but I doubt we
will have any impact on the majority of programmers who write <...>
for all inclusions.

--Beman