Learning a new langauge by using it for new (always urgent) projects doesn't often work.

You find yourself sitting there frustrated to hell trying to work out how to do something trivial, that you know you could do in your old langauge almost without thinking, and you usually revert because of the time pressure.

And if you try coding all those nice-to-have-but-you've-never-found-the-time-for projects that have been kicking around, you'll never finish them because there's no pressure to do so.

Make a list of the old language programs that get most use or require most maintenance and sort them by loc ascending in a scratch/private directory. Then try setting aside 30 minutes a day or whatever suits your schedule and go through them one-by-one in order and do your best job of converting them to Perl. Don't sweat the detail of testing and robustness on your first pass. None is ever going to use them for real. The idea is to force yourself to just work out a way to tackle those familar tasks in the new language. As you go through them, you'll find better ways of doing things you've already done earlier. Maybe go back and update those earlier attempts..

One day soon you'll need to write something, and the Perl solution will pop into your head before the old language solution does. At that point, if the new script is anything other than a throw-away, you'll probably still need to write it in the old language because of testing/robustness. But then is the time to set up a more formal attempt to do the same thing in Perl in parallel, and perhaps post your attempts here for critical review.