"Perl is unDead"

schwern on 2008-06-17T15:59:11

This morning, at 8:15am when only the dead should be awake, I gave my "Perl Is unDead" talk. I did it before at YAPC::Asia but this time it took half the time and I felt the presentation was better. Maybe because I've done it before. Maybe because I wasn't feverish. Maybe because the microphone didn't work and I got to shout the talk. It really lends itself well to shouting. :)

---

summer re-runs?

gabrielle on 2008-06-17T18:22:38

Will $pdx_lug get a repeat of this? The slides are killing me, would love to see the real deal.

(8:15? only the dead? I've had breakfast twice by then.)

Re:summer re-runs?

chromatic on 2008-06-17T18:27:21

(8:15? only the dead? I've had breakfast twice by then.)

Schwern meant 8:15 am.

Re:summer re-runs?

schwern on 2008-06-17T21:29:03

Sure, I can do a re-run at PDXLUG. It's a fun talk to give.

What cultural values do we have?

btilly on 2008-06-17T22:46:07

Back in the 90s there was a Perl programmer who thought as you are encouraging people to think now. Nice young kid, he set out to make useful applications in Perl. They solved real problems. They were easy to install. They were easy to customize. Those were his criteria, and by those criteria he succeeded brilliantly.

The Perl community hated him. For years afterwards his name has been vilified. We've complained horribly about the bad impression he gave Perl. We still complain about that.

Why? Because Matt Wright didn't understand security. He didn't design his code for reuse. He didn't follow basic best practices. And many beginning programmers used him as a model for how they should program. Every one of those criticisms is valid.

Now go look at some of those PHP applications out there. Examine them. You'll find that Matt Wright's spirit lives on in PHP, in spades. Similarly look at how Rails handles databases. If you care about database design, that's not the way to do it.

Then go read Worse is Better and see the parallels. Characteristics which make software worse from one perspective often make it better from another. In particular maintaining programming ideals around areas like security, good design, code reuse and the like often get in the way of making software that will become popular.

What are our values? How much are we willing to bend in the name of popularity?

Re:What cultural values do we have?

schwern on 2008-06-18T18:46:35

The trick is, how do you make code easy to install while remaining sane and modular? I think we can do better than throwing out modularity and backsliding to single file installs.

Re:What cultural values do we have?

btilly on 2008-06-18T21:56:53

We can make it more sane and modular than single file installs, but not without making the install process more complicated than a single file install is.

That said, we've gone a long ways towards making things simpler. Just compare how difficult it is to get a Perl library installed versus installing a useful Java library. We can make it simpler still, but Perl already has a pretty good story to tell.

However it is never going to be as simple as a single file install. It just ain't going to happen unless we're willing to compromise our values.

Excellent, Mr Schwern, as usual.

Tim Bunce on 2008-06-18T08:37:23

Mind if I pinch a slide or two for an update to my Perl Myths talk?

Re:Excellent, Mr Schwern, as usual.

schwern on 2008-06-18T18:40:58

Only fair, I pinched some of yours.

Is there video?

Aristotle on 2008-06-18T19:23:35

Of the talks recorded at YAPC::Asia this is the one I really wanted to see, but due to the terrible reverberation at that venue and my non-native English speaker ears, the speech was completely unintelligible to me almost half the time. Is there video available from YAPC::NA?

Re:Is there video?

schwern on 2008-06-18T20:57:31

There was a camera in the main auditorium and I believe Krishna was filming. So look for it.

Fair's fair

mr_bean on 2008-06-21T06:02:09

The funniest part of the talk is:

Gandhi said:
                First they ignore you...
                Then they laugh at you...
                Then they fight you...
                Then you win.

We won.
Then what?

                We became the Empire
                We ignored... (Ruby)
                We laughed... (at PHP)
                We fought... (with Python)

Now they may win.

I think (following Franklin D Roosevelt, who was apparently following someone else) the only thing to fear is fear itself.