Functional Programming Doesn’t Work (and what to do about it)

Why in the world would you write 3 to 5 lines of code when one will suffice?

if (a > 0) {
  a1 = a + 1;
} else {
  a1 = a;
}

should be

a = (a > 0) ? a + 1 : a;
or
a1 = (a > 0) ? a + 1 : a;

It’s controlled, it’s functional, and it’s perfectly readable. 

I don’t understand why people make these things so much harder than they need to be, even when making a point.

As for your point:
>> Making it a true global counter removes a whole layer of data threading..

Unless you’re using a singleton, globally scoped variables are near-terminable offenses, IMHO, especially if repeated.

Seriously, there is no reason for, in the case of a game, a score to be globally scoped. It’s all about accessibility that makes sense, and a good design up front solves that, every time.

Yes, FP is hard if you start with this premise:

“Now make these dependent on each other: function A behaves differently depending on the number of times function B has been called and vice-versa.”

This is BAAADDD!!!!

That means that these functions are codependent, they are highly coupled, and they need to be redesigned.

For FP to work, each function result can only depend of the input parameters (which can be another function too).

If your app architecture requires that highly coupling functions (and I would doubt first the architecture that has that requirement), then FP is the wrong tool for the job… but then again, I would distrust the relationship between those functions first.

@jorge

The level at which the codependency is bad is too abstract.  It might not be an app architecture rule, it can be an 11th hour business rule change, or a human interface improvement for tracking consecutive mistakes, etc.

If you need to have a fully realized understanding of the problem at hand, and the solution to that problem, in order to employ FP effectively, that’s a very crippling requirement for a vast majority of software projects.  I don’t think the situation is quite that bad, but I agree with the authors point.

Jorge: it’s not BAAADDD if it’s a requirement.  Complexity should be minimized and managed, not shunned.

lol, you’re pretty bad at this aren’t you.

In your second to last paragraph did you really mean “...and all of a sudden a large part of the code has shifted from imperative to functional”.

If I understand you then wouldn’t the shift “Make those globally accessible and modifiable…” have happened the other way around?

If not then, surprise surprise, I didn’t understand you.

@stanley

If you know the final specification of any system before you write it then you are not working on the same kind of systems that I am.  My users are constantly changing/adding to the spec in ways that totally break even java code (where you obviously have globals/statics etc to hack around stuff) and mean you need a major rewrite.  This doesn’t go down well with users if you have to do a major redesign cos they asked for F1 to depend on F2 as in the original example, especially if they know that they could just fix it if the system was written in their favourite language (VB).
If you’re going to get functional languages into mainstream then this is an important topic that needs proper coverage imho.

Uh, all six of those URLs at the top of the post seem to point to the same article (this one).  Could someone fix the links?

I have never known that functional programming will not be much useful and now I am quite surprised because there are many people who actually preach about the benefits of functional programming. I believe that these guys are going wrong somewhere which they may realize that a very later stage. - Jordan

Previous comments have covered the Func A dep on Func B problem - a very poorly chosen example - and a strong clue that you have not fully grasped the basic ideas of the functional paradigm.  You’ve also missed the idea of binding and expressions in a functional context.  In erlang, I have so rarely written an “if” block like the one you presented I could probably count the number of times on one hand.  An “IF” statement in an imperative language is an IMPERATIVE construct.  In functional langs like erlang it is just a short-cut version of CASE.  And CASE is just a guarded expression.  And in good functional fashion we are only really interested in the VALUE of expressions.  In that sense we “try” to avoid side-effects in our funcs because that is something more than just a simple value.  Its not so much that its harder to think about programs that have side-effects, its that its harder to think about programs where side-effects are firing off all over the place and harder still to promote them the (now prevailing) concurrent context.  It is unrealistic to have “pure, pure, pure” functional code - perhaps a reason why erlang (which I consider a hybrid) thrives commercially and haskell largely remains an intellectual curiosity,

I have to agree with Randy. Good design trumps almost anything. I was talking to my father the other day and he was telling me about the “old day” when he had to program on punch cards with 16k of memory on the computer. Things had to be precise and there was no room for excess. Now with so much hard drive space and memory available to programmers, they have gotten lazy and messy with their code. When my father was working on disaster recovery software the name of the game was no excess code, even if the developers thought it was easier.

Thanks for providing this information. It’s quite informative and helpful. Most especially to those notices users. Two thumbs up for the contributor.Best Binoculars Under 100

awesome article

Can’t seem to view more than a few minutes of this presentation, then it dies - tried more than 10 times over the past couple of days.
Duvet Sets
Do you have a mirror of the presentation anywhere?

Well I think both of the first commenters have made valid points on their claims. Regards, Angel of territory mapping software company.

I’m having the same issue with the presentation cutting off after about 10 minutes. Anybody knowhow to resolve the problem? I’m on firefox 3.6 if that helps. Steve from san diego virus removal

I liked the posts and cool layout you have here! I would like to thank you for sharing your experience and the time it took to post!! Two Thumbs up!
dolce gabbana perfume

I just want to say that it is an elegant blog that have useful information. I will visit this site again for more information. Take care & carry this job of sharing nice information with others.
ripenaxrau hjd electroluxap fgd tpersonalcar

This doesn’t go down well with users if you have to do a major redesign cos they asked for F1 to depend on F2 as in the original example, especially if they know that they could just fix it if the system was written in their favourite language (VB). i don’t agree with that at all.
Bail Bonds Los Angeles

I have to agree with Randy. Good design trumps almost anything. I was talking to my father the other day and he was telling me about the “old day” when he had to program on punch cards with 16k of memory on the computer. Things had to be precise and there was no room for excess. How to Consolidate Debt