Salvador Ortiz

@Xliff: No, Rakudo has it's own tests in https://github.com/rakudo/rakudo/tree/master/t/04-nativecall

There is already some code in DBDish's guts in that direction: At the DBDish (generic) level see StatementHandle's column-type attribute, that is initialized at prepare time with default values (the...

I have a few comments: If in Perl6 a Type is a first class citizen, why reduce it to its Str name? In particular the fallback in your getter just...

@kaare, in perl6 a "builtin" Type is indistinguishable from a non-builtin one. And I still think that the setter/getter approach isn't really needed. Something like: ```perl6 role TypeConverter does Associative...

As the associated PR addresses the originally stated issue, I'll merge @kaare's current work and then push a few minor cosmetic changes on top of it (like remove the now...

I like the idea. In fact I now use a hardcoded `typed-nulls => False` in "t/40-SQLite-common.t" I think that a simple `method capabilites` at the driver level can be enough....

Well, a concrete use case is in `Test::DBDish`. Now we have three drivers that returns typed NULLs and SQLite that can't. The test suite now adapts the related tests to...

With podman in F33 the recipe fails with: ``` … / # curl -sSL -o moarvm.tar.gz https://github.com/MoarVM/MoarVM/releases/download/$MOARVM_VERSION/MoarVM-${MOARVM_VERSION}.tar.gz /bin/sh: curl: not found ``` The build succeeds adding `apk add curl git`

@nxadm, > Same result. Success or failure? With your new recipe my build succeeds.

The problem may be that the *composed* `=` operators are all synthetized with the same precedence that item assignment operator `=`, while the mutator method call special operator `.=` has...