I dunno ... software tends to live forever in whatever language(s) it was started in.   Perl(5) continues to do what it was designed to do for those systems, and this alone I think will keep it in the main stream for decades to come.   Like COBOL, it will forever continue to be the implementation language of revenue-producing legacy code.   For that reason alone it will continue to be important.   Perl(5) still is a Swiss Army® Knife for an awful lot of things.   Well-implemented, and able to leverage a truly vast library of tested contributed code, this tool still moves the freight that pays the bills.

But, probably, not for new systems.   (And I will say the same for PHP and, at least within web circles, Python, all IMHO.)

Just as ADD 1 TO COBOL GIVING COBOL. was more-or-less always an academic curiosity, I don’t think that Perl(5) will ever have, nor that it ever needed, a “successor.”   It would have been far better to call the successor to Perl(5) Ruby Rakudo, or something, anything else which did not imply that it actually was an upward-compatible successor to anything.   (The fact that the team was fractured over wildly divergent implementation choices and wanted to run on JVM didn’t help matters, either.)   The choice of a Perl6 name was an attempt to “bootstrap” upon the success of the Perl5 language and its enormous contributed library.   A clear message never developed, and maybe that clear message never was there at all.   It offered no implementation advantages to mitigate the cost and the business risk of materially changing “a Perl5 code-base that is right now successfully running the business.”   But, you could never say such a thing to any of the language’s now-shrinking number of pundits at that time.   They just wouldn’t listen.   Maybe they still won’t now.   Will the last pundit left standing please turn off the lights.

Edit:   Perhaps this is a little bit too harsh.   After all, new(!) programming languages do appear all the time, and most of them are indeed serious attempts to refine and to build upon what has come before.   In this particular case, though, I think that this entirely-different set of implementors (and, Larry Wall himself, as the registered owner of the Perl® trademark) erred by trying unsuccessfully to link this new(!) language to their (his ...) previous wildly-successful creation.   It is not properly a “Version 6” of anything, and I fear that it has suffered grievously ... and perhaps, unnecessarily ... from the pretense.

Many other programming languages have tried hard to follow in Perl’s footsteps, and the most-notable of these that comes to my mind is Ruby.   Ruby copied many key concepts from Perl, and yet it never pretended to be “Perl++.”   From day one, it established itself as a clearly-separate brand with a clearly-distinct ancestry, and it always seemed to know what it wanted to be when it grew up.   By comparison, Perl6 produced only brand-name confusion, and it seems to IMHO that the market responded by ignoring it.   Which just might be a damned shame.