Perl

Perl is a programming language. That means: a fixed notation with which people give instructions to a computer. The American Larry Wall released Perl in 1987. He was a linguist and a system administrator at the same time, and the language shows both influences. Perl was built to quickly search, transform, and pass on texts and files. Unlike large software projects, Perl programs often consist of only a few lines that handle a single tedious task.

Why Perl was regarded as the glue of the internet

In the 1990s, a large part of the web was built on Perl. If a webpage had a form, a Perl program usually ran behind it. It accepted the input, checked it, and wrote it to a file or database. This technique was called CGI, a simple standard through which a web server could call external programs. Perl was the obvious choice for this because it handled text effortlessly.

The nickname “glue of the internet” describes exactly this role. Perl rarely wrote the main program. It connected systems that didn’t actually fit together. A Perl script would fetch data from an old database, transform it, and hand it off to a new system. Such connecting pieces are invisible, but without them operations would come to a standstill.

Perl also passed a lot on to other languages. Its handling of search patterns in text became a role model. Today, Python, JavaScript, Java, and many other languages use a pattern syntax explicitly referred to as “Perl-compatible.” The language’s influence thus reaches far beyond its own distribution.

Regular expressions and the language’s motto

Perl’s strongest tool is regular expressions. These are short search patterns for text. Instead of searching for a word literally, you describe its shape. A pattern might read: a sequence of digits, then a period, then letters. This lets you find and replace all date entries in a file with millions of lines within seconds. In Perl, such patterns are built directly into the language, not merely bolted on.

Perl is a scripting language. The code is not translated into machine instructions beforehand but is read and executed immediately upon startup. You change a file and run it again, nothing more. This makes Perl fast for experimentation and popular for small helper programs. For very computation-intensive tasks, however, it is slower than compiled languages like C.

The language’s best-known motto is: There’s more than one way to do it. Perl allows many different notations for the same goal. This is convenient when writing but often tedious when reading. Perl is therefore considered hard to understand, sometimes mockingly called a “write-only language.” Python deliberately implements the opposite and favors a single obvious way.

Where Perl still runs today

Perl is not dead, but it has disappeared from the spotlight. For new web projects, hardly anyone chooses Perl today. That role has been taken over by Python, JavaScript, and PHP. In programming language popularity rankings, Perl no longer appears among the top ranks. New job postings for pure Perl development have become rare.

Nevertheless, a lot of Perl code continues to run. On Linux servers, Perl scripts are embedded in maintenance tasks, backup routines, and log file analyses. Banks, government agencies, and research institutions operate systems whose core was created in Perl back in the 1990s. Such systems run reliably, and no one rewrites them without a good reason. Well-known services like Booking.com relied heavily on Perl for a long time.

In the news, you therefore mostly encounter Perl in two contexts. First, in legacy systems for which experts are lacking, because hardly anyone learns the language anymore. Second, in connection with AI: language models can read and explain Perl code because it exists in large quantities on the internet. That is exactly what makes them a useful tool for maintaining such old programs.