walterbyrd <[EMAIL PROTECTED]> wrote: > On May 18, 10:24 pm, [EMAIL PROTECTED] (Alex Martelli) wrote: > > > > > I think that Ruby, which roughly speaking sits somewhere between Python > > and Perl, is closer to Python than Perl is. > > I don't know much about Ruby, but it does not seem to be commonly used > for anything other than web-development. It may be that Ruby could be > used for other purposes, but I don't seem to see it happen much. > > I know that PHP can used at the command line, and could be used for > the same sort of sys-admin tasks for which, Perl and Python are often > used, but I don't seem to see that happening either. > > I'm not sure if Ruby, or PHP, are as general purpose as Perl or Python.
PHP was definitely born _for_ webpages; Ruby wasn't, just like Perl or Python weren't, it just became very popular for webpages when Rails appeared. so i tried search queries for [ intitle:X intitle:Y ] where X is each of various languages and Y one of two words connected with non-web traditional application programming, and these are the number of hits I saw: Y==scientific: perl 334 python 703 ruby 452 php 423 java 2370 c++ 3340 fortran 3200 Y==payroll: perl 81 python 1 ruby 8 php 176 java 382 c++ 101 fortran 1 >From these numbers it would seem that Ruby (and PHP) aren't really more web-specific than Perl (and Python). In general, these days, when you're interested in how popular each of a certain set of technologies / subjects is, search engines can come in handy (with due precautions, of course: for example, "php" appears in SO many web pages (2.5 billion, vs e.g. 282 million for java) that you need to restrict it cleverly (e.g., I used the intitle: attribute -- that gives 20.1 million for php vs 21.4 million for Java, and biases numbers towards pages that in some sense are "about" that technology rather than just happening to mention it as an aside, a part of an URL, etc:-). "c" is very hard to search for because many occurrences of that single letter have nothing to do with the language (here you may try quoted sentences such as "c programming": without precautions, "c" has 2.86 billion hits, but "c programming" 1.22 million vs 1.09 million for "java programming", which again puts things in better perspective). I say "technology" because this doesn't apply to just programming languages: e.g., "COM", a classic Microsoft technology for component interconnection, is swamped by the homonimous occurrence of ".com" in URLs, so you need the intitle: trick or something like that. An interesting alternative would be to use search engines which do semantic annotation, but I'm not very familiar with those, myself; I'd be interested in details if anybody does try that. Anyway, if you're interested in popularity issues, I believe these techniques, for all their defects, will work better than asking a few people or trying to generalize from unsystematic observations. Alex -- http://mail.python.org/mailman/listinfo/python-list
