I see where you're coming from, but I fixed this because someone was
annoyed that it used to work and considered it a regression, and the
problem wasn't in any "real" code but in a constraint being
unnecessarily strict. Pointers to structs fully support type compile
time interface of an output range w/o any changes (besides in template
constraints) to any code that assumes it's getting a struct by value.
On 9/15/2010 6:55 PM, Andrei Alexandrescu wrote:
I'm getting a tad worried that we embrace everything that could
possibly work under the ranges juggernaut. Each decision in part is
justified (delegates, sealed ranges, opApply, now pointers) but at the
end of the day we should clarify a charter. For example until
yesterday I had no idea and no intuition that pointers to output
ranges should be supported. What's next? Pointers to input ranges? etc.
I'm not saying we need to eliminate something from this diff or the
current definition. But a discussion about the ranges' charter would
be in order.
Andrei
On 9/15/10 17:41 CDT, dsource.org wrote:
phobos commit, revision 2008
user: dsimcha
msg:
Support pointers to structs in std.range.put().
http://www.dsource.org/projects/phobos/changeset/2008
_______________________________________________
phobos mailing list
[email protected]
http://lists.puremagic.com/mailman/listinfo/phobos
_______________________________________________
phobos mailing list
[email protected]
http://lists.puremagic.com/mailman/listinfo/phobos
_______________________________________________
phobos mailing list
[email protected]
http://lists.puremagic.com/mailman/listinfo/phobos
