Oloh is a a really cool site that analyzes the source code from thousands of open source projects. This is not just a static analysis either. They track every checkin and know how many new lines of code are added over time, in what language and who added them. This enormous database lets them see some interesting trends, like these from their PHP Eats Rails for Breakfast article:


Best of all, this fantastic resource was written in Ruby on Rails!


To add insult to insult, the article link now just displays "Application error (Rails)". Sigh...
The really entertaining part is clicking on the links in this article and getting "Application Error (Rails)".
Now it says "Ohloh is temporarily offline for maintenance. We'll be back shortly." which is a little better.
Something that article doesn't mention, but the commenters do, is that Oloh seems to be including forks of large PHP code bases. I also didn't see anybody mention that maybe Ruby just doesn't require nearly as many lines of code compared to PHP. I'd say that those charts are an incomplete analysis of the true situation.
Is not correct comparing roug numbers about lines of code (LOC)... a feature could be implemented in 100 LOC in PHP and 10 LOc in Ruby and 5 LOC in whatever-else....
Surely, Ohloh is a great tool...but incomplete.
(IMHO)
I'd see the second chart (and the site's choice of implementation language/framework) as more significant. Unless I read it incorrectly, it seems to be telling me that Ruby projects have been three times more numerous than any of the other languages mentioned, or to put it another way, more Ruby projects were started than PHP, Python and Perl put together. I suspect the charts are also telling us that Perl is heading back to the die-hard *nix sysadmin world that it came from. Good news: if I never have to figure out another line of author-only Perl it'll be too soon...
You always have to be careful with statistics, they can just as easily be misleading as enlightening.
I would tend to agree with the other commenters here that lines of PHP code vs. lines of Ruby code may easily be misleading. On the other hand, while I tend to put more stock in the number of new Ruby projects, this too could be a bit misleading. Ruby's popularity is fairly new and I'd think many newcomers would want to fill in their favorite holes in available Ruby libs.
Project complexity and status (0.alpha or version 7.4) should also be taken into account -- MediaWiki and Bricolage are huge and actively maintained a mod_perl hello-world may be instructive and elegant, but not really comparable.
Also some languages might be more amenable to interleaving various libraries (as in Unix pipes/filters). It would be nice to see some statistics demonstrating such reusability -- though honestly I can't visualize how you'd graph that.
Of course, it might also be argued that Ruby on Rails strength is the ability to write very specialized applications so that many of the PHP projects could be implemented in such a small amount of Ruby code that no one would ever feel a need to actually start a project.
Back in October 21, chromatic commented on this project and article. The URL is here
http://www.oreillynet.com/onlamp/blog/2006/10/lies_sloc_lies_and_advocacy_st.html
To me, the two graphs indicate that it takes WAY more PHP to get the same amount of work done. The number of ruby projects has exploded, but there's only a small increase in the percentage of the total lines of all code. If I were looking to reduce maintenance and harness more productivity, to me it's obvious which platform is best...
image return superman image return superman [link=http://islandcrete.com/hotcms_filestorage/12/comment-290.html]image return superman[/link] mexican gangs mexican gangs [link=http://tattifattoria.com/prova_php/foto/001/topic501.htm]mexican gangs[/link] maple grove campground pa maple grove campground pa [link=http://artecorrente.it/images/user_images/thumbs/text-2174.html]maple grove campground pa[/link]