Sometimes I worry about myself, when passages like this one from A conversation with Alan Kay will invariably push my buttons:
I think a lot of the success of various programming languages is expeditious gap-filling. Perl is another example of filling a tiny, short-term need, and then being a real problem in the longer term.
I shouldn’t be this emotionally attached.
Looking at the wider context of the conversation, he is talking about the historical view, and looking at earlier iterations of Perl, I have to admit that it was just not a very good language back in the time. Only since Perl5 can it be considered a solid language, and a lot of the problems we have with a massively crufty legacy Perl code bodies are caused either directly or indirectly by the Perl4 heritage.
Later, he goes to differentiate between agglutinative vs style languages. Stuart Feldman, who is interviewing him, says in summary that
I would characterize style languages as those with a very rigorous kernel that describes them intellectually.
This is the culmination of a stretch in the discussion that touches, among others, on how Lisp and Smalltalk are defined extensively in terms of themselves.
Most certainly, in its sixth iteration, Perl is coming of age.
What I worry about is that if anything, Perl seems to accumulate more arcane syntax as it grows, however much symmetry goes along with it. For my sensitivities, examples of Perl6 code too frequently look like a bad pun on APL. I hope this is just curmudgeonly resistance to change, with no basis in real trouble.
On his closing point, Alan shows that he and Larry most likely do not disagree at all:
I could go on and on. I feel like my answers are quite trivial since nobody really knows how to design a good language, including me.
Now that sounds familiar…