Skip to content


Subversion checkout URL

You can clone with
Download ZIP
Fetching contributors…
Cannot retrieve contributors at this time
236 lines (173 sloc) 8.22 KB
~ readme ~
~ p ooOOOo tion ~
oO %% a little
Oo fast language.
___/ /
/` \
/v^ ` ,
p2 is a she. She is a perl5 and possible perl6/nqp backend, based
on why the luck stiff's potion. See `README.potion` why potion is exciting.
parrot example/fib.pir 28 0m1.746s
perl example/ 28 0m0.439s
p2 -B example/fib.p2 28 0m0.177s
potion example/ 28 0m0.013s
p2 example/fib.p2 28 0m0.013s
parrot example/fib.pir 40 3m36.447s
perl example/ 40 2m19.752s
p2 -B example/fib.p2 40 0m54.560s
potion example/ 40 0m3.512s
p2 example/fib.p2 40 0m3.512s
For most small examples comparable performance to lua and C.
For bigger examples optimizable via static types/early binding.
Its exciting points for perl folks are:
* Radically simplified and functional core.
* Should parse most of perl5, just some XS and B tricks not.
* Signatures, fast mop, types, native threads, call-replacing
macros for perl5.
* No worries about pseudo-scientific discussions, project management,
performance bugs and feature deadlocks. Leaving the slack behind.
* A lightweight generational GC, based on Basile Starynkevitch's
work on Qish, with ~4ms per GC on average with < 100MB heaps.
The compiler will add hooks for objects leaving scope and add
DESTROY method calls if they are visible at
compile-time. Run-time created DESTROY methods will not be
automatically called.
* Using most of Damian Conway's p5i recommendations:
i.e. classes are immutable and final by default.
See <>
* libp2 is usable as a fast backend for nqp/perl6, but parsing and supporting
p6 natively without going through nqp and the rakudo bootstrapping process
should be easier and faster.
Tokuhiro Matsuno already wrote a p6 parser, pvip.
* Native and fast calling convention, no XS stack args on the heap.
XS is replaced by extern (dynamic ffi), and/or shared libraries
with normal function calls into the vm.
* Just-in-time compilation for x86, x86-64 and ppc done. arm later.
* Intermediate bytecode format and VM. Load and dump code. Decent
speed and cross-architecture. Heavily based on Lua's VM.
* int, str, num are objects, you can optionally type all args and
lexical variables. In fact everything is an object as in parrot,
even the parsed AST, the vm state.
* Sized arrays are non-autovivifying and initialized with undef,
resp. if typed 0, 0.0, "".
my $a[3]; print $a[3] => compile-time error: array out of bounds
my int $i[3]; print $a[2] => 0
my num $n[3]; print $n[2] => 0.0
my str $s[3]; print $a[2] => ""
* A new pragma "no magic" applies to all visible objects in scope.
no magic;
use Config;
print $Config{'ccflags'};
=> compile-time error: Invalid use of tie with no magic
no magic;
use Config ();
print $Config::Config{'ccflags'};
* p6model: via scoped mixins, :multi, roles. Like Moose, just better
and roughly estimated 800x faster.
No multiple inheritance yet, \@ISA == 1.
Only one parent, depending on another parent, rather multiple
mixin compositions.
* ~~ match works because all data are typed objects.
A destructuring-bind alike new match operator which matches
structures (list, trees, hashes of hashes) against structures,
values and types.
* Bootstrapped "id" object model, based on Ian Piumarta's soda
languages. This means everything in the language, including
object allocation, parser, compiler and interpreter state are
part of the object model. (See COPYING for citations.)
In fact all objects and classes are closures, which simplifies
scoping, and makes p2/potion a functional vm.
* parsers are scoped and run-time loadable for ffi, p6, and
other languages.
* Interpreter is thread-safe and reentrant. This should
facilitate parallel interpreters initialized in a threadpool.
* Small core. Under 10kloc. Install sloccount and run: `make sloc`.
* Reified AST and bytecode structures. This is very important to
me. By giving access to the parser and compiler, it allows people
to target other platforms, write compilers, debuggers, code analysis
tools and even fully bootstrapped VMs. I'm not as concerned about
the Potion VM being fully bootstrapped, especially as it is tied
into the JIT so closely.
* Memory-efficient classes. Stored like C structs. (Although the
method lookup table can be used like a hash for storing arbitrary
* The JIT is also used to speed up some other bottlenecks. For
example, instance variable and method lookup tables are compiled
into machine code.
* Fast asynchronous IO via libuv
However, some warnings:
* Ints cannot hold pointers and need to be boxed. This should be
transparent though. Use WeakRefs for pointers.
* Strings are immutable (like Lua) and byte arrays are used for I/O
buffers. All strings and symbols are utf8 internally.
* Limited platform support for coroutines.
This affects exceptions.
* No regular expression bindings yet (pcre or sregex)
* The parser is not GC safe yet. This affects eval.
Do not waste too much memory inside eval.
* No OS threads yet.
* No signal handling yet.
* The compiler is mostly good enough, as in lua, but there are no
tradional compiler optimizations yet. We'd might need to add
constant folding and propagation, simple local basic block
optimizations, peephole optimizations, loop optimizations,
tracking side-effects. And tailcalls are not yet detected.
* No type checks, bounds-check elimination and type optimizations
~ building and installing ~
$ make
$ make install
Look inside the file called `` for options.
~ feverish and fond thankyous ~
why the lucky stiff left the ruby community. potion was his last
public project. It's a work of great art and mastership. I could not
think of anything better. I think it is better than Go, if potion
would have had added auto-threads and channels.
Rob Pike for Go, and the parrot community for parrot.
Larry Wall for perl5. We miss his leadership and technical excellence.
Without him nothing got done, and if something was done it was wrong.
Besides defined-or, which left Tom Christiansen behind.
why the lucky stiff thanks:
why is gravely indebted to Basile Starynkevitch, who fielded
questions about his garbage collector. why favors French hackers
to an extreme (Xavier Leroy, Nicolas Cannasse, Guy Decoux,
Mathieu Bochard to name only a portion of those he admires) and
is very glad to represent their influence in Potion's garbage
Matz, for answering why's questions about conservative GC and
for encouraging him so much. Potion's stack scanning code and
some of the object model come from Ruby.
Steve Dekorte for the Io language, libgarbagecollector and
libcoroutine -- He referred frequently to all of them in
sorting out what he wanted.
Of course, Mauricio Fernandez for his inspiring programming
journal housed at and for works
derived throughout the course of it -- extprot most of all.
Many of my thoughts about language internals (object repr,
GC, etc.) are informed by him.
Ian Piumarta for peg/leg. We use a re-entrant customized version of
it, but the original library is sheer minimalist parsing amazement.
Final appreciations to Jonathan Wright and William Morgan
who pitched in, back in the wee hours of Potion's history.
~ license ~
See COPYING for legal information.
potion is MIT licensed, which lets you do anything you want with this.
p2 and perl6 code is ARTISTIC 2 licensed.
Jump to Line
Something went wrong with that request. Please try again.