"... it seems likely that the differentiating factor was libraries, not the core language. ..."
Absolutely the libs make the difference. [0] But are there other forces going on that limit language choice and adoption?
I'm thinking of say Perl. Perl has CPAN, which kicks pythons pants , hands down when it comes to the variety of usable tested modules. Is it the Perl syntax or core language that scares off the hackers?
My own pet theory is that it's not the hackers who ultimately decide the language(s) they use at work. Additional libs don't matter as much as the initial (fad) language the code base is written. It is the companies (and PHB's) that hackers ultimately work who determine language choice. In the end hackers just give up.
One more reason to "start a startup". You can choose the best tool(s) for the job.
[0] A friend of mine, a strictly "ANSI-c" man wondered how I could code the rings off him. Then I showed him the Python libs I could choose from. He had to download the source code, compile it, learn how to use it at 'c' level and then do the job. For example regular expressions.
Maybe the hacker who can produce error free code in Perl does not exist? If even hackers sometimes make programming errors, perhaps most of them were smart enough to eventually figure out they could minimize that problem by moving away from Perl.
Most companies I know actually asked their developers which language they should choose. Typically there is a decision eventually to concentrate on a limited set of languages, but choosing that or those languages is not usually being done without consulting the developers.
"... perhaps most of them were smart enough to eventually figure out they could minimize that problem by moving away from Perl. ..."
It does require dev's have to be better programmers. There are plenty of ways to make Perl code safe. Maybe you have to work harder at it?
"... Typically there is a decision eventually to concentrate on a limited set of languages, but choosing that or those languages is not usually being done without consulting the developers ..."
I agree with the limiting of languages. But I've yet to see languages chosen purely for good technical reasons. [0] And this is one reason I think Lisp gets pushed out.
Libraries are important but code has to be easy to read. You have to be able to easily read the example code for the library you want to use or to read other's code and learn from it. That's what scared me away from Perl and really why I still mainly use Python over Lisp. But each person finds different things easy or hard to comprehend. It mainly depends on your background and what kind of patterns the brain becomes accustomed to.
Comments
"... it seems likely that the differentiating factor was libraries, not the core language. ..."
Absolutely the libs make the difference. [0] But are there other forces going on that limit language choice and adoption?
I'm thinking of say Perl. Perl has CPAN, which kicks pythons pants , hands down when it comes to the variety of usable tested modules. Is it the Perl syntax or core language that scares off the hackers?
My own pet theory is that it's not the hackers who ultimately decide the language(s) they use at work. Additional libs don't matter as much as the initial (fad) language the code base is written. It is the companies (and PHB's) that hackers ultimately work who determine language choice. In the end hackers just give up.
One more reason to "start a startup". You can choose the best tool(s) for the job.
[0] A friend of mine, a strictly "ANSI-c" man wondered how I could code the rings off him. Then I showed him the Python libs I could choose from. He had to download the source code, compile it, learn how to use it at 'c' level and then do the job. For example regular expressions.
Maybe the hacker who can produce error free code in Perl does not exist? If even hackers sometimes make programming errors, perhaps most of them were smart enough to eventually figure out they could minimize that problem by moving away from Perl.
Most companies I know actually asked their developers which language they should choose. Typically there is a decision eventually to concentrate on a limited set of languages, but choosing that or those languages is not usually being done without consulting the developers.
"... perhaps most of them were smart enough to eventually figure out they could minimize that problem by moving away from Perl. ..."
It does require dev's have to be better programmers. There are plenty of ways to make Perl code safe. Maybe you have to work harder at it?
"... Typically there is a decision eventually to concentrate on a limited set of languages, but choosing that or those languages is not usually being done without consulting the developers ..."
I agree with the limiting of languages. But I've yet to see languages chosen purely for good technical reasons. [0] And this is one reason I think Lisp gets pushed out.
[0] PHB's dont like you re-writing code from scratch so you could be forced to work with what ever was previously chosen ~ http://www.joelonsoftware.com/printerFriendly/articles/fog00...
Libraries are important but code has to be easy to read. You have to be able to easily read the example code for the library you want to use or to read other's code and learn from it. That's what scared me away from Perl and really why I still mainly use Python over Lisp. But each person finds different things easy or hard to comprehend. It mainly depends on your background and what kind of patterns the brain becomes accustomed to.