http://qs1969.pair.com?node_id=11151337


in reply to Re^2: Did ChatGPT do a good job?
in thread Did ChatGPT do a good job?

G'day Bod,

I use module-starter. I use the Module::Starter::PBP plugin; I like the templating features and have written my own custom ones; I'm less interested in the "Perl Best Practices" features and have removed a fair bit of that.

With this, I can create skeleton modules including POD, a number of standard test files, and other associated files. Just like you, I need to add the program code and POD details. My results are consistent every time, which you won't get with ChatGPT. Furthermore, it does a lot more work than ChatGPT and, I'm reasonably certain, with less effort and in a shorter time. Consider the following which only took about a minute.

ken@titan ~/tmp/pm_11151331_module_starter $ module-starter --module=Nod::To::Bod Added to MANIFEST: Changes Added to MANIFEST: lib/Nod/To/Bod.pm ... multiple similar lines for other files created ... Created starter directories and files ken@titan ~/tmp/pm_11151331_module_starter $ cd Nod-To-Bod/ ken@titan ~/tmp/pm_11151331_module_starter/Nod-To-Bod $ perldoc lib/Nod/To/Bod.pm ... displays a couple of screenfuls with TODOs where details need to b +e added ... ken@titan ~/tmp/pm_11151331_module_starter/Nod-To-Bod $ perl Makefile.PL; make; make test Checking if your kit is complete... Looks good Generating a Unix-style Makefile Writing Makefile for Nod::To::Bod Writing MYMETA.yml and MYMETA.json cp lib/Nod/To/Bod.pm blib/lib/Nod/To/Bod.pm Manifying 1 pod document PERL_DL_NONLAZY=1 "/home/ken/perl5/perlbrew/perls/perl-5.36.0/bin/perl +.exe" "-MExtUtils::Command::MM" "-MTest::Harness" "-e" "undef *Test:: +Harness::Switches; test_harness(0, 'blib/lib', 'blib/arch')" t/*.t t/00-load.t ............. 1/1 # Testing Nod::To::Bod 0.001 t/00-load.t ............. ok t/99-00_pod.t ........... ok t/99-01_pod_coverage.t .. ok t/99-02_manifest.t ...... ok All tests successful. Files=4, Tests=5, 1 wallclock secs ( 0.01 usr 0.03 sys + 0.29 cusr + 0.50 csys = 0.84 CPU) Result: PASS

And, of course, I have the added benefit that none of this code ever hallucinates. :-)

— Ken