Memories of 20 Years of Perl

Dec 21, 2007 by chromatic

Proving Them Wrong

Around 1991 I wrote a very useful program, in C, which took a bunch of files and then sorted them into groups according to which files had identical contents. A lot of sysadmins at the time wrote to thank me for it. But when I boasted about it at Usenix that year, people told me “oh, you should have written then in Perl.”

That was pretty annoying, so I got the Camel Book (pink in those days) so that I could learn Perl and prove that they were wrong. But it turned out that they were right.

Mark Dominus is the author of Higher-Order Perl

My First CGI Program

It was the year 2000, and I was working at a software startup in San Francisco. I was tasked with writing a simple form handler with an auto thank you email. I had been a C programmer for several years, a Fortran programmer for a few, and this was essentially my first Perl program. It was your standard CGI gateway which presented a form to the user, did some error checking, and sent a thank you email to the user.

After a few hours of learning Perl and putting my form handler together, it was put live on our website. I was delighted that I was able to pick up this language so quickly and produce results in a short period of time. I never like programming C that much (although that has changed), due the fact that it got in my way. Perl just worked.

I came into work the next day and reviewed how my program was doing. It turns out that my first bug had surfaced; the thank you email function managed to get caught in a loop. One poor soul who filled out my form had received 800 thank you emails! I was able to quickly fix the bug.

In honor of my first Perl program, I would like to extend a hearty 800 thank yous to the Perl community! I have been using Perl ever since and love it.

Fred Moyer is just another mod_perl hacker

Perl and the University Student

One cannot imagine how useful Perl proves sometimes to a university student. I can recall several occasions in which I used Perl to facilitate a task or check my homework. Of them, there is one that I still remember very clearly.

It was the course “Introduction to Computer Networks” and we learned about the various variations of networking protocols (Stop-and-wait, Go-back-N, and Selective-Repeat). We were given a simulation of these protocols written in C and compiled to run on Windows. The simulation could be ran with several parameters and would output a verbose file with the parameters of the simulation, the simulation itself and then some statistics of the simulation.

We ran the program several times and got several files in return. Now we had to somehow insert the statistics into Excel so we can analyze them, process them, and create charts out of them. But the statistics were scattered over several different files, all with the same format, but nothing that Excel can understand (at least not without a massive amount of Visual Basic for Applications code).

Without thinking for a moment, I started writing a Perl script that will process the files, extract the corresponding data and output a tab-delimited file that can be inputted into Excel. It took some time to write the script, and meanwhile my partner decided it may be faster to do it by hand. Thus, he occupied the nearby station, and started extracting the data himself. I finished a few minutes after that, though, (while he was just beginning in his manual labour) and we were able to input the data into Excel and continue the assignment. It took about 15 minutes or less, all in all.

Later on I talked to a few fellow students about the assignment. One of them claimed it took him 3 hours to input everything into Excel. (!) Another said it took him one hour, which is still much worse than 15 minutes. Needless to say, none of them knew Perl.

Enough said.

(Originally published at Perl Success Story, Israel.pm.)

Shlomi Fish has worked with Perl since 1996 and considers himself a happy user, developer and advocate of Perl and other open-source technologies.

How To Become a Guru

In early 1999 I started a new job as a system administrator. In my previous position I’d taught myself Unix and GNU/Linux, and ended up writing a small tracking application for a customer service group in Java.

As a new SA, I took over a pile of work from my predecessor, including some small Perl programs he’d downloaded, installed, and modified to add his name to the comments. Over the next couple of months, I picked up the Camel and the Perl Cookbook, and taught myself enough Perl that I could skim comp.lang.perl.moderated and answer some of the questions in my head.

About that time, I started to do a little work on the Everything Engine – not much, but a little bit – and so I was the second external person to register on PerlMonks when it started. In those days there was no voting, no XP, and there were just a few people racing to reach the milestone of a hundred posts.

In between troubleshooting problems at work, I’d play with little programs, read whatever tutorials or books I could get, and answer any question I could on the site, and so I learned Perl that way.

I remember the rush to find an idea – any idea – worthy of putting on the CPAN, and thinking in 2000 that every problem that anyone could solve, someone had already solved. I remember my first patch to Perl 5, then realizing that I hadn’t actually run the tests, and resolving to improve the tests because they didn’t actually do what they said they should.

I remember getting job offers from my postings, and meeting some of the top Perl programmers in the world for the first time, and being accepted because I did (some of the) things I said I would, just because no one else was doing them.

That, I think, is the secret to become a contributing member of any community. Look for something that needs someone to do it and do it. You don’t have to have permission, just a little bit of determination and stubbornness and some time.

I’m a little sad that I missed the first eleven years of Perl’s life, but I’m glad to have caught up in the past nine years.

chromatic does a lot of things, some of them even sometimes productive.

How an English Major Saved Christmas

Right before Christmas of 1998 I was a fairly new employee at Amazon.com. Not a CS grad hacker with 30,000 shares, but an English grad customer service rep with 250. I knew about the 29,750 share disparity from picking up a fax for a star employee in the apps group. Instead of letting it get to me, I started to look into why it was so. I bought Learning Perl and spent two of the most painful weeks of self-edification in my life discovering how the lack of chmod +x was preventing me from getting through Chapter 2.

Free at last I wrote, in two days, a badly needed and overlooked tax + shipping costs calculator for customer service for the new product tab launching that week. It was the kind of script that would take any decent Perl hacker 30 minutes. A former art critic saved hundreds of reps and tens of thousands of customers a lot of time and aggravation. I got the company’s “Just Do It” Award. If it had been C or Java or anything but Perl I wouldn’t have been able to do it.

If I’d come to anything but Perl, I would not have returned to coding–I dabbled in BASIC and Assembly as a kid–and I wouldn�t be a software developer today.

Ashley Pond V is a New Mexican writer turned Seattlite software developer, currently working with Catalyst applications, who credits Perl with saving his soul as he’d probably have gone into marketing otherwise.

Smells Like Wet Camel

Standing out in my memory is the day in college (either in late 1993 or early 1994) when my grandmother had emergency eye surgery. Originally, she only had a regularly scheduled checkup, and my mother could take her to the appointment before work began, but not pick her up. The doctor was one street over from the college (more or less) and I was conscripted to go over and take her home after her appointment and my first class. The day was rainy, increasing in intensity as the day grew older.

Everything changed when I arrived at the doctor’s office, because the doctor had found something that required immediate attention. She had to be taken to a specialist immediately, and I began improvising. Each eye appointment took a long time, and they would only get longer as my grandmother was worked in to the specialist’s schedule as an emergency patient. So I had time to take her to the next appointment, leave her to wait for what might be hours, go to my next class, eat lunch, and come back and get her.

I was trying to keep up with my college work, and brought my O’Reilly Perl book along so I could work on my computer science project, figuring I might as well do something useful while I was sitting around. My project involved writing an e-mail processing system in Perl, so I had bought what was for me at the time an almost impossibly expensive book to help me learn the language. On the way to the car, in the hardest and coldest rain I can ever remember, I was trying to help my grandmother and juggle the umbrella, car keys, car door, and everything else. The book slipped out from under my arm and landed in a puddle. Somehow, it landed on its edge, and had about an inch of muddy water soak into it. My new book! Ruined! Nothing to do but keep going, to the next appointment, and back to my class. I knew that to leave the college after eight a.m. was a guarantee of not being able to park anywhere near the building for the rest of the day, because the only parking spaces left were in the lower area of an overflow lot far from any building I needed to go to. Without even a sidewalk near this lot, I had plenty of time to think about my ruined book and what was happening to my grandmother as I trudged through the mud, in the pouring rain, to get to my next class.

I also, in these days before mobile phones, had to find a pay phone to tell my mother about the abrupt change of plans. My grandmother eventually got settled in the hospital, where it was at least dry, and she pulled through the eye surgery fine. My waterlogged book with a brown bottom and hastily scribbled notes on the blank pages in the back was a good enough starting point; I graduated.

Scott McMahan has been writing Perl code since 1991.

“I Couldn’t Believe That Perl Even Worked”

My first exposure to Perl was a web server with – I think – Perl 4.036 installed. This would be 1995 or so. I wanted to write CGI scripts so I started reading everything I could find about Perl. I nearly lost heart when I read that the parser was, effectively, heuristic. Coming from a background in Pascal and C I couldn’t believe that Perl even worked.

Fortunately Perl was the only option for my script. I persevered and discovered that – not only did Perl work – I rather enjoyed it. Within two weeks I had a CGI script that implemented a kind of ad-hoc PHP: chunks of Perl embedded in HTML. It was ugly – but Perl had made it possible.

At some time between then and now – after digressions into Java and even LotusScript – Perl became my main language. At the end of 2006 I decided to concentrate on Perl, release some modules, proactively seek out things I didn’t know about the language and learn them.

As a result 2007 has been the happiest year of my professional career. I’ve written loads of code, most of which works. I attended my first YAPC in Vienna and came home with a bunch of new friends and a renewed enthusiasm for cranking out code.

I’ve still got plenty to learn. Perl may be easy to pick up but mastery takes years. And if you love programming that’s part of the fun. However good you think you are there’s always a way to improve.

I dabble with other languages – because if you take programming seriously you must. What do they know of Perl that only Perl know? There are things about Perl that grate. It’s not perfect but it’s, well, loveable I suppose.

Thank you Perl community. Thank you Larry. Thank you for a lovely language.

Andy Armstrong is a compulsive Perl abuser based in Cumbria, UK.

From awk to perl

In early 1990, I was working with a large set of data that needed to be massaged and formatted so that it could be statistically analyzed.

I started the task in awk, but quickly ran into trouble because awk could only open one file at a time. A quick search through the Usenet comp.lang group found Perl 3.0, which had just recently been released.

I had to get the source code and build it on my machine, but it compiled cleanly and I was able to try some simple stuff. Worked real good too. As I had already a large awk program, that I didn’t want to re-edit for Perl, I ran it through a2p and the perl version produced the same results. I was hooked. When I got stuck, asking questions on comp.lang.perl almost always got instant answers. There has been an active perl community for a long time, and they were fabulous! (Just like now). I subsequently re-factored my code for perl and produced vast quantities of data to be analyzed. I have been using Perl ever since.

Roe McBurnett is a systems engineer for a telecommunications company and has been working on telephony related projects as a developer, systems engineer, and software tester since 1985.