Post: Re-frameworking
As you have no doubt seen if you’ve visited my site , my site recently underwent a bit of a cosmetic overhaul (i.e. it stopped looking like shite). That out of the way, I’m now contemplating the prospects of another redesign. This one isn’t with the front-end of the blog – I’m happy with that for the time being – but with the software powering it.
As some of you will no doubt know from my incessant whining occasional criticism, I have certain issues with WordPress and the language (PHP) on which it’s built. I won’t go into them here for fear of this turning into another rant, but suffice it to say that I’m not entirely happy with my current choice of CMS.
As far as I can see, I have three choices:
- Stick with WordPress: it works well enough and my time might be better spent working on other projects.
- Switch to another CMS with a sharper focus on web standards, semantic HTML and clean design. Only problem here is that I don’t know what’s good: I’ve been immersed in the world of WordPress for too long.
- Roll my own. This is looking increasingly appealing, especially as it gives me the perfect excuse to play around with a clean language (Python or Ruby, most probably) with a clean framework on top (such as Django or Rails). Problem here is that it would be a pretty mammoth task, even with the help of one of the aforementioned frameworks, especially if I want to keep all the features WordPress offers.
So, thoughts? Which route do you reckon I should take?
I would probably ‘roll my own’, but only because I am a control freak
Perhaps the only major downside is than in getting from Point A to Point B, you often find yourself having to create Points C-Z just to get there 😉
If I could offer one piece of advice, not that you need it, but it would be to manage this like a project with the 4Ds of project management: define, design, deliver and develop.
Defining the scope is crucial, as it is designing the overall approach you will take. From my own experience I know all about ‘function creep’, as mentioned above. You have to limit your scope, whilst designing in such a manner as to keep things open and extensible for the future. Delivering may seem obvious, but try and incorporate ‘gates’ into your thinking, that is: do I continue? I have spent literally years on some projects, wasting time and resources, when really they should have never gone on that long. And finally developing your idea is all about learning from the experience so you capitalise on what works whilst not making the same mistakes in thr future. This is maybe the greatest asset you can take away with you, especially on non-paid projects.
It all seems common sense, and once explained maybe it is, but it’s valuable nonetheless
WordPress is OK — at least you can easily customise if you want to. I used to run Movable Type and found that difficult because I can’t hack Perl at the worst of times. If you threw some energy into improving WordPress (either as part of the project or just for yourself) that would probably be a better use of your time.
I think you’ll need to go into a rant, they’re always fun. What’s so bad about PHP? RoR is a fad and will probably die out soon 😉
Design note: At higher font sizes, your enumerations begin to be printed off the LHS of the page.
Ben, there are a number of things bad about PHP: a type system which is loose to the point of being diarrhoeic, inconsistent function naming and behaviour, various bits of behaviour (references, GET and POST globals) which are inconsistent across different versions. What it all boils down to is that the language was designed sloppily and encourages sloppy design in turn.
To “Your mum”: noted and fixed.
Your allowed tags has decided to become indented again, to roughly the middle. (In IE).
To Jo: fixed, and placed down below the comment box where it’s less obtrusive.
Much nicer
I’ve found ROR to be very comfortable for development. It’s all about what you want out of your weblog though. Currently I am using Textpattern for two main reasons: I can hack up PHP without too much thought and the data structure makes enough sense to roll out something quick for “adding” functionality to my site. WordPress is a lot like this as well. The only reason I didn’t end up with WP at the time was I had this crazy idea of wanting to manage my whole site through the CMS. For the adventurous, http://www.opensourcecms.com/ lets you see the default install and play around with almost every well known PHP/MySQL thing out there. From weblogs to wikis.
If you’re not ready to roll your own, but still want to give the ruby/rails environment a shot, Typo http://www.typosphere.org/ has stepped it up a notch and might be a very good springboard to develop from.
It just looks damn good now.
(not the point of this post I know, but still)
Thank you: much appreciated. Clearly, though, you haven’t taken a look at it in Internet Explorer recently.
For what it’s worth, I plan to keep this design if and when I change the CMS.