TIOBE is worthless and everyone knows it. It's nothing but a count of Google hits. If you want to actually look at jobs, try these graphs from indeed.com. As you can see, both Perl and Python continue to grow, but Python has a long way to go before it's in the same ballpark for number of jobs.
More developers know Python than Perl? Dream on. This might be true for Java, but certainly not for Python. If you compare the number of jobs out there for Python and Perl, it becomes pretty obvious that Python still operates in a small niche by comparison.
And I have no sympathy for coders who try to be cool by writing obscure or complex code either. Those people are BAD programmers. Perl is not the problem there.
No, it doesn't. You just don't know perl's tools. For example, the "scriptlet" thing you mention has nothing to do with perl -- it's a feature of some specific framework you chose. Perl's Template Toolkit is very similar to Django templates, and perl has great ORM tools like Rose::DB::Object.
Did you notice the explanation of how they create the "TIOBE Index"? They run some Google searches and count the results. That's it. I wouldn't take it too seriously. And this person is asking about Ruby, which is miles below Perl on this index. In terms of actual jobs, this chart seems to show that Perl is doing quite a bit better than these other dynamic languages.
I suspect Perl's strength in the job market has to do with what you mention as a weakness: it's association with things beyond just web development.
Perl is Amazon's main web development language now, and their last few stores (clothing, for one) are entirely perl. Yahoo uses lots of perl, python, and PHP. You can't always tell from their URLs, but they have discussed this at conferences. TicketMaster is all perl, and it is in the top three retailers on the web in terms of revenue (along with Amazon and Dell).
I have friends who work for AOL. They do the hardcore stuff in C, and a lot of other stuff in Java. Tcl is used for certain large web properties of theirs, but certain not for "everything."
All of these scale well enough that for most sites you won't care, and none of them really scale well enough to support something like cnn.com.[...]If you want something that will scale to absolutely huge levels (like the aforementioned CNN), J2EE is about the only game in town.
Yahoo uses PHP and Perl. Amazon uses Perl. TicketMaster uses Perl. These are some of the biggest sites in existence. The only truly large site using Java is eBay, and they are not using all the standard J2EE stuff. And I can't believe you are claiming that Tcl is more maintainable than Perl.
Operator overloading is a very rarely used feature. I've never seen anyone use it for anything more than improving the string representation of objects. Good code doesn't do things like that when they aren't necessary.
There's no point in talking about all the ways people could intentionally obscure their code. Good coders who intend to write readable code have no problem doing so in Perl.
You're fooling yourself if you believe that PHP is going to be inherently easier to read and maintain than well-written Perl. Bad PHP code and bad Perl code are both awful to look at.
Now I won't need to bother getting opinions from multiple sources, since there will only be one source. Thanks FCC!
Re:You can fry an egg on my head right now...
on
CPAN Shifts Focus
·
· Score: 1
I think you should learn more about languages and their capabilities before you start slandering them. I'm programming in Java again these days, and have built an e-commerce site for a large comapny with it. It's definitely much better now than the last time I used it. However, in my previous job I worked on a very successful commercial website written in Perl, and it was a nice, well-designed object-oriented system with excellent performance. It had a multi-tier structure with clean separation of the model, view, and controller layers. The templating system was far superior to JSP, even with add-ons like Struts. It was developed by a mid-sized team in a relatively short time. What's more, it wasn't very hard.
You could certainly do the same thing in Java, but it would not necessarilly be faster or better in any clear way. Most of the performance issues on Slashdot are probably from the database interaction. Some clever caching helps in those situations, but the particular language in use doesn't make much difference.
There is a great deal of dissent, even within the Java world, about the use of EJBs. Using them to model data gives relatively poor performance, while using nothing but stateless session beans makes them little more than RMI + JavaBeans with extra overhead. Recently there was an interesting report released from Rice University, which showed that in their benchmarks of a bookstore and an auction application, PHP (yes, a simple scripting language) outperformed Java servlets, and plain Java servlets outperformed EJBs. IBM has released benchmarks saying similar things and advises their consultants to stick with plain servlets when at all possible, and just use JavaBeans for modelling data.
All this to say, don't be so quick to put down programming languages that you haven't done any real projects with. Perl (or Python, or PHP, or whatever) might not be your style, but there's no reason to think it can't power a site like Slashdot just as well as Java can. You can write good or bad code in any language.
Here's another option: instead of wasting hundreds of thousands of dollars on Teamsite, install CVS, rsync, Perl (which, incidentally, is not spelled "PERL"), and Template Toolkit (a set of Perl modules). CVS gives you version control, and does it a lot better than Teamsite. rsync gives you the OpenDeploy functionality, but more efficiently. (Yes, I know OpenDeploy can theoretically do database stuff too, but you have to do so much customizing to make that happen that you're better off writing your own simple script.) Perl gives you the ability to quickly write CGI scripts to input data into a database or flat file, and Template Toolkit gives you a templating solution vastly superior to the one provided by Teamsite.
If you think this sounds like a lot of work, you should take a hard look at how much work it will take to get Teamsite to do what you want. Teamsite is practically useless out of the box, and you basically program it to do whatever you need. For the same money, you could pay a good programmer to write exactly what you need instead of fighting with Teamsite and going to months of training, and you'll end up with something that meets your needs more closely.
it is often said that 99% of people only use 5% of the features. However, each person uses a different 5%.
It is often said, but no one ever gives any evidence, and I just don't believe it. I suspect that the vast majority of word users all use the same 5%.
Re:I don't think there is a problem
on
Google Juice
·
· Score: 2, Interesting
We don't have some advertising hack sitting behind a desk on Madison Ave. saying "Make it so" and pushing a site to the top of Google.
Well, actually we do. There are networks of advertisers who run this software that generates pages with links that push up the Google rating for whoever is paying for those words. I've seen people use this to push their e-commerce affiliate sites higher in the Google ranks. One of them got his site to come back as the first return for "etoys" -- higher than the actual etoys.com site!
Actually, the video quality of TiVo in it's "normal" quality is pretty bad. Very visible compression artifacts. My cheap VCR kicks its ass. To get decent quality, you have to crank TiVo up to a setting that eats up disk space quickly.
I have NEVER in three years seen one person claim to intend to attack anyone.
Then you weren't looking very hard. I posted a link to a comment containing an attack script in this thread. There were others, like one that ran lynx in a loop. There were also many comments encouraging this kind of behavior. One of the articles in that series mentioned a DoS attack by a group called RTMARKin a positive way.
Also Slashdot rarely points out a company as a bad thing.
You must be joking. The whole "Your Rights Online" section is littered with attacks on various companies. I think many of them are at least partly justified, but there's no question that Slashdot editors often say that specific companies are evil in one way or another.
Blanket comments such as yours is exactly how a whole group gets misunderstood and a reputation it does not deserve.
I never said anything about Slashdot readers as a group. I said that Slashdot editors have not felt any need to speak out against DoS attacks when the attacks were on targets they don't mind seeing DoS'ed, even when the code and techniques for the attacks were being discussed on Slashdot itself.
There are plenty more like that. Some use lynx in a loop, some use Python, some use fancier Perl. There are also lots of comments saying "let's DoS them."
My point was that for any large commercial site the Slashdot Effect is just a blip in traffic. I've been Slashdotted and I've been DoS'ed, and the Slashdot Effect was nothing in comparison.
We're on the side that says information is not a crime, but attacking someone is.
You are on that side, but not everyone is. I've seen stories about companies that Slashdot criticizes fill up with comments along the lines of "I'm DoS'ing them now, and here's the script I'm using." Never heard a word of protest about this from the Slashdot editors before.
How many companies has/. forced out of business with our very own special brand of DoS attack?
Answer: none. The swell of traffic brought on by a link on Slashdot is really no big deal for any large commercial site. Most of the sites that get affected are small, often hobbyist or academic sites. You don't see the New York Times site get slow when Slashdot links there, do you?
I know you were just joking, but some people really overestimate this whole Slashdot Effect thing.
No, that's complete bullshit and let's just stop it right here. Pick up a copy of the book "Peopleware" and read about the origins of this ridiculous management myth. It is just not true.
Slashdot Mirror
TIOBE is worthless and everyone knows it. It's nothing but a count of Google hits. If you want to actually look at jobs, try these graphs from indeed.com. As you can see, both Perl and Python continue to grow, but Python has a long way to go before it's in the same ballpark for number of jobs.
More developers know Python than Perl? Dream on. This might be true for Java, but certainly not for Python. If you compare the number of jobs out there for Python and Perl, it becomes pretty obvious that Python still operates in a small niche by comparison.
And I have no sympathy for coders who try to be cool by writing obscure or complex code either. Those people are BAD programmers. Perl is not the problem there.
Django provides much that perl does not
No, it doesn't. You just don't know perl's tools. For example, the "scriptlet" thing you mention has nothing to do with perl -- it's a feature of some specific framework you chose. Perl's Template Toolkit is very similar to Django templates, and perl has great ORM tools like Rose::DB::Object.
What does "mundacity" mean?
Did you notice the explanation of how they create the "TIOBE Index"? They run some Google searches and count the results. That's it. I wouldn't take it too seriously. And this person is asking about Ruby, which is miles below Perl on this index. In terms of actual jobs, this chart seems to show that Perl is doing quite a bit better than these other dynamic languages.
I suspect Perl's strength in the job market has to do with what you mention as a weakness: it's association with things beyond just web development.
Perl is Amazon's main web development language now, and their last few stores (clothing, for one) are entirely perl. Yahoo uses lots of perl, python, and PHP. You can't always tell from their URLs, but they have discussed this at conferences. TicketMaster is all perl, and it is in the top three retailers on the web in terms of revenue (along with Amazon and Dell).
I have friends who work for AOL. They do the hardcore stuff in C, and a lot of other stuff in Java. Tcl is used for certain large web properties of theirs, but certain not for "everything."
Yahoo uses PHP and Perl. Amazon uses Perl. TicketMaster uses Perl. These are some of the biggest sites in existence. The only truly large site using Java is eBay, and they are not using all the standard J2EE stuff. And I can't believe you are claiming that Tcl is more maintainable than Perl.
There's no point in talking about all the ways people could intentionally obscure their code. Good coders who intend to write readable code have no problem doing so in Perl.
You're fooling yourself if you believe that PHP is going to be inherently easier to read and maintain than well-written Perl. Bad PHP code and bad Perl code are both awful to look at.
Now I won't need to bother getting opinions from multiple sources, since there will only be one source. Thanks FCC!
You could certainly do the same thing in Java, but it would not necessarilly be faster or better in any clear way. Most of the performance issues on Slashdot are probably from the database interaction. Some clever caching helps in those situations, but the particular language in use doesn't make much difference.
There is a great deal of dissent, even within the Java world, about the use of EJBs. Using them to model data gives relatively poor performance, while using nothing but stateless session beans makes them little more than RMI + JavaBeans with extra overhead. Recently there was an interesting report released from Rice University, which showed that in their benchmarks of a bookstore and an auction application, PHP (yes, a simple scripting language) outperformed Java servlets, and plain Java servlets outperformed EJBs. IBM has released benchmarks saying similar things and advises their consultants to stick with plain servlets when at all possible, and just use JavaBeans for modelling data.
All this to say, don't be so quick to put down programming languages that you haven't done any real projects with. Perl (or Python, or PHP, or whatever) might not be your style, but there's no reason to think it can't power a site like Slashdot just as well as Java can. You can write good or bad code in any language.
If you think this sounds like a lot of work, you should take a hard look at how much work it will take to get Teamsite to do what you want. Teamsite is practically useless out of the box, and you basically program it to do whatever you need. For the same money, you could pay a good programmer to write exactly what you need instead of fighting with Teamsite and going to months of training, and you'll end up with something that meets your needs more closely.
It is often said, but no one ever gives any evidence, and I just don't believe it. I suspect that the vast majority of word users all use the same 5%.
Well, actually we do. There are networks of advertisers who run this software that generates pages with links that push up the Google rating for whoever is paying for those words. I've seen people use this to push their e-commerce affiliate sites higher in the Google ranks. One of them got his site to come back as the first return for "etoys" -- higher than the actual etoys.com site!
Actually, the video quality of TiVo in it's "normal" quality is pretty bad. Very visible compression artifacts. My cheap VCR kicks its ass. To get decent quality, you have to crank TiVo up to a setting that eats up disk space quickly.
Then you weren't looking very hard. I posted a link to a comment containing an attack script in this thread. There were others, like one that ran lynx in a loop. There were also many comments encouraging this kind of behavior. One of the articles in that series mentioned a DoS attack by a group called RTMARKin a positive way.
Also Slashdot rarely points out a company as a bad thing.
You must be joking. The whole "Your Rights Online" section is littered with attacks on various companies. I think many of them are at least partly justified, but there's no question that Slashdot editors often say that specific companies are evil in one way or another.
Blanket comments such as yours is exactly how a whole group gets misunderstood and a reputation it does not deserve.
I never said anything about Slashdot readers as a group. I said that Slashdot editors have not felt any need to speak out against DoS attacks when the attacks were on targets they don't mind seeing DoS'ed, even when the code and techniques for the attacks were being discussed on Slashdot itself.
There are plenty more like that. Some use lynx in a loop, some use Python, some use fancier Perl. There are also lots of comments saying "let's DoS them."
My point was that for any large commercial site the Slashdot Effect is just a blip in traffic. I've been Slashdotted and I've been DoS'ed, and the Slashdot Effect was nothing in comparison.
You are on that side, but not everyone is. I've seen stories about companies that Slashdot criticizes fill up with comments along the lines of "I'm DoS'ing them now, and here's the script I'm using." Never heard a word of protest about this from the Slashdot editors before.
Answer: none. The swell of traffic brought on by a link on Slashdot is really no big deal for any large commercial site. Most of the sites that get affected are small, often hobbyist or academic sites. You don't see the New York Times site get slow when Slashdot links there, do you?
I know you were just joking, but some people really overestimate this whole Slashdot Effect thing.
No, that's complete bullshit and let's just stop it right here. Pick up a copy of the book "Peopleware" and read about the origins of this ridiculous management myth. It is just not true.
Since they did release their source, I guess ID won out over Ego.
The Orion server is so nice that Oracle bought it and now sells it as their J2EE server.
Um, Linux is available for free to everyone. Does that make it the defacto standard?