Live data from Hacker News

Home at Last, or, The Last Programming Language I Will Ever Learn

colinsteele.org

71–77 of 77 posts

Re: Home at Last, or, The Last Programming Language I Will Ever Learn

#71
strongly recommend not stopping with clojure. Find other languages outside your comfort zone, that require you to learn about different technologies and different ways of thinking.

You should know at least one language with a strong type system that's closely modelled on the hardware environment. (eg C or Java)

Databases and SQL.

OS-level tracing/debugging languages like Dtrace or SystemTap.

Puppet/system administration

Javascript/Browsers

Mobile/? (I don't know much about phones)

An ML-style language.

Re: Home at Last, or, The Last Programming Language I Will Ever Learn

#72
post #36

The punchline is...."So about six weeks ago I picked up Cloure" - If you're linkbaiting your article, you should always spellcheck.

flag . Not something that I generally do, but (a) clojure won't be in any spellchecker, and (b) your snark adds nothing to the discussion. If you wanted to say "and that language was clojure", you could have left it at that.

Not sure why this is being down voted. It was a comment that added absolutely nothing to do the discussion.

Re: Home at Last, or, The Last Programming Language I Will Ever Learn

#73
post #58

Earlier quoted context omitted.

Then I urge you to look at it again - who cares what the users believe? What matters is what you get out of it.

I hacked a lot of CL back in 2002-2003 and I've dabbled in Clojure lately. Clojure has a lot of promise but I'm waiting for some of the rough edges to get smoothed out (stacktraces in particular).

If you're adventurous enough to try out the 1.3 alphas you'll find the stack traces immensely improved.

Re: Home at Last, or, The Last Programming Language I Will Ever Learn

#74

This is exactly the kind of cultish triumphalism that put me off Common Lisp. The inability of many lispers to see lisp, like all languages, as a collection of design tradeoffs instead of the one true way just clashed too sharply with reality for my tastes. It's certainly true that lisp has been an important testbed for many concepts that have later made their way into the programming mainstream, but I suspect that m…

As Lisp is very good at adopting paradigms from other languages - the only design tradeoff that I've consistently heard is the syntax - and this usually from people too blind to realize their syntax is hardly better.

So what do you see as a fundamental design tradeoff for Lisps?

Re: Home at Last, or, The Last Programming Language I Will Ever Learn

#75

This is exactly the kind of cultish triumphalism that put me off Common Lisp. The inability of many lispers to see lisp, like all languages, as a collection of design tradeoffs instead of the one true way just clashed too sharply with reality for my tastes. It's certainly true that lisp has been an important testbed for many concepts that have later made their way into the programming mainstream, but I suspect that m…

As Lisp is very good at adopting paradigms from other languages - the only design tradeoff that I've consistently heard is the syntax - and this usually from people too blind to realize their syntax is hardly better. So what do you see as a fundamental design tradeoff for Lisps?

people too blind to realize their syntax is hardly better

That's exactly the kind of categorical statement I'm so tired of hearing from lispers. Other syntaxes score better on every metric (terseness, readability, expressiveness) except flexibility. Lisp makes the same design tradeoffs the other dynamic languages do but bets the farm on macros. Over 40 years of historical evidence suggests that most programmers do not find the power of macros adequate compensation for a syntax they find highly unpalatable. It takes a special kind of arrogance to dismiss all of these people as blind.

Re: Home at Last, or, The Last Programming Language I Will Ever Learn

#76

Earlier quoted context omitted.

As Lisp is very good at adopting paradigms from other languages - the only design tradeoff that I've consistently heard is the syntax - and this usually from people too blind to realize their syntax is hardly better. So what do you see as a fundamental design tradeoff for Lisps?

people too blind to realize their syntax is hardly better That's exactly the kind of categorical statement I'm so tired of hearing from lispers. Other syntaxes score better on every metric (terseness, readability, expressiveness) except flexibility. Lisp makes the same design tradeoffs the other dynamic languages do but bets the farm on macros. Over 40 years of historical evidence suggests that most programmers do no…

Except syntaxes die. Syntaxes tend to encode what is possible to describe in a particular language. Eventually people decide what a language expresses is not adequate, they write a new language w/ new syntax to describe a new thing. Lisp adopts the new thing with little in the way of syntactical change.

How do you account for that?

  * Over 40 years of historical evidence suggests that most programmers do 
    not find the power of macros adequate compensation for a syntax they 
    find highly unpalatable. It takes a special kind of arrogance to dismiss 
    all of these people as blind.
No 40 years of historical evidence suggest that your syntax will die. C and C++ have survived because of their intimate nature w/ popular operating systems. Lisp has survived because it can evolve. How can you not chuckle that C and C++ are just now getting lambdas and closures?

In anycase Alan Kay says it better than I, http://queue.acm.org/detail.cfm?id=1039523.

Re: Home at Last, or, The Last Programming Language I Will Ever Learn

#77
I am old enough (late 50s) that settling down on just one programming language would not be a bad strategy if I was not a consultant. Maybe.

Over the last several years, I have been hoping to mostly just use Ruby for future work, but in the last 5 years almost half of my consulting jobs used a Lisp language (Common Lisp or Clojure) and the rest mostly Ruby, with some Java.

The Clojure language and libraries are very nice, and I am sure that problems with the development experience like poor stack traces (good work already in progress), etc. will get ironed out soon.

If you are an entrepreneur with your own product or web services based company, then pick one great language and use it until a better great language comes along.

For those of us who enjoy consulting, we will mostly use what best fits customers' needs.

Post reply on HN