Perl 7 Announced
Perl 7, really Perl 5.32, has been announced.
https://www.perl.com/article/announcing-perl-7/
It's not the hippest language, but it's pretty fun. It's also much better then writing shell scripts.
Perl 7
- How excited are you?19 votes
- Yes!5.26%
- No26.32%
- $731.58%
- Amitz36.84%
Tagged:
Comments
Soon we'll need PERL SELEKTER
"A single swap file or partition may be up to 128 MB in size. [...] [I]f you need 256 MB of swap, you can create two 128-MB swap partitions." (M. Welsh & L. Kaufman, Running Linux, 2e, 1996, p. 49)
Anybody up on the early history of perl? I never saw a discussion by Larry Wall or by anybody else with inside knowledge of the early days about what specific perceived deficiencies in sh originally led to the development of perl. Or maybe the origin was different, not caused by sh deficiencies, possibly something more like there was another language which Wall liked, so he build perl to be like that other language. Thanks!
MetalVPS
It was something to do with report processing. That's all I know.
As for shell scripting, shells don't handle complex data structures well (it's clunky), and life gets easier by moving to a full programming language. Shell is fine for automating a series of commands which don't need to share data.
Perl is fairly ubiquitous, it works a lot like shell scripts, and it's a really good system administration tool due to it's historical usage as such. Other languages have better syntax, but they're trying to solve different problems.
That's my experience anyway.
Exactly! When I talk with perl people, it's always this. Plus they often say that perl is Turing complete whereas sh isn't. But somehow, to me, all this, while true, doesn't seem quite to satisfy the historical perspective that I seek.
MetalVPS
whoa whoa ;-) shots fired, lol
That's definitely incorrect, anything that can do conditional jump one way or another is considered Turing complete. And also "Turing complete" has never been strictly defined.
http://stedolan.net/research/mov.pdf
Apologies! I never meant to fire shots at anybody! I meant to report what perl folks have said to me over the years.
Thanks @comi for the cite to the Dolan paper, which says that the mov instruction from the x86 instruction set plus a single unconditional branch at the end of the program is enough to simulate a Turing machine. So I am corrected about sh not being Turing complete. And I now appreciate that something as simple as one instruction can be Turing complete.
But whether either or both of perl and sh, is Turing complete, or not, incompletely
addresses my historical question: how and for what reasons did perl originally come into being? I've been wondering for years, asked people, searched a bit, and come up empty. If anyone knows or can point to the answer, I again would be grateful. Thanks!
MetalVPS
Although it may not address your precise question, the page "The Timeline of Perl and its Culture" is immensely informative: http://history.perl.org/PerlTimeline.html
Perhaps closer to addressing your question, here are slides from a talk by Larry Wall (the creator of Perl) in 1994 entitled "The Taming of the Camel": https://www.shlomifish.org/lecture/Perl/Newbies/vhll-slides.pdf
Finally, here's an interesting article from Salon.com in 1998 entitled "The joy of Perl": https://www.salon.com/control/1998/10/13/feature_269/
"A single swap file or partition may be up to 128 MB in size. [...] [I]f you need 256 MB of swap, you can create two 128-MB swap partitions." (M. Welsh & L. Kaufman, Running Linux, 2e, 1996, p. 49)
Main issue I have with Perl, is that I'm unable to decode my own code, and if it works, it often does something quite different than what I wanted it to do ... (I'm no great programmer, and Perl allows so many different styles, quotes etc. Haven't used for many years.) As I understood it, Perl 6 developed in a direction that not all liked, became a separate project and then changed name to Raku in 2019. So Perl 7 would be the continuation of Perl 5 ...
@flips @angstrom Thanks guys for your informative comments. I will check the links that @angstrom posted.
MetalVPS
As I understand it, since Perl 6 was to break backward compatibility with Perl 5, there was no practical choice but to make it (effectively) a different product. Perl 7 will preserve backward compatibility with Perl 5, at least to a very large extent.
"A single swap file or partition may be up to 128 MB in size. [...] [I]f you need 256 MB of swap, you can create two 128-MB swap partitions." (M. Welsh & L. Kaufman, Running Linux, 2e, 1996, p. 49)
Oh man, I did a fair bit of Perl hacking, back in the day; this brings back memories. When Perl 5 came out, it was like the sky was falling -- what was a scripting language doing with OO?
There was a time when I really bought into the "literate programming" idea and tried hard to make my code read like English ... all a big waste of time. The fact there are a zillion ways to do anything hurt readability rather than helped.