On Tue, Oct 23, 2007 at 12:40:17PM -0700, chromatic wrote:
> ... except don't break CONST_STRING across lines, because apparently icc
> can't
> count and won't compile.
>
> One might assume that being able to count is less of a burden than handling
> pointer math correctly, but I never took a
On Tuesday 23 October 2007 12:32:21 Allison Randal wrote:
> You will need to bundle up the 'sub' argument in a hash (the low-level
> equivalent of named arguments like 'initial_sub'). So, using your
> existing init_pmc, it'll be something like:
>
>void init_pmc(PMC* sub) {
> PMC *classobj
chromatic wrote:
now, the creation of Parrot::Coroutine becomes (r21929) :
.local pmc coro
.const .Sub coro_sub = "enumerate_tree"
coro_class = get_class 'Parrot::Coroutine'
coro = coro_class.'new'('initial_sub' => coro_sub)
So, how translate it in PMC ?
I think it's some
On Tuesday 23 October 2007 09:40:29 François PERRAD wrote:
> I've the following problem in languages/lua/pmc/luathread.pmc :
>
> when the creation of Parrot::Coroutine
> (runtime/parrot/library/Parrot/Coroutine.pir) was :
>
> .local pmc coro
> .const .Sub coro_sub = "enumerate_tree"
>
I've the following problem in languages/lua/pmc/luathread.pmc :
when the creation of Parrot::Coroutine
(runtime/parrot/library/Parrot/Coroutine.pir) was :
.local pmc coro
.const .Sub coro_sub = "enumerate_tree"
coro = new 'Parrot::Coroutine', coro_sub
in luathread.pmc, I wrote (t
