develooper Front page | perl.perl6.users | Postings from September 2022

Re: BEGIN {} question

Thread Previous | Thread Next
ToddAndMargo via perl6-users
September 1, 2022 21:09
Re: BEGIN {} question
Message ID:
On 9/1/22 00:45, Richard Hainsworth wrote:
> Raku and Perl are two different languages in the same family. They 
> evolved with different targets, perl to react quickly to internet 
> requests, Raku to be a better programming language. This may not be the 
> take the actual developers have, but it's what I think happened.
> So the thing you designed for Perl will be faster because it fits the 
> pattern. But it's not the best way for Raku. Even though stage parse 
> will get significantly faster over time because Raku was designed to be 
> optimisable, I think Perl will always be faster.

Hi Richard,

I do not think I have been clean enough
in my intentions.

I have two goals: long and short term.

My short term goal, which I have expounded on at
length, is to assist me with a program for my
business where I have to constantly keep up with
it and have to do a lot recompiling.  .precomp
is not helpful in that instance.

My long term goal is future programs (as I get
better at this stuff and if I EVER have a chance
to learn GLADE).

What I am after is NOT a race between Perl 5 and
Perl 6 (Raku) or any other programming language.
I really do not care if Perl 5 compiles 300 times
faster.  I care what the customer's "perception"
of what the start time is.

Ten microseconds or two seconds, the users
can't tell them apart.  Ten or twenty seconds and
users think something is wrong and start
several more instances of the program. This
is the "professionally embarrassing" part I
have written about.  I constantly have to tell
users (of non Perl 6) programs to just wait
or you will continue to crash things.

A widows work around is to have them start slow
to start programs by right clicking on the icon
and left clicking on "Open".  That way they know
they did not flub the double click.

Think of it this way. An NVMe drive is about eight
times faster that a mechanical drive.  Do the user's
programs run any faster once loaded from either
drive?  Unless they are doing substantial (YUGE)
continuing writes and reads from their drive, no
they do not run any faster.

But!!!!!  They load substantially faster, giving
the user the impression of a fast, snappy program.
This colors the rest of the user's experience.
And they don't start five instances of the program
thinking they flubbed the double click.

So I DO NOT CARE if Perl 6 is slower to compiler
AS LONG as the user/customer does not think
something is wrong.  One or two seconds is
just fine with me.  I am not after microseconds.


Thread Previous | Thread Next Perl Programming lists via nntp and http.
Comments to Ask Bjørn Hansen at | Group listing | About