Preface:

After spending a few months toying around with various prepackaged web systems (the most notable included WordPress, Drupal, and the Play Framework) along with a number of other static blog generators (like Jekyll, Pelican, and PieCrust), I decided to create my own python based website generator using a combination of prepackaged Python modules that included Markdown, SASS, Jinja2, and CoffeeScript.

While one might be skeptical as to why I decided to go for such a radical solution given my available options, I will defend my rationality by saying that my needs were rather unique for a number of reasons.

Rational:

For starters, while I think the overall notion of a prepackaged web framework is a rather sound concept (at least in theory); however, some of these more popular systems are horribly organized monsters that assume server inode simply grow on trees ("for the love of the light!!!" Wordpress, did you truly need to fragment yourself into 2,766 files for a vanilla install?) and if that was not bad enough the API documentation is so horribly organized that a 4 year degree in the internal workings of such systems are required prior to even thinking about templating anything (Oh joy I get to Google what the function "_e" does in your web system woo!!!).

Likewise, with this being said, it is inevitable that some folks in the peanut gallery will call me out as being a cynic on this issue (although I must confess that I am a firm believer in sarcasm); however, in my defense my decision was not limited to this particular attribute alone. To elaborate, because I do most of my development work on a windows machine and the more prevalent static generator tools available were designed to work with the Linux OS, the learning curve to use such web systems does tend to rise substantially upon factoring in the inevitable cross compile compatibility issues that generally arise (noting that I was looking to mitigate such occurrences, although ruby/node has helped mitigate this over the years).

While the difficulties associated with such systems were not really an issue for me (as I tend to figure out most software problems rather quickly) it was more about the easiness to quickly modify the final result along with having to install additional software that was the deal breaker here.

Nevertheless, I will admit that some of these contemporary static systems look easy to work with (particularly if you are running Linux or opt for a ruby/node install), and although someday (I swear) I will completely swap over to Linux (just as soon as the main Linux forums stop answering questions with "I know the one line of terminal code to fix your problem but I am not telling you because you should know this information already!") but alas i needlessly digress...

Now, I will concede that such attributes were only a small part of my decision to roll my own system since (in truth) I wanted a web system that was mostly static with a touch of Ajax/PHP, had a small data profile (permitting the usage of SQLite), inode compressible (I will write an article on this at a later date), utilized Markdown syntax, and could be managed inside of Scrivener.

Concluding Remarks:

Overall, I think the web system is looking rather good, and while I am sure a few tweaks and improvements are needed, I am content with the result and maybe I will release the source on Bitbucket or GitHub when everything is said and done.

Regards!