I've just been asked to "upgrade" a Perl project, last updated in 2012, to PHP 8

I'm not a #Perl expert

Or a #PHP expert

Current mood:

@cypnk That's a weird description of the task to rewrite a whole software in a different language.

Well, you can probably actually upgrade the app to latest Perl 5.42. #Perl has very strong longterm stability and the app might just run fine. Then you can do small incremental improvements from there.

Maybe better have a modern Perl app than no functional app at all?

@dboehmer That's the common sense approach

I would suggest that they keep the project as Perl, but they're very strongly motivated to switch languages. I still think it's inappropriate to make a huge change like this since there are just too many edge cases

I think my week is going to be full writing a whole new thing in PHP instead of trying to convert this

@cypnk @dboehmer
Using the old system as de facto requirements is the sane pivot, if it's small enough (which it sounds like, from very little hints). Finding all the hidden features may be easy or hard.

[Otherwise, strangler fig pattern as s/o else mentioned, retrofitting layers of tests, reverse engineering would be the path, but that's not one person week.]

@BRicker @dboehmer It's luckily not a large enterprise application or anything like that. Just an evolved "Small Series of Shell Scripts" (but in Perl) that are now doing much more over decades and integrated into their workflow

That's still a huge number of things they're doing, but small sections at a time can be removed without taking everything down

This will definitely be a slow slog

@cypnk
Best wishes for you.

If you wind up needing Perl5 deciphering clues, and don't have budget for senior advice by the hour (OTOH ping me if you do), you can ask under hashtag here, Perl FB groups (one is polite), @Boston_PM mailing list.