microsoft / GSL

Based on the cpp core guidelines I expect to be able to write code like this
int length(zstring p);
but with this gsl implementation the code must look like this (at least on my compiler)
int length(zstring<> p);

I infer from the unit tests that this problem is not specific to me or my compiler. See for example:

GSL/tests/string_span_tests.cpp

Line 988 in 8a4b9ed

czstring<> str = name.assume_z();

Not sure if this is a problem with the core guidelines or the gsl implementation but based on #770 it seems like the type alias shouldn't be templated.

zstring<> is only marginally worse than zstring but the time wasted figuring out why the example code wouldn't compile in practice is my primary concern.

I've reached out the C++ Core Guidelines editors for comment on whether or not we should redefine zstring to remove the empty angle brackets.

Echoing this question from the Guidelines issue: Do we have data about whether the extra parameter is used in the field?

I ran a quick search on Github for zstring and czstring to see how users are using *zstring and found that the majority of uses just have empty angle brackets. I came across a couple uses where the user does specify an extent, here is one such case:
https://github.com/luncliff/coroutine/search?q=zstring. Most of the other hits that I found were clones or copies of this coroutine library.

whoops, didn't mean to close.

This issue has now been resolved in PR #998

zstring family require empty template parameter list for use