Has anyone sucessfully walked accross the line of midsize (tens of thousands of lines of code) application into the domain of huge business to business corporation while keeping perl's flexibility intact?

I've seen some pretty big perl projects, but not at "tens of thousands of lines." However, I used to code in C extensively, and I am always aware that a perl program can do the same thing in 1/20th of the lines of source code that another language can. Or less. And do a better job of it. :)

Could perl be turned into a VM kind of model?

Probably not in the way that Java is, but Perl has it's own set of strengths. As far as I am concerned, it's practically a VM language now - I can write a perl script and run it on a VMS, DOS, Solaris, or Linux box, and have them all run the way I intended it, without worrying about internals like the size of a short or how the file system works.

Will PalmPerl ever exist?

As I said in an earlier post (too busy to dig now), it is being worked on, but has a big obstacle in that Palms (and other PDA OSs) have a very, very small pool of memory to work with, which we've all but taken for granted the last few years as modern computers have memory undreamt of 10 years ago. I think it is possible, but it will be a scaled-down version, and it will require a little more memory than most PDA's currently have.

Will perl byte code ever be viewed as an object that exports a clean well documented API for developers, or even automated systems to make calls against?

Interesting question. On the one hand, it could be argued that perl is already well documented, and you can make your code as "strict" (PI) as you want it to be. On the other hand, the "More Than One Way To Do It" is definitely a major strength, and I would hate to see it go the way of VB and have to use StrictlyDefinedReallyLongNamedFunctions that never do exactly what you want them to. :) What would you improve about the current state of Perl? (Meant as an honest question - the API point is interesting)