Lars Gullik Bjønnes wrote:

larsbj@gullik.net (Lars Gullik Bjønnes) writes:

| Both in boost/archive/detail/oserializer.hpp(l.117) and | boost/archive/detail/iserializer.hpp(l.124). | | Because of some commented code 'flag' in unused and gives a warning. | | Would be nice if this was fixed before 1.34 proper.

Here is the patch:

I would be OK with this, Robert? Thomas -- Thomas Witt witt@acm.org

"Robert Ramey" <ramey@rrsd.com> writes: | If someone want's to apply such a patch I have no objection. | | Actually, different compilers given warnings for this and that and the other | thing. Can you rephrase please? | I don't think its a good idea to start down this road. What road? The road of producing warning free code for more compilers? -- Lgb

Thomas Witt said: (by the date of Thu, 12 Apr 2007 11:47:41 +0100)

Robert Ramey wrote:

...

If someone wants to apply such a patch I have no objection.

Actually, different compilers given warnings for this and that and the other thing. I don't think its a good idea to start down this road. I agree that this is low risk. But they're all low risk. Better is to get 1.34 out without even one minute more of delay.

Take my word for it there is time for this change. That being said it's your call.

This "unused parameter warning" is much more frequent in boost::serialization, perhaps Robert has cleand it all in CVS-HEAD. But applying proposed patch to 1_34 would mean that a dozens of other lines should be fixed too. That's what I suppose Robert wants to avoid. OTOH I'd really appreciate if boost::serialization would not produce warnings during compilations of correct code. Some people (like me) have a warning-free compilation policy, and this library interferes with that... I raised this problem before, but it was not resolved: Should boost users expect a correctly written "hello world" to compile without warnings? -- Janek Kozicki |