The monstrous upgrade
As of a few hours ago, eVirtus!net is on running a brand new version of Spark CMS! After more than 8 months of work I froze version 0.3.5 on build 200 and uploaded it to the servers. Internally dubbed "the monster" (due to the fact that it accumulates more than 8 months worth of work), it represents a major milestone in the project; The frontend features are mostly completed.
By mostly completed I mean that the complete framework and (most of) the required feautures to actually generate a viewable website from the data stored in the db is ready. At this point, any missing functionality really cannot be completed before the backend closer to completion. As with most other software products though, there's probably several things that becomes apparent needs changing over time, but for all intents and purposes Spark CMS's framework should be able to facilitate the functionality I want to implement (at least at this stage). The final test of this was the implementation of the server status monitor on this site, a feature that really incorporates a wide variety of functionality (external gets, XML parsing, db updates, task automation, cache-control, and several others).
With that information you might be wondering why there's several apparent short-comings on this site. Well, the simple answer is that eVirtus!net is currently running with several features disabled. Mostly because I haven't focused on the site and its content yet, but also because the smaller numbers are easier to work with and makes it easier have a complete overview and identify problems (I manually read through all log files!). Some features are also intentionally hidden due to the fact that I don't want to give away everything just yet... ;-)
Short overview of the new stuff
A major portion of the development time on this version has been spent on the feedback system (for comments etc), with the exception of a couple administrative functions this is all done. Another major new feature (although basic in nature) is the addition of a general documents format, the result being a product that can be used for more than posting blog entries. I've also spent some time creating a bot for retrieval of external content (feeds etc), and some initial methods of actually utilizing that data.
A number of bugs and minor glitches have been identified and eliminated as well, most of which either relates to time (see below) or standards (XHTML 1.1 Strict) compliance in one way or another. Most notably server side caching now actually works as intended, not just because of some lucky coincidence with time offsets on the web server. A couple of response header bugs have been fixed as well, so now (working) expiration and content-type headers are actually being served for pre-generated content as well.
Although it doesn't mean anything in itself, the size of the codebase have grown from 137kb to 223kb (62% growth). It's my (of course highly subjective ;-) opinion, that the actual gain in functionality is higher the percentage growth in size. This should however give a small indication to that amount of work I've been doing.
I'm about to...
Tomorrow I'm finally starting the work of piecing the backend functionality of Spark CMS together. I actually started working on it earlier today (which, to be correct, have become yesterday), and I must admit it's been a long time since I've been this enthusiastic about working on the CMS. It's nice to see that all the hard work I've put into this actually become something usable. But right now I need some sleep (yes, the time-stamp on this post is correct)...
- on the 29th of September 2005 @ 02:44
- in SparkCMS