I agree that few studies are done using the right context and variables. However it is just as rare that my past experience applies directly to whatever variation life throws me next. Furthermore I am well aware of how easy it is to fool myself.

Therefore I like to seek an understanding that encompasses as broad a range of experience as possible. That means my experience, the experiences of friends, and studies.

I don't take the studies' claims as gospel, but if I disagree with what they concluced, I strongly prefer at least having an understanding of why they would have seen what they saw.

Guess why books for children use short sentences.

Because they are geared toward people who are still learning the requisite language skills.

Guess why most programming introductions start with printing Hello, world.

Because they are geared toward people who are still learning the requisite programming skills. And Mathematical introductions don't begin with integrals and derivatives either, but that misses the point as entirely as your two examples.

One of the tasks of a programmer is to manage complexity. The size of a routine is a trait that is causally unrelated to complexity except in trivial cases. It is true that relatively shorter routines are often a by-product of good complexity management, but not by virtue of shortness itself. The aim should never be about making something smaller, it should be about making it simpler (and simpler for a routine can be with respect to several variables, both locally and globally).

    $simpler != $shorter;
[download]

In fact, as some studies have shown, simpler (in terms of errors, changes required, and even comprehensibility) is not even correlated with smaller size (within upper bounds ranging from 100 to 500 line routines). And with that in mind, I'll paraphrase Einstein: routines should be made as simple as possible, but no simpler!. People who equate smaller with simpler often overshoot the target.

I do not know of any study. I don't care much about studies either, since few are done using the right context and variables.

Well, the head-in-the-sand tactic is a common defense of personal opinion.

One of the tasks of a programmer is to manage complexity.

I like to manage complexity by breaking down the complexity into classes and methods, and keep those methods small AND simple. I know that short doesn't have to mean simple. See Attribute::Property. I also know that simple doesn't have to mean short. But more often than not, they have everything to do with eachother.

Well, the head-in-the-sand tactic is a common defense of personal opinion.

Personal study has shown that the sand is nice and warm.

Juerd # { site => 'juerd.nl', plp_site => 'plp.juerd.nl', do_not_use => 'spamtrap' }

Im very glad you spoke up in this column. I don't agree that short subroutines are necessarily more understandable and easier to manage. In fact on the contrary. I've found that short subroutines and methods make understanding the code flow of a program much more difficult. (As yourself and others such as Corion have pointed out.) Your comments here have articulated what I've thought (or felt) when I've seen this argument before. I hate dealing with code where methods only do one or two lines of action and there are many many such methods and subs. By the time I trace through a horrible maze of tiny subs each almost identical to the next my head hurts, I can't get a mental model of the program flow (and I typically want to fly to Australia to strangle the author responsible :-)

As you said elsewhere, subroutines and methods should be a small and as simple as they can be and get the job done. They shouldn't do multiple things and be huge monstrosities but at the same time they shouldn't be sliced up into bits that are only going to be called from one place.

For me a sub/method should do a single conceptual task. If that means it ends up quite long (over 50 lines or so) then so be it. I certainly don't spend a lot of time worrying how long my subs are.

Once again thanks. Tis a pitty you are anonymous or I would follow the nodes you write on a regular basis.

One last thought though. I think this particular subject has different levels. Experienced programmers can probably argue against tillys advice. We have the knowledge to decide when to sacrfifice a rule of thumb like "keep your subroutine short" on the altar of pragmatism. But for a beginner I think the advice "keep your subroutines short and your methods shorter" is probably sound. Until they have developed a more sophisticated basis by which to decide the rule is probably a sound design/craft guideline. The more you know a subject the easier it is to decide "when to break the rules".

So thanks to tilly for yet another thought provoking thread, and thanks to yourself for a bit of evidence of the contrary view.

So thanks to tilly for yet another thought provoking thread, and thanks to yourself for a bit of evidence of the contrary view.

I don't think mine and tilly's views are actually different (just different angles into this interesting topic). He mentioned an OO colleague of his having problems with McConnell's information regarding longer routine sizes, and he noted he found the OO does tend to have reduced routine (method) sizes (perhaps explaining the OO programmers bias). I agree OO tends to have more shorter routines than equivelant procedural code, and I even waxed (hypo)theoretic on one possible reason this might be so (OO allows organized coupling, freeing the programmer from managing it). Not that OO code can't be found (all too easily) that suffers from overly monolithic routines or overly fragmented routines (there is no spoon, I mean, panacea).

Tis a pitty you are anonymous or I would follow the nodes you write on a regular basis

You still can. You just won't know if its one of my better nodes or just some drivel until after you look at it :-)

I hate dealing with code where methods only do one or two lines of action and there are many many such methods and subs.

So do I, so do I. I like short methods, not tiny ones :) I think screen page (24 lines) size is a good size for subs (methods or non-methods) and anything much larger should be taken apart.

Subs with only one or two meaningful lines are not what I was talking about when I meant short subs. They're a little too short :)

I do prefer one-line subs to five-line subs when it comes to often repeated code.

sub property1 {
    my ($self, $value) = @_;
    @_ >= 2 and $self->{property1} = $value;  # update: s/>/>=/
    $self->{property1};                       # update: s/property/pro
+perty1/
}
sub property2 {
    my ($self, $value) = @_;
    @_ >= 2 and $self->{property2} = $value;  # update: s/>/>=/
    $self->{property2};                       # update: s/property/pro
+perty2/
}
[download]

versus

sub property1 : Property;
sub property2 : Property;
[download]

Juerd # { site => 'juerd.nl', plp_site => 'plp.juerd.nl', do_not_use => 'spamtrap' }

Why are you assigning to $self->{"property$n"} and then returning $self->{property}?

Your @_ test also makes little sense to me. If you are trying to write get/set routines, you would want to get on a single routine, set on 2, and think about throwing an error on 3 or more. Your methods as written require an extra argument to say, "set, pretty please".

I was also going to chide you for having methods called property1, property2 without thinking through what you are going to do if you had more than 2 properties, and then I realized that you were likely talking properties of an object, not a computer representation of physical properties. :-)

Oh, right. If you are going to have a lot of very similar code, then I think that autogenerating can be good. (Assuming, of course, that the person who will take over the code is capable of figuring out autogenerated code...) For instance:

foreach my $accessor (@object_properties) {
  no strict 'refs';
  *$accessor = sub {
    my $self = shift;
    $self->{$accessor} = shift if @_;
    $self->{$accessor};
  };
}
[download]

(Yeah, yeah. A million CPAN modules implement variations on the above for you.)

I was reviewing some of my code and it looks like my sub lengths (excluding short utility subs that would probably be macros in C or inline in other languages) seem to be around 50-150 lines.



---
demerphq



First they ignore you, then they laugh at you, then they fight you, then you win.
-- Gandhi