This is a flame-out or a troll by some random JoS forum poster (not Joel Spolsky); it has no more merit than hundreds of similar eruptions that get posted to the usenet every day. I don't think such posts deserve comment.

The potential for illegibility in Perl code is an often discussed topic, both in PM and elsewhere. It is not necessary to broach the subject this way. IMHO it was an unwise decision to post this in PM. Even if you are quoting someone else, it is still garbage.

Also, I think the title of your thread is very misleading, because it gives the impression that the rant's author is Joel Spolsky. I have no idea what Spolsky's opinion of Perl programmers is. For all I know, he may have an even more negative opinion of them than this, but these are not his words.

Go look at posts by tilly, tye or Ovid. To the best of my recall none of them are particularly "into" obfu. Sure they might appreciate one occasionally but I can't imagine any of them advocating seriously use of obfu in production code. tye in particular is a good example. I work with his code often. He definately isn't in to putting to much on one line (im much worse about that).

Its a rant. Hes got some points, but theres not much point in discussing them as they are well known and a big part of the reason we are all waiting on perl6.

Actually, answer something first. Why are you posting this? What are you going to get out of this? There are far too many flames already and personal style/legibility has been discussed at length.

Finally, I suggest you retitle the node. It's misleading. Joel Spolsky didn't write that rant.

I've updated the title - didnt realised it's potentialy misleading untill you folks pointed that out. Guess it's the way you look at it(TMTOWTDI).

- Why are you posting this?
It got me interested and wanted to hear someones opinion on it.

- What are you going to get out of this?
Looks like only --'s :)
OK, seriosly. I didnt expected anything except to hear what some people think about few of things that were pointed out.

As I wrote, I dont like the way guy wrote it nor I liked everything that he wrote. But I must say that he and others (others were more "constructive") pointed out few things that I doo agree with.

It got me interested and wanted to hear someones opinion on it.
Maybe you should have just pointed it out in the Chatterbox instead. I have had to deal with some really bad Perl code in the past written by folks who didn't understand Perl in the least, and while the example is apparently by someone who does understand Perl, the fact is ... Perl is hard. I coded Perl for 2 years before i even considered myself a decent Perl programmer, and then i came to the Monastery. I'm still decent at best, IMHO -- but, as one commenter to that cantankerously crass commentary wrote:

Only the top 1% of programmers appreciate perl.

This also clears up the recent confusion about who is a top 1% programmer.

Now, that's clearly a hyperbole as well, but there is a lot to be gleamed from that statement -- it takes a lot of discpline to code Perl in a manner that is readable and maintainable by others. Perl loads the gun, cocks it, and aims it at your foot. But it is ultimately you who pulls the trigger.

OK, seriosly. I didnt expected anything except to hear what some people think about few of things that were pointed out
? Did you miss the comment by the same author (Lies of Society) that said:

Here's the problem [with Perlmonks]. All posts (nodes) can be upvoted or downvoted by anyone who's attained user level 2 (takes a few days, maybe a few weeks). They don't have to say why they downvoted--what their justification is--and there's no form of metamoderation like at Slashdot. Your reputation (XP), the thing that lets you move up to other user levels and gain new privileges, is impacted when you're downvoted...

Surely if you had read that then you would expect to be downvoted by posting such a negative comment here at the Monastery. <sarcasm>After all, we are a bunch of cowards who participate in group think and ostracize any and all who do not adhere to our doctrine!</sarcasm> Well, i'll tell you why i downvoted you. First, you could have asked this in the chatterbox. Second, even if you did chose to post it, you didn't have to post the contents. Now, anyone who knows me personally knows that i swear like a sailor on a regular basis. So many expletives spew from my mouth on a daily basis, but even i think this kind of language is not necessary here at the Monastery. A simple link to the article would have sufficed. And the third reason i downvoted? Well, because you didn't point out which points you thought were good. You didn't point out which responses you thought were good. You just posted the rant without adding any value of your own.

jeffa

L-LL-L--L-LL-L--L-LL-L--
-R--R-RR-R--R-RR-R--R-RR
B--B--B--B--B--B--B--B--
H---H---H---H---H---H---
(the triplet paradiddle with high-hat)

OK, seriosly. I didnt expected anything except to hear what some people think about few of things that were pointed out.

And what things do you think he "pointed out"? That bad programmers can write bad code? Hardly a ground breaking observation. That he hates various members of the perl community? Well, that's slightly more original, but lacking any credible documentation, so we can probably ignore it. After that, what is left that he points out?

To me the entire thing reads more like a joke. A deliberate troll trying to get responses.

It's a fair criticism to say that perl culture encourages code that is tricky, sneaky, or magical in some way, and that tendency has consequences on how people view perl and perl programmers. It's quite a touchy issue though, as you can see whenever someone suggests that writing simpler code would make for easier maintenance and then gets attacked by people saying that there's no point in toning down code for people who don't know perl well enough.

I'm not sure the author here chose a very good example. This one looks like someone who didn't know about fancier things like IPC::Open2, not someone trying to be tricky. He's right about combining statements with "and" though.

This one looks like someone who didn't know about fancier things like IPC::Open2, not someone trying to be tricky. He's right about combining statements with "and" though.

he was talking about C::Scan originally written by Ilya Zakharevich (ilyaz) and now being maintained by hv (the current perl pumpking). I doubt they didn't known about IPC::Open2!

For me it's understandable, they were just focusing on the important thing, parsing the C code, and leaving secondary things like portability for later... or just going for a "it works for me but patches are welcome" mode.

Er, ex-pumpking - see The perl 5.10 pumpkin passes.

I'd say his point about C::Scan is perfectly fair, and the sort of thing that isn't said often enough. I'd rather have had a patch, though, and a slightly more temperate approach.

Hugo

Perl can be used by some people to write crappy code, Moscow in flames. Film at 11.

I'd probably prefer it if you had posted a small snippet and a link, since this way you are creating a new home for something that already has a perfectly good home. But that's a trivial point.

The post reminds me of Star Wars ep3 in a way -- watching it, I enjoyed the story that I could easily sense behind the obscuring cloud of miserable acting and obnoxious special effects. It wasn't a brilliant story, or an original or particularly nonobvious one, but I found it to be a rather good yarn nonetheless. The post is the same way, although its qualities are covered up by the results of letting loose a satisfying tirade rather than mere incompetent directing.

If you want to get something out of it, read it as if you were overhearing a disgruntled user of your software bitching to a colleague. Hopefully you've all had something of that experience before -- the user is so frustrated that she'll go way overboard in her abuse of your product, but buried within the stream of obscenities and exaggerated complaints will be a rich vein of suggestions and ideas for improvements. You'd never be able to intentionally dig as much from your usual user base no matter much you beg for suggestions. You just have to set aside your pride long enough to detect the value.

I can't say I have anywhere near the visceral reaction to the quoted code that the post's author did, but I do still think he's making some good points. I've written code similar to (and worse than) either of the snippets above. I like to think that I've improved. There's nothing wrong with the open or die idiom, but chaining on an and clause does bother me. That said, I've read code that used it heavily, and after you see it enough, it starts to make a certain kind of sense. (I'd still avoid it for maintenance reasons.)

The second code snippet is harder to defend. It's a good example of a fine idea taken to a disastrous extreme. The author is making a more general point, though: bad Perl code leads to a bad reputation for everything Perl. That resonates with me. At my current company, I've witnessed the negative reactions of far too many people who couldn't possibly have an informed opinion about the language. Marketing people, CEO's, vice presidents, etc.

I also think that many of the obvious places to look for examples of cood coding (eg the most heavily used modules) are anything but. Perl feels like it's resting on shakier and shakier foundations, as the old bedrock modules and toolkits get further and further behind the times. The newer modules and frameworks don't really seem to be keeping pace with the larger community, and are more often than not imitations of or eclipsed by those from other environments. (Where's our Ruby on Rails? PyGame? PHP?) It doesn't feel so much like the Perl community is tired or anemic or stupid; it's more that the richness of CPAN is becoming a liability. Consider a hypothetical lurking genius who previously would have created some great new package. Today, he finds so many existing modules that cover part of the territory he wanted to stake out that he gets demoralized and doesn't bother. Or perhaps his initial user base is too small because his users can't distinguish between the alternatives. Or he's conscientious and researches all the alternatives before beginning and chooses to contribute to the most similar candidate, but belatedly discovers that the original author's vision is very different from his, and that the initial similarity was more by coincidence than design. Or maybe they're well aligned, but he just can't muster up enough enthusiasm for being just one of the contributors of someone else's baby.

Whoa there. Sorry, that was quite a tangent. I guess I'd better stop now so you can get on with the downvoting.

I seem to have problem with communicating this way (written english) and somehow I dont realy say what I ment so people get bad judgment about that - somehow they dig up the worst out of it. This isnt the first time, not just here, devshed forums also ...

Anyway I admire the way you said it. And especialy agree with :

The author is making a more general point, though: bad Perl code leads to a bad reputation for everything Perl. That resonates with me. At my current company, I've witnessed the negative reactions of far too many people who couldn't possibly have an informed opinion about the language. Marketing people, CEO's, vice presidents, etc.

it sounds like someone who hasn't really worked with Perl as anything more than a scripting language

It doesn't sound that way at all to me. It sounds like someone who has gotten into arguments with his co-workers about their excessive uses of $_ and other idiomatic perl that makes their code hard to read. And as others just pointed out to me, the code he's criticizing is part of the core perl distribution.

Just to nitpick slightly, how exactly does "idiomatic perl" make code hard to read? Isn't the definition of idomatic along the lines of "something that is easily understood"?

Let it rest. We all have our opinion of various languages and communities. And we'll stick to them, no matter what.

Yes I'm obviosly new around here, and yes, I still havent seen/understanded all that PerlMonks have to offer.

So? What does that mean? That I shouldnt post, ask questions, whatever?

Once I readed somewhere that no question is stupid question ...

Think about this way. If the post you quoted had been posted directly to PM, it would have gotten downvoted to hell and back, irrespective of whatever grains of truth it may contain. The downvotes would be in response to the tone and the complete lack of anything that could be regarded as constructive or even original.

So why should a quotation get any better response? More specifically, why should quoting something circumvent PM's scheme to keep its content useful?

If I really wanted to know other opinions about the rant in question, I guess I would have posted some brief notice in CB, along the lines "Hey, what do you folks think about this rant?", and provided a link to it. I'm sure that this alone would have gotten me plenty of useful perspectives.

the lowliest monk

Whoa. That makes me a bit sad.

---

Jeff japhy Pinyan, P.L., P.M., P.O.D, X.S.: Perl, regex, and perl hacker
How can we ever be the sold short or the cheated, we who for every service have long ago been overpaid? ~~ Meister Eckhart

That post is relating to Dan's work as a developer, not a user. Perhaps he's abandoning Perl and Parrot use altogether, but the impression I get is that he is simply not going to be part of the development anymore (for now).

FYI, Dan is indeed still a user of both. His current work involves porting another language to Parrot and he was at a recent Boston.PM. See also more recent thoughts on parrot.

--
In Bob We Trust, All Others Bring Data.