Isn't one of the main cool things about Perl6 the Parrot "master" virtual machine? The VM that's supposed to slice, dice and be a platform for every language from Perl to Prolog?

I hope that I'm wrong, but I have a feeling that the Perl6 people are trying to do too much at once. How many years will it take to delouse a complex language AND multi-language VM and produce something ready for safe production use?

Parrot is already shaping up great and shaping up rapidly. The VM will be mature long before the language it is destined to carry. This is the main advantage of doing Perl6 and Parrot separately. It's likely that several other languages will have been ported to and stabilized on Parrot by the time the first serious effort at implementing Perl6 for real even takes place.

Makeshifts last the longest.

But at least they would have to be backwards compatible, so one wouldn't have to change the way one has been programming in for decades.

Abigail

How? Fixing OO is just one example that would require breaking old habits anyway, whether it happens under the guise of Perl5 or is deferred to a wholly new design.

Makeshifts last the longest.

Your not of the opinion that the P5 sources have just reached the point where further enhancement involves so much hacking, conditional compilation, and intertwining of the new with the old that it has effectively been rendered almost impossible to maintain never mind extend?

In the three and a half years that the Perl6 project is on its way, perl5 has produced 5.6.1, 5.6.2, 5.8.0, 5.8.1, 5.8.2, and even 5.9.0. In the mean time, perl6 hasn't even decided how their operators will look like.

Granted, modifying the perl5 sources is hard, but that's caused a large part by having to be backwards compatible. If, just like perl6, that condition could be lifted, it would be easier to develop 5.10 or 5.12.

Perl5 is evolving. Perl6 isn't there.

Abigail