One interesting possible side-effect of a post like this is that it points out the extent to which Perl is or can be used for more than something extremely simple.
I suspect there may be lots of programmers who've used Perl, or seen it used by coworkers, for throwing together quick scripts, but don't think of it as suited for larger projects, or as having any particularly interesting or advanced linguistic features.
Chromatic lists features that are built into Rakudo Star, but he also lists how they're available as add-ons to Perl 5. That may bring positive attention not just to Perl 6, but to Perl 5 as well.
And there are lot of places where Perl5 is being used in big applications or processes. The City of London and The Genome Project spring to mind but there are many more that discreetly just carrying on doing their job without too much fuss: http://news.ycombinator.com/item?id=1537848
I agree with the post completely - it's correct, detailed, etc. Yet, I think it completely misses the point of the original comparison. Most of the time when I use Perl, it's because I need something extremely simple - something like sending some specifically generated bytes over TCP, or running perl -e 'print "a"x40' in the console. I don't need "Parser manipulation", "Advanced object system", "autoboxing", "Coroutines", "Parse tree manipulation", etc.
So why are they loaded at all? Is Perl6 supposed to be only for "proper applications" and medium-sized systems with their own DSLs? Are the famous "Perl one-liners" dead as far as Perl6 is concerned? I agree with the comments on LWN that said - yeah - you can optimise it, but to bring it down to perl5 level, you'd need to start dropping features loaded "by default".
Or maybe that's not the point? Maybe Perl6 is supposed to be a new language for writing applications? Noone complains that Java is starting too slowly to write scripts in it (or they're insane), then again previous versions of Java weren't primarily used for scripting (so there are no expectations like that).
If Perl 5 suffices for your applications, by all means use Perl 5. Perl 5 will continue to be a useful and usable language for a long time.
David Skoll's original post missed the point that plenty of people use these features in Perl 5 I added in my comparison. Adding them to the language as intrinsic features helps everyone who uses them, and everyone who might want to use them, and everyone who currently can't use them because their shared hosting providers won't install them or their system administrators haven't vetted them or they're incompatible with existing projects they've had to write themselves in their own organizations.
In short, Perl 6 exists in large part to improve on things Perl 5 did not do well enough. One of those things is programming in the large.
We'd also like Perl 6 to be usable for one liners and small, short-lived applications, and we're working on mechanisms to avoid paying the cost for features you don't. Yet Rakudo Star has always been "a usable, useful subset of Perl 6" and never "a drop in replacement for everything you use Perl 5 for right now." Bugfixes, new features, modifications, and optimizations will come.
I suspect there may be lots of programmers who've used Perl, or seen it used by coworkers, for throwing together quick scripts, but don't think of it as suited for larger projects, or as having any particularly interesting or advanced linguistic features.
Chromatic lists features that are built into Rakudo Star, but he also lists how they're available as add-ons to Perl 5. That may bring positive attention not just to Perl 6, but to Perl 5 as well.