MoarVM VS Inline-Perl5

Implement return prioritization #1786: superseeded

Now, as to how that's done at the low level, and whether the way it's done is a good way to do it, well, I can't help. Suffice to say, the main place to look for how this is handled at the low-level is MoarVM. I dug around for a couple minutes and have a link to a commit from nearly a decade ago. Does that help?

For a class the default representation is P6Opaque. This is one of a few dozen stock representations that Raku requires compiler backends implement as standard. See, for example, the 46 .c/.h pairs of C89 source files in the relevant MoarVM directory. A quick glance at the names of the source code files should paint a broad picture. A look at their code will fill in some details.

I'm hoping someone who knows C89 can take a quick gander at the C89 code implementing 47 representations in here and comment on it.

This is (yet another) very un(der)documented Raku feature (over a decade after it was introduced!), but one can browse what looks to me like reasonably clean and commented C89 code implementing 47 representations in the relevant MoarVM directory.

FROM alpine:latest as base ARG RAKU_RELEASE=2021.12 ENV PKGS="git make gcc musl-dev perl linux-headers bash" RUN apk update && apk upgrade \ && apk add --no-cache $PKGS \ && git clone --depth 1 --branch ${RAKU_RELEASE} https://github.com/MoarVM/MoarVM.git \ && cd MoarVM \ && perl Configure.pl --prefix /usr \ && make --print-data-base \ && make install\ && cd .. \ && git clone --depth 1 --branch ${RAKU_RELEASE} git://github.com/Raku/nqp.git \ && cd nqp \ && perl Configure.pl --backends=moar --prefix /usr \ && make install \ && cd .. \ && git clone --depth 1 --branch ${RAKU_RELEASE} https://github.com/rakudo/rakudo.git \ && cd rakudo \ && perl Configure.pl --backends=moar --prefix /usr \ && make install \ && ls /usr/share/nqp/ FROM alpine:latest ARG UID=1000 LABEL version="0.5.0" maintainer="[email protected]" raku_release=${RAKU_RELEASE} raku_user_uid=${UID} COPY --from=base /usr/lib/libmoar.so /usr/lib COPY --from=base /usr/share/nqp/ /usr/share/nqp COPY --from=base /usr/share/perl6/ /usr/share/perl6 COPY --from=base /usr/bin/moar /usr/bin/nqp /usr/bin/raku /usr/bin/perl6 /usr/bin/rakudo /usr/bin/ RUN mkdir /github \ && addgroup -S raku && adduser -S raku -G raku --uid ${UID} USER raku WORKDIR /home/raku ENTRYPOINT ["raku"]

Visit the MoarVM project's home page and/or its github repo.

MoarVM's representation is very cool https://github.com/MoarVM/MoarVM/blob/master/docs/strings.asciidoc

Verbatim from the documentation at https://github.com/niner/Inline-Perl5: use Inline::Perl5;

> Perl 6 was treated as the successor of Perl 5 -- and that was the mistake. It meant Perl 5 started dying,

Perl 6 took a long time to make, but how much did that matter? What was Perl going to do about Rails, Clojure, Go, Rust, JS/TS, and more? The world of programming languages used to be a lot smaller than it is today.

> Perl 6 had a new different syntax.

Inline::Perl5 [3] allows running legacy Perl 5 code in Perl 6 codebases.

[1]: https://docs.raku.org/language/5to6-nutshell#Regular_express...

[2]: https://github.com/atweiden/voidvault

[3]: https://github.com/niner/Inline-Perl5

Perl can be faster and defiantly nicer to work with than PHP but it's nowhere near as fast as some other options like Rust but that has the downside of being much harder to write and get working. Perl has the advantage of being prepackaged with the base system and on OpenBSD it's kept up fairly well with the system version currently being only one version behind the latest yearly Perl release. There are some patches added to the OpenBSD version so updating it each year takes time. The easiest way that I've found for running Perl websites is to leave the system modules as is and to install all of the modules that you need into a users home directory with local::lib and cpanm. That way the system install isn't contaminated with the extra modules and it makes deploying easy because it's all contained in one users home directory that can be rsync'd or tar'd. Another nice thing is you get access to pledge and unveil through Perl so you can lock down your website even further. I've recently been trying out doing things in Raku (Perl6 was renamed to Raku in 2019). It's slower than Perl most of the time but I find Raku to be so much more beautiful and expressive that it's a joy to code with. There aren't as many modules available right now for Raku but it gives you access to all of CPAN through the Inline::Perl5 module.

Interfacing with Perl 5 with Inline::Perl5:

I'll be happy to dialog about this compatibility, but the key thing is to start with an open mind; a recognition that the original vision predated Parrot; and a recognition that Rakoons have never relinquished that original vision even while many of us are delighted that Perl folk are keeping Perl healthy as a separate thing in its own right.

To be clear, the realization of Larry's original vision is not constrained to interop with Perl. Raku has extraordinary potential, some already realized, for pan-language compatibility via its [Inlines](https://modules.raku.org/search/?q=inline).

The steady improvement of [the exemplar Inline](https://github.com/niner/Inline-Perl5) for the last 6 years; its evident maturing; its suitability as a solid blueprint for upgrading all the others; all of this bodes well for being able to use a lot of the world's best existing code from within Raku this decade, with Inline::Python quite plausibly shining in this regard within the next 2-3 years.

[1] https://stackoverflow.com/a/45181464/1077672

The most polished of the Inlines, namely Inline::Perl5 (github repo) enables devs to instantiate Raku objects that are instances of Raku classes that are sub-classes of Perl classes. It does this even though Perl has pluggable OO, essentially arbitrary OO, with dozens of different OO systems available, all of which differ from Raku's, for example having a variety of MRO linearizations that aren't C3. So such things can be done.