About a week ago, Pat Eyler, who writes the "On Ruby" blog contacted me to ask if I would be willing to participate in one of his "Questions Five Ways" theme articles.
Pat picks five people who he thinks would make an interesting mix, throws a question at them to start a discussion, then sits back and lets the exchanges fly, then weaves them into an article.
So I had a very enjoyable several days discussing the topic with Steve Yegge, James Edward Gray II,, Diomedis Spinellis, and Don Stewart.
The result can be seen here.
I just ran across a reference to this article by Alex Sandler, on how C++ implements "object-oriented" concepts.
It's a more detailed, and probably more recently researched, coverage of a topic I briefly covered in my RubyConf 2008 talk. If you understand this stuff, you have an appreciation why a compile-time static typed, run-time weakly typed language like C++, as compared to a run-time typed language like Ruby or Smalltalk, makes it crucial to avoid tricking the compiler into thinking that an object is of the wrong type.
I pushed version 0.0.6 of RiCal to GitHub. So far it seems to be getting a fair bit of interest. To-date 52 github users are following the project, and 7 have forked it, several of whom have send pull requests for patches and or additional specs when they discover bugs.
I should also mention that I've set up a google group for discussing the library. Anyone can join, but I've got it set so that I need to moderate a new users initial posting.
Run>Code>Run is also keeping an watchful eye and doing continuous integration testing of pushes to github. So far once Run>Code>Run installed the tzinfo gem all the commits have successfully tested.
And finally, Mirko Stocker, who writes for InfoQ, contacted me shortly after I made my low-key announcement of RiCal, and asked me if I would do an email interview. You can see the results of that request here.
This started out with a question on the RSpec group about why debugger invocations in Cucumber steps seemed to be ignored.
Now, I didn't know the answer to that, but I did have a bit of related knowledge since one of my contributions to RSpec was adding the --debugger/-u option to the spec command as an analog to the same option in the script/server command in Rails. As far as I know that option hasn't been added to Cucumber.
Which led me to refreshing myself on how Rails implements that option in order to help answer the Cucumber question. So I brought up the rails gem in textmate on the latest version of Rails, and had a minor epiphany.
Well, "a day or two" turned out to be a little more like a week, and the official release is still a bit in the future, but I've just published an unofficial version of ri_cal on github.
I've been working with Adam who is using github for a Rails app. He's been feeding me bugs to fix and features to add. Most of the work in the past week has been on developing the "DSL" for creating calendars and events programatically. There are a few loose ends to tie up but if you are adventurous, you might want to play with the library. There will likely be some api changes before the initial release.
When I'm ready to make the "official" release, I'll publish it to RubyForge as well as to git hub, and the version will be bumped to 0.x.0 once I've decided what x should be.
RiCal represents a considerable effort over the past several months. I hope someone besides Adam finds it useful.