3 ms·

Perl is not going away any time soon. 1,20,000 packages[1] ( compared to Pypi : 32621 [2] and Rubygems : 59,519 [3] )is nothing to scoff at. [1] http://www.cpa

Perl is not going away any time soon. 1,20,000 packages[1] ( compared to Pypi : 32621 [2] and Rubygems : 59,519 [3] )is nothing to scoff at. [1] http://www.cpan.org/ http://www.cpan.org/ [2] https://pypi.python.org/pypi https://pypi.python.org/pypi [3] http://rubygems.org/ http://rubygems.org/

gautamsomani 13y ago
Just curious, are those packages being re-written to make them better and faster (if they can be, that is)?
btilly 13y ago
This is true, BUT there is a law of diminishing returns. The value of a large library scales nonlinearly. The added packages are more likely to be unrelated to whatever you are doing, and so become progressively less likely to be useful. That said a culture of testing all of the time makes the Perl packages less likely to have serious conflicts. Which is a huge value proposition. (For the Ruby fans who jump up to object, do you run unit tests by default every time you go to install gems? Do you have people volunteering lots of machines with different things installed for the purpose of running unit tests on everything in Rubygems? No? Then you don't know what a "culture of testing all the time" is.)
- VLM 13y ago
  "The added packages are more likely to be unrelated to whatever you are doing, and so become progressively less likely to be useful." I've not found that to be a problem in actual development. It just gives the search engine extra work with no impact on my productivity. For example, the existence of "CNC::Cog" has no impact on me, although it gives the search engine extra work when I search for "Astro::FITS". The value of Perl to me, is if its computationally possible, there's probably a module on CPAN to make it easier for me. Therefore its the easiest to use, most powerful language, because I can do the most with the least effort, and time is money. Shrinking CPAN would merely make me move to the currently second most powerful language/library team. Donno, java or ruby or python or something like that. I do agree that in an academic setting, a homework assignment like "here's a list of 10 random Perl modules, now make something interesting out of them" would be painful. However, the programming equivalent of the "Chopped" TV cooking show might be highly amusing for a youtube video series or some kind of competitive multi-player startup game company.

gautamsomani 13y ago

Just curious, are those packages being re-written to make them better and faster (if they can be, that is)?

btilly 13y ago

This is true, BUT there is a law of diminishing returns. The value of a large library scales nonlinearly. The added packages are more likely to be unrelated to whatever you are doing, and so become progressively less likely to be useful. That said a culture of testing all of the time makes the Perl packages less likely to have serious conflicts. Which is a huge value proposition. (For the Ruby fans who jump up to object, do you run unit tests by default every time you go to install gems? Do you have people volunteering lots of machines with different things installed for the purpose of running unit tests on everything in Rubygems? No? Then you don't know what a "culture of testing all the time" is.)

VLM 13y ago
"The added packages are more likely to be unrelated to whatever you are doing, and so become progressively less likely to be useful." I've not found that to be a problem in actual development. It just gives the search engine extra work with no impact on my productivity. For example, the existence of "CNC::Cog" has no impact on me, although it gives the search engine extra work when I search for "Astro::FITS". The value of Perl to me, is if its computationally possible, there's probably a module on CPAN to make it easier for me. Therefore its the easiest to use, most powerful language, because I can do the most with the least effort, and time is money. Shrinking CPAN would merely make me move to the currently second most powerful language/library team. Donno, java or ruby or python or something like that. I do agree that in an academic setting, a homework assignment like "here's a list of 10 random Perl modules, now make something interesting out of them" would be painful. However, the programming equivalent of the "Chopped" TV cooking show might be highly amusing for a youtube video series or some kind of competitive multi-player startup game company.