develooper Front page | perl.perl5.porters | Postings from March 2000

Re: RE: A plea for more pollution

Thread Previous | Thread Next
M.J.T. Guy
March 17, 2000 09:30
Re: RE: A plea for more pollution
Message ID:
Horsley Tom <> wrote
> Now, if I'm understanding all this, we are going to explicitly look
> through the source for colliding names and automatically cause those
> modules to be POLLUTed?
> What was the point of all this again? :-).

Good point.    I was proposing to look for both "PL_na" and "na" etc.
If a PL... symbol was found, we can assume the module has been updated
so no POLLUTE is needed.    I'd only force POLLUTE if there were
polluting symbols present and *no* PL... ones.

So the case that would get bitten is where a module makes *no* use
of Perl's external symbols, and uses symbols which are hit by pollution.
So in particular, the module could never be built before 5.005 .
So in practice it won't be one of the "big name" ones.

Would it all be a better risk if the search was restricted to symbols
which are (a) relatively common (b) although nominally polluting,
extremely unlikely to clash in practice?   I have in mind
sv_undef sv_yes sv_no.    And of course perl_destruct_level and perldb
would be good risks, tho' unlikely to get any hits.

I'll sit and think about this a bit more before rushing in.

An interesting exercise would be to identify those modules which suffered
from pollution and caused this whole exercise to be initiated.    Then
make sure the algorithm does the right thing for them.    Does anyone
remember which they were?

Mike Guy

Thread Previous | Thread Next Perl Programming lists via nntp and http.
Comments to Ask Bjørn Hansen at | Group listing | About