The evolution of Moose then Moo is a good example.
It's a good thing that people can scratch an itch, throw something up on CPAN, and it gradually become popular (or not). In contrast to languages with very rigid control over their evolution (Go, Java, etc)
Coming back to my point...
Perl 6, I think, breaks dramatically from this key strength. It may too evolve in such a way, but with such a gap between the too languages, it seems like a very sensible idea to rebrand perl 6 to something. I have no stake at all in it, but merely offer the advise as to ensure its an "available" name in terms of other projects, trademarks, copyrights, prior art, domain names, seo etc.
]]>]]>Over time, it became clear that though Perl 6 was in the same family as Perl 5, a straightforward migration path was unlikely.
If things remain as they are, nothing will change. The downward decline will continue.
If perl6 is renamed, and perl5 "officially" has its own future. Things might change.
Of course there are many other factors, but the choice seems binary and the "do nothing" option seems to have a certain outcome.
]]>I think this would send several messages to the outside world:
Thoughts ?
]]>As for the new name I don't particularly like Raku. For a Brazilian Portuguese speaker it sounds harsh. But I think it's mostly about having time to get used to it.
Some time ago I thought that "Gerl" would be a nice name for Perl 6:
On the other hand, the new name should probably be very different from Perl in order to avoid perpetuating the confusion. Raku fits this bill very well.
So be it.
]]>I mean, the big user-visible improvement in Perl 5.30 was "The upper limit 'n' specifiable in a regular expression quantifier of the form '{m,n}' has been doubled to 65534", which was pretty underwhelming. And so far, the most notable change in the dev releases is that "vec" is not allowed on code points above 0xFF. I'm sure these are good and valuable changes, and yet at the same time, it's not exactly what I'd want to see in a "What's new in Perl 7" document.
]]>As an example of another possible problem, suppose you have the following list of names: abc cbd dec cba baz. The only name with no successor is baz, but baz is not part of the longest sequence. The longest sequence is (abc cbd dec cba) or (cba abc cbd dec). Computing the list backward from baz will not lead you anywhere near the longest sequence. So, while your optimization may work well with the Pokémon list provided by Mohammad, I don't think it works in the general case (again, assuming I've understood your proposal correctly).
Best, Laurent.
]]>I try again:
Yeah, well, I understand what you mean and I mostly agree, Yet, but, strictly speaking, this is true only for x > 0. You have to add or subtract pi when x is negative, and atan2(y, x) is usually defined for x == 0 (at least for values of y other than 0), even though y/x would lead to an exception when x = 0. I chose to give a geometrical interpretation of atan2 (the rectangular coordinates of a point in the plan) in my update to avoid having to deal with all these edge-case difficulties in algebra.
]]>