[forwarded submission from a non-member address -- rjk]
From: Nathan Torkington <[EMAIL PROTECTED]> Date: Mon, 22 Oct 2001 12:31:50 -0600 Subject: Re: Parrot and .Net and Java and migration from Perl 5 To: Boston Perl Mongers <[EMAIL PROTECTED]> Steve Tolkin writes: > Q1. Will there be a reasonable subset of Perl 6 > that can run *safely* in .Net? I might be willing, > even happy, to live with restrictions such as: > must type all variable, must not use eval string, etc. > > Q2. Can Perl 6 run unsafe in .Net? > > Q3. Can Perl 6 emit Java bytes codes, e.g. *.class files? Perl 6 is the language. Larry's still working on it, and until he's done and we write a compiler for it, it doesn't exist. You can see the work in progress by reading the Apocalypses and Exegeses at http://www.perl.com/ As the language isn't finished, and implementation hasn't begun, there doesn't seem to be much point in my making up answers to those questions :-) We fully expect that people will turn Perl 6 into .NET managed code and Java bytecodes, but without a complete (or even near-complete) language spec those people haven't had a chance to write that code. > Q4. At previous talk Dan explained that Parrot used a register > model rather than a stack model, to permit generating more > efficient machine code. (Most hardware uses registers, > not stack.) However both .Net and Java use a stack model. > I suspect this is a cased where the famous principle > of "worse is better" applies. > http://www.jwz.org/doc/worse-is-better.html > Briefly: It is better to be compatible than to be technically better. This wasn't a question :-) If you were asking "why did you do this?" or "won't this break your compatibility?" then Dan can answer those. Nat
