Hacker News Reader

Comment by xref | Hacker News Reader

Why is it that Perl (apparently) didn’t survive the transition from 5->6, but after years of acrimony Python looks like it will survive the 2->3 transition?

View on HN

Perl6 was a reaction to Perl5's loss of mindshare, but it took a very long time to arrive and still cannot be used as a drop-in replacement for the latter. It has rather different strengths and weaknesses and thus has to find a new niche to live in.

> Perl6 was a reaction to Perl5's loss of mindshare

That's not the way I remember it. The Perl 6 effort was announced in 2000, and (at least by the ever-fickle TIOBE index), Perl marketshare kept growing until 2003 and did not start declining until 2005:

https://www.tiobe.com/tiobe-index/perl/

By that time, Amazon for example had already decided that Perl's time had passed.

Another case in point, look at the Google trend for "perl tutorial"[1]. Unfortunately, that only starts at 2004, but you can still see by that metric, Perl popularity already was in sharp decline when Tiobe still had it fluctuating around a peak.

[1] https://trends.google.com/trends/explore?date=all&geo=US&q=p...

It took a while to realize how great the distance would be between perl 5 and perl 6, and how far off into hyperspace the latter would go.

The GP spoke of mindshare. Marketshare shifts frequently lag shifts in mindshare by years.

Some Perl fora had started to become hostile by 1999. It had clearly begun to lose core talent and something needed to be done about that.

There was plenty of hostility in 1996, too. By 2000, certainly some people decided something needed to be done about the hostility, but I don't recall "core talent" leaving over it.

Maybe I'm wrong. My memory is too hazy on it partly because I was just an outsider looking in. I know some "important people" quit p5p, but could well believe they didn't quit the community. [1]

[1] https://www.perl.com/pub/2000/05/p5pdigest/THISWEEK-20000507...

Fairly early in the life of Perl6 it was basically decided there wasn't going to be a transition. Not like the Python one at least.

Also I've heard about projects that transitioned from Python2 right to Go. (The way it sounded, it was easier for them than transitioning to Python3.)

---

Perl6 isn't trying to replace any language. Why does everyone think this has to be a zero-sum game?

C++ didn't replace C. D and C# didn't replace C++ or C. Ruby didn't replace Perl. C didn't even replace Assembly.

There is a long list of languages that didn't replace other languages.

I mean even Cobol is still an actively used language.

---

Also I think it is dumb that Python didn't copy Perl and make all of those changes work in a backward compatible way. (I have yet to see a change in Python3 that isn't reminiscent of one I've seen in Perl5.)

Perl was already dying by the time 6 came around. Python was on the upswing when Python 3 came around.

2to3 works and forward/backward compatibility was handled very well.

It really wasn't, but I suppose it clears the low, low bar of perl 5/6

Data science libraries and adoption by teaching institutions.

Reply on news.ycombinator.com