This is the mail archive of the libstdc++@gcc.gnu.org mailing list for the libstdc++ project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [RFC/Patch] libstdc++/14320 aka class type streamoff doesn't work :(


Paolo Carlini <pcarlini@suse.de> writes:

| Paolo Carlini wrote:
| 
| > Gabriel Dos Reis wrote:
| >
| >> | Keeping streamoff as a class type, basically amount to closing 14320
| >> | as "WONTFIX"
| >>
| >> That is not what that would imply.
| >>
| >> One of the testcase says that operator*= is not defined.  Another
| >> wanted to insert or extract. It occurs to me that thr right course of
| >> action is to provide those instead of removing the class.
| >
| > And what about the first testcase? And the fourth?
| 
| Also, besides the details of the various testcases, please consider
| that the name of the PR is "istreambuf_iterator::difference_type is
| not a signed integral type", therefore, clearly, we are talking about
| signed-integral-type-ness, not about changing a class to behave "as
| if" it were a C++ signed integral type. The testcases, are just
| exemplifications of the general point.

You can give a PR any title you want. If you want to rename it to 
"libstdc++-3 is not standard conforming", please do.   That would be
even clearer and more close to the truth.

| Again, I'm convinced that we cannot answer the spirit of the PR while
| keeping class streamoff.

I care about useful abstractions and real problems, I don't care about
the spirit for random PRs.

-- Gaby


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]