Content Management with ByteSpring

I have been looking to either create my own CMS or find an open source CMS for use in my church’s web site. I have found several that look promising, but I have consistently found documentation for installation and customization lacking. Then I stumble across Jason Sheedy’s ByteSpring CMS, a simple but surprisingly capable little CMS with great documentation.

Simplicity is key for me as I have little time to train the parish staff. Much like Contribute, which we currently use, simply navigate to the desired page to edit, make your changes, click save and that’s it. No HTML required.

Bytespring CMS is built on ModelGlue and ColdSpring, both of which are provided with Bytespring. Installation was pretty simple. Drop the files into your web root and navigate to home page. I have not tried skinning yet, but it looks like pretty straightforward CSS. I feel confident I could have this thing up and running inside of an hour.

User management is there and it looks like we can expect some expansion on that in the future. There are two user types currently–Administrator and guest. Guest doesn’t appear to do much of anything right now, but I suppose it’s just a matter of time. There currently is no workflow to allow for a content approval process, but as with user management this could be up and coming.

There is no database required at all. Content is stored in XML files in the go2 folder, while user, news, and menu data is stored in a WDDX packet, which means plaintext. This might be a deal breaker for some. For a simple church web site, I don’t think it’s an issue.

Bytespring, simple as it is, offers plenty for the small mom & pop operation wishing for a web presence without the headaches. My testing so far certainly makes it a viable candidate for our relaunch later this year. I’ll be putting other CMS packages through their paces this summer as well, but Bytespring certainly was a good start.

Selenium Test Scripts – To record or write?

So I’ve gotten the Selenium Core installed on a staging web server where we run development code.  Additionally, I have installed the Selenium IDE Firefox extension to record test scripts.  I have been recording a few test scripts on some very basic functionality of an application I have been working on, then uploaded the test suite to a testing directory within my application directory structure.  Then it’s time to run the Selenium TestRunner and access my test suite.  For some reason, I get small failures that while not indicative of a problem with my code, but rather idiosyncracies with Selenium or my JavaScript deficiencies.  I’m getting very minor errors about elements not being found which I’m fairly sure is due to me not being in the habit of assigning an id attribute to everything especially form elements, so that Selenium can find them.  I’m also wondering if it is better to manually write test scripts for Selenium, which sounds too time consuming, or is it OK to just use the Firefox extension to record them.  I’ll do a bit more digging to see what I can find.  Selenium is available at http://openqa.org.