I'm glad I know a little perl...

[kiat]

Hi monks,

I followed the thread 355190 with interest as I was curious to read the responses from the monks here.

Anyway, I came across an essay (by Paul Graham) today that I found really enlightening. The url is:

http://www.paulgraham.com/javacover.html

Here's a little quote from the essay:

Historically, languages designed for other people to use have been bad: Cobol, PL/I, Pascal, Ada, C++. The good languages have been those that were designed for their own creators: C, Perl, Smalltalk, Lisp.

Perhaps Wassercrats might be interested to have a read.

update I'm surprised that on this very site, you can neither say good nor bad things about perl.

[chromatic]

I'm surprised that on this very site, you can neither say good nor bad things about perl.

If you have a problem with the voting and experience system, please explain it in concrete terms. (If you can provide a reasonable suggestion for fixing it, even better.)

Whining obliquely that you received downvotes (two, as of this writing) is counterproductive. I'd really like to see it stop everywhere.

[calin]

that you received downvotes (two, as of this writing)

Knowing that some uber creature knows who votes what (and doesn't at least keep it to himself) is also frightening. You can't blame people for their imagination running wild...

I'll write some day a node about the climate of grace present here at the monastery.

$chromatic += you_know() ? 1 : -1;

[chromatic]

The gods can see the number of plus and minus votes a node has received. This feature helps us to identify rare cases of voting abuse.

Since the original poster complained about receiving downvotes, I checked the numbers. At that point, the node had several times the number of upvotes as downvotes. In my mind, there was no evidence of abuse.

I have no idea what you mean by the phrase "climate of grace", but if you're concerned about what any one god can and cannot do, any other god can and will look into the matter.

[Nkuvu]

It doesn't take an uber creature to see how the voting has gone so far. Just vote on a node, and *poof*, you see the reputation.

So if you vote on a node and see it's got a -2 rep... well, you can reach the same conclusions I can, I hope.

Note that I'm not saying one way whether that's the case here or not -- chromatic may very well have used godly powers to see the reputation or voting history. Just that knowing a node has at least 2 downvotes does not imply lack of discretion.

[Joost]

update I'm surprised that on this very site, you can neither say good nor bad things about perl.

I like the particular column by Paul Graham you are linking to, but I think you might have recieved a more positive score on this post if you'd kept the tone a little more neutral.

Anyway, I'll give your a ++ vote, just to get people interested in Paul Graham's columns.

[l3nz]

-troll mode on-

I personally will take Java over C any time of the year (and C over C++, but I happen to heartily hate C++). And I frankly don't miss Smalltalk. And I know quite a number of people who used and loved Pascal. So what?

Reading Paul Graham, it looks like he likes language in virtue of their similarity to Lisp. I like what I can do and how much time it takes me, and how expressive a language is. So what?

-troll mode off-

[jepri]

To troll properly, you have to make a general assertion that will tick off most of the community, like "It's obvious to everyone who isn't a bonehead(like most here) that Java is in every way superior to any dinky little hand-rolled language you simpering fools might prefer" It doesn't work properly if you make it clear you're just expressing a personal opinion.

--, please practise your trolling harder

OK, I'm just kidding. Personally, I share your opinion.

I think Paul Graham gets away with his critiques because he makes it totally clear that he is completely partisan, but in a very non-hostile way.

Plus I have a lot of respect for someone whose basic attitude seems to be "I'm having a great party over here all by myself. Come and boogie anytime you want." It's got me looking, which is something no other advocate has ever managed.

[l3nz]

Ok I'll go practice my trolling harder..... :-)

I love reading Paul Graham - his On Lisp book - available freely - is a collection of jewels (written in Lisp, of course...) but you'll have to learn from it in any case, whichever your language of choice might be. Still I don't get convinced to leave my Java for Lisp in real world system building. :-) A 500hp roadster might be fun for the experienced driver, but I'd rather drive a sensible sedan every day going to the mall....

[BUU]

Since we're wandering in to the topic of language debate, personally I can't stand C, I much prefer C++, probably because I'm bitter over using lots of C code that tries to be OO, but with none of the OO features that make OO practical. *mutter.

I see this sentiment a lot though (people hating C++ but liking C), and I still don't understand why. I tried asking why in another node some what similar to this one and got some of my diferences between the two (c and C++) nit picked, but I never really got what I felt was a substantiative answer.

So, why do you hate C++ and prefer C?

[tkil]

So, why do you hate C++ and prefer C?

Personally, I much prefer C++ over C. But I have a few theories as to why others might prefer C:

• It's a smaller, simpler language. K&R 2nd ed. is a slim volume, about 150 pages or so. Stroustrup "Special Edition" is somewhere north of 1000 pages.
• When I transitioned from C to C++, I found myself spending 90% of my time working on header files. While my overall time to a functional solution went down, it did sometimes feel like I wasn't coding as much as I used to under C. In C, it is easier to just start coding, as it does not force you to think about larger-picture issues as often — or, at least, as early — as C++ does.
• C is "closer to the metal". People can mentally map a given chunk of C code to assembler (or at least they think they can; modern compilers and CPUs are liable to do much stranger things to the code than the average programmer).
• The (mostly mythical) issues of efficiency. Using C++ in all its power does inflict overhead, both space and time, at compile time, link time, and runtime. Whether this overhead is noticable, and whether it is recovered by the progress of technology in the amount of time you save by using higher-level languages, is less certain.
• Earlier versions of C++ (both the accepted language definitions and the compilers that supported it those to varying degrees) had issues and shortcomings that could turn people off in a hurry (remember generic.h?). Also, some widely-used frameworks did their best in those limited languages, with painful results (hello MFC). If coders have not dealt with a modern compiler and modern library, they might still be thinking of 1994 C++.
• C++ with proper use of the STL can often leave coders feeling that they are just monkeys connecting components together and never writing anything fun or interesting.
• As you mention, there are a lot of C libraries that try to do OO-ish things, and those models often fail to interact gracefully with the C++ object model. this can cause people to avoid C++ entirely.
• Some of the features of C++ make it difficult to interact gracefully with any libraries (templates, exception handling, RTTI, virtual functions).

As I said above, I personally prefer C++ over C for almost every task where I am forced to use a low-level language. The above are some of the reasons I've heard others give as to why they don't care much for C++, though.

[McMahon]

I knew a C programmer once who liked pure C simply because there's no magic. The programmer is in direct control of every aspect of the program.

I've never written any C, but I used to debug a fair amount of it, and I really appreciated this programmer's code. It was really tight, easy to read (for C), and rarely contained surprisess.

[Anonymous Monk]

It is amazing how much pointless discussion results from one monk's choice to express his surprise and disappointment.