I haven't done a "what's in Chris' brain" posting for a while, so I thought as part of the Ideas of March it was as good a time as any to resurrect it.

I had been a Delicious user and had not used it much except as a dump for links I intended to read later. I often didn't come back but it was nice to have those links around for those times when the coding part of my brain wasn't working and I needed to some more research into topics I had flagged for further study. Then the infamous leaked graphic that showed that Yahoo was planning on sunsetting Delicious started making the rounds. It got me to thinking about the dangers of relying on a free service for long-term archiving of information. Delicious could just go POOF one day, taking all the stuff I found interesting with it. I decided to move my links to Pinboard and have introduced time to go over the links I store once a week into my workflow. I even read up a but on their architecture and found their development philosophy refreshing.

They are using PHP on the front, Perl in the back, with MySQL and S3 as data stores. As they put it, very vanilla. No framework being used, and I'm starting to see that perhaps my framework "obsession" is having a drawback other than me having a mind full of how to do things in 10 different ways. Anyway, feel free to browse through my bookmarks: no secrets in there and I'd be glad to discuss anything you come across. I wanted to go over a few things I recently "pinned" to my brain.

Beyond Frameworks is a presentation that Stuart Herbert gave at the PHP UK Conference this year. He talks about the challenges of trying to prevent the frameworks you build your applications with from swallowing everything up around them. When is the last time you actually upgraded an application written in one version of a framework to work with a newer version? Personally, I always like to keep upgrading the framework for my applications to the latest stable versions and dealing with errors by making sure I have decent test coverage. Others are not so lucky to have the kind of control that I am allowed to exert over the code base. Heck, I don't know how much of a task it would be to update a Zend Framework application I'm using right now at Moontoast. Maybe I should go and look.

Anyway, I want to take a deeper look at the architecture he proposes: keep the framework-specific stuff to a bare minimum and write as many standalone components that can then be shared by other applications in your company. His use of PEAR as the method of organization and installation is just genius: why reinvent the wheel. I watched the presentation once while distracted by other things, I will take a run through it again. Jason Gilmore is the publisher of my CakePHP book, co-founder of the awesome CodeMash conference and a pretty prolific author. He recent blogged about the tools he used to write his latest book. I have a supersekret project I am working on and are using a lot of these same tools at Jason's suggestions. Writing a book was a very interesting project, and I agree that you should not be fighting with the tools you are using. Jason's tool chain has allowed him flexibility in the formats he can release his book in. As someone who just bought a Kindle, I can appreciate how important it is to have ebook options.

Interested in know what some of the PHP 5.3+ frameworks are shaping up to look like? Check out Symfony2, Alloy HMVC and what I consider to be the "first to the post" PHP 5.3+ framework Lithium. I'm not sure what framework I would pick if I was starting a PHP project from scratch, but I do know that I would be avoiding using some of the previous full-stack options available to me. Especially if I want to move into a layered architecture as proposed by Mr. Herbert.

My sister got me an Arduino starter kit for Christmas (still haven't actually received it due to scheduling conflicts as she lives 250 miles away) and I can't wait to hack around with it. It uses it's own little language called Wiring but I found some info on how to use Python to control Arduino devices.

Finally, I've been working on a new version of the web site for the simulation baseball league I've been a part of for 14 seasons and counting and I've been building it using Django and a few other things. I'm also using this project to get myself more familiar with BDD and are using an awesome tool called Lettuce to write my tests. I have a sneaking suspicion that since I could do some Facebook integration testing using Cucumber (I linked to it in a previous post) I might be able to use Lettuce along with some Facebook libraries in Python to do that too. We'll see, but Lettuce has proven to be easy to get up-and-running with.

So, my third Idea of March: share the things you've been bookmarking and tagging for later use. You never know what other people will find useful from them.