I’ve been working in the Perl ecosystem professionally for almost 18 years at time of writing this. One thing I keep coming back to is how fragmented the ecosystem is.
Let’s consider writing a web application. First you must decided on a framework:
- Catalyst - Full MVC framework, some strong opinions, but tries to be flexible first.
- Mojolicious - Real time framework with support for websockets.
Then you get to decide how to interact with your database:
- DBI - Bare bones, connect to the database and
send SQL, receive back hashes and arrays.
- This, at least, is pretty standard.
Now lets look at options for validating form data before storage:
You may be thinking, “but choice is good, right?”. Choice is good. What’s bad is silos. When you’re working on a Validation library without considering how it will interact with an ORM. DBIx::Class::Validation attempts to integrate with the ORM but, thanks to segregating missing and invalid fields, makes it onerous to deal with errors in web forms.
When looking at other options in the web space, I find the most compelling to be those with a solid core of modules. Look at C# with ASP.NET MVC and EF Core, or Python with Django, or Ruby with Rails. Each of these frameworks carry with it a solid set of included functionality. You may stray beyond that initial set, but for 90% of projects, the basics just work. Additionally any solution which wishes to compete for that 90% must be as good or better than the default.
I do not mean to “bash” on Perl. Perl is a wonderful language with a cornicopia of talented developers giving their time and effort to the ecosystem. But could you recommend Perl to a new programmer who wants to make web apps? I’d have a hard time suggesting a stack to work with, given the lack of a standard set of known-good tools.
Disagree? Did I miss something? I’d love to hear from you. Hit the contact button on my site.