Damien Conway interview: Perl 6, Java, his PhD

http://fyi.oreilly.com/2008/08/the-mind-of-damian-conway-scie.html

by gtani • 18 years ago

14 11 18 years ago

fyi.oreilly.com

astrec2• 18 years ago

Ah memories! I consider myself very fortunate to have been one of Damian's students. The container box store mnemonic took me way back.

He's not overstating the diversity thing either. In those days the CS program was C, Lisp, Prolog, C++, sh, MIPS assembler, operating system design, AI, digital logic, algorithms, formal methods, and discrete mathematics.

These days, the kids I hire from the same school mostly cruised through IS style degree with only Java/C# under their belts.

bootload• 18 years ago

"... Ah memories! I consider myself very fortunate to have been one of Damian's students. ..."

Ahh those whacky CS students from Monash. I went to RMIT & Swinny and missed "The Damian". What kind of lectures did he run? I've heard some of his talks & they are pretty good ~ http://itc.conversationsnetwork.org/shows/detail880.html If your lucky you can still hear him talk for nics at Melbourne Perl Mongers.

astrec2• 18 years ago

He taught the first year intro to programming (C) with Ann Nicholson (AI researcher), and supervised a few post graduates. Ann also taught the Lisp course.

Damian was a fantastic lecturer with a rare ability to enthuse and engage his students while he delivered otherwise dry material. A true master of the story.

albertcardona• 18 years ago

I loved the part that said: in diversity there is power. You should know as many programming languages and paradigms as possible. And code every day to stay fit.

martinr• 18 years ago

Am I the only one who considers Perl6 dead and irrelevant? I haven't seen anything approaching a usable product out of the efforts, and how long has the project been underway now? By the time it's finished, Perl5 is irrelevant as all the kids will be using Python, Java, Ruby, maybe D. Legacy applications won't switch from Perl5.

gaius2• 18 years ago

Perl 6 is an academic research language, which is weird, normally languages come from academia into commercial use.

neilk• 18 years ago

That's an interesting insight. It's not truly "academic" because such a language would not be considered interesting in a research sense; it has no clear vision, no central metaphor for computing.

But if you consider the Perl community a sort of academy with its own tastes and goals which have nothing to do with mainstream acceptance, then yes, it is academic.

asjo• 18 years ago

I think you are not alone. But then again, there are also many people who look forward to get to use Perl6 for real.

It has been underway for a long time, since 2000. Do notice though, that Perl6 isn't just a revision up for Perl5, it is a thorough redesign.

As for all the kids - well, if they are going for the (percieved) latest and greatest language all the time, they will be quick to try Perl6 when the right christmas comes around.

I think Perl6 will hit a sweet spot, but we'll see. Maybe your dire predictions will be fullfilled; maybe not.

I don't think age necessarily is a bad thing for a language - I bet a lot of programs you use are written in C, for instance.

greyman• 18 years ago

Perl6 is not dead. I think the basic idea behind perl6 was good, but they had to release a few years ago to be relevant. Now they will have hard time competing against ruby or python. We will see, thought.

gaius2• 18 years ago

A lot of people were burnt by unmaintainable Perl codebases in the 90s. Right now, even if I wanted to, I couldn't start a new Perl project at my company, its reputation is too bad. People who left Perl for Python or Ruby won't be back - ever.

SwellJoe• 18 years ago

Am I the only one who considers Perl6 dead and irrelevant?

No, lots of people have expressed that opinion. But, the reality is that Perl (5) is still more popular than Ruby and Python (probably combined). I suspect that if anything really derails Perl, it won't be Ruby or Python.

It also certainly won't be Java or D, which are going after systems-level development tasks, and, in the case of Java, accidentally ended up being somewhat popular for poorly fitting tasks, like web apps. Perl and Java, actually, pretty much came along at the same time, and both have coexisted mostly peacefully at two ends of a spectrum for many years (and now, over at the Perl end, though maybe not quite as far down the road, Ruby and Python also coexist mostly peacefully).

I suspect that if anything derails the popularity of Perl it'll be that JavaScript will have a good server-side implementation before Perl 6 becomes a solid reality, and it'll also derail Python and Ruby from ever reaching a level of true pervasiveness (on the order of PHP, which will also lose out significant market share). And, of course, this theory only applies to web apps, where the current server/client dichotomy is less comfortable than it should be, that this will make any sense at all, and it doesn't mean any language will stop being used for web apps, at all. Millions of lines of code don't stop existing overnight. I just think that when the choice is between "use one pretty good language on the client and server" or "use my favorite language on the server and a pretty good language on the client and try to make it all fit together seamlessly" and apps become more and more rich on the client-side, it'll end up being obvious to choose the single language option.

Of course, a lot of software is not a web app, and so there will remain systems-level languages, and dynamic languages for various types of text processing and other such things, and Perl and Python and Ruby will be among them. But, if a good implementation and good libraries show up for JavaScript on the server, it will have a huge impact on all current web application languages popularity.

Another interesting game changer is dynamic languages on the JVM and CLR (and, probably, Parrot). I suspect that will also have a dramatic impact--but it won't lead to the death of any language. It'll make more languages flourish, as it removes the requirement that a language have a reasonable standard library before it can gain any traction--the library will be "anything else that runs on that same VM".

is irrelevant as all the kids will be using Python, Java, Ruby, maybe D

The kids are already using Python and Ruby. It doesn't mean kids aren't also using Perl. And, of course Perl 5 hasn't been sitting still...5.10 has some nice new bits and pieces pulled back from Perl 6.

Legacy applications won't switch from Perl5.

They won't have to. Like all Perl versions for the past 10 years, or so, Perl 6 is backward compatible. It's possible to "port" a Perl 5 app to Perl 6 instantly, and then begin making use of Perl 6 features as appropriate. Of course, you won't get the benefits of Perl 6 in sections of your code that are Perl 5, but they can be ported as time allows and as refactors happen naturally.

Damien Conway interview: Perl 6, Java, his PhD

Related Stories