Tom Scogland

That seems reasonable, I normally use `auto` so I suppose I didn't think about it. I'm beginning to think the entire resource setup might need some re-thinking as we do...

It should be const at least, probably also noexcept on the concrete resources but it would likely have to be conditional noexcept on the type-erased version. On 6 May 2021,...

Ok, it's time to get some eyes on this, the current version appears to work portably and improves the camp test compile time for ICC from 40 seconds to 11...

Not yet, RAJA will need some work to run with this cleanly, it shouldn’t be much but I’ll do it shortly. Sent from my iPhone On Nov 19, 2019, at...

🤦‍♂️ ugh, sorry, bonehead mistake on my part. It's working in here as is (or was, hopefully still is) the main rework is to deal with two issues, an EDG...

One concrete example I'm reminded of for the second issue is cases like this: ```c++ template using ht_int = DefineConcept(has_type(val() + val())); //desugars to template constexpr std::true_type valid_expr(T &&...); template...

That is absolutely fascinating, thank you so much for sending the test code @corbett5, finding good tests for this stuff is really tricky. It probably wont, but just in case...

Well, @corbett5, I have both good and bad news for you. The bad news is that you're completely right that the detection for that pattern is broken, and that even...

Thanks @MrBurmark!

@rchen20 what's the status of this? Should we take this, close it, rework it..?