I sort of apologize for the relative quiet on the blog the last week or so. I’ve been off geeking my way through a project, and it’s borrowed most of my free time attention.
What I’ve been working on is a way to organize the data stored within my blog’s “For your Consideration” areas — lots of reviews and recommendations, but it’s a blob of data that really isn’t very useful once the initial “post to blog so the readers see it” time passes. I’ve felt for a while that better organization to make that material browsable would give it more visibility, and since it ties into amazon’s affiliate marketing, might generate some incremental revenue for the site.
(quick digression: I’ve done some minor testing over the years, and my audience seems best oriented towards Amazon or similar type advertising; I have always kept it low-key and intend to continue that, but it’d be nice if this site ultimately paid for itself, which it currently doesn’t. that’s my problem to solve, not yours. But, it goes without saying, if you like the kind of stuff I write here, and wish I’d write more, then go down to the footer and go buy yourself something on Amazon through the link. It costs you nothing, and amazon chips a few bucks into the pot for me).
I love all aspects of building sites, but if I could only do one aspect, the structure and organization the most. Figuring out how to tie everything into a usable format is both non-trivial and a lot of fun. That’s probably why I enjoyed DBA work so much, back in the day. I’ve found working out these kind of structural issues is a natural for early and quick prototyping — you have some idea how it should look, then you take your data and start pulling it together, or at least a subset of it. As you do, problems in your approach show up early (and usually, often), and you can solve it as you go along, until you either find it working, or it collapses in a heap and you tear it all apart and go back to the drawing board.
My ultimate goal is to turn “For your Consideration” into its own site. It was obvious early on that the best solution would to to use a database back end and spit the site out on demand, but I wanted to avoid building that until I knew it was worth the investment. Because of that, I’ve been exploring whether I could build it as a static site in a way I liked enough I could push public and explore how people reacted to it (or whether they ignored it, or laughed at it).
What the last few days of exploring the manual prototype has confirmed is that doing this manually doesn’t scale, even to a small scale experiment. I’m not disappointed; I had a feeling that might be the case going in. I could mitigate that somewhat with templates, but even so, I don’t see a “simple” (i.e. manual) solution working beyond 15-25 records. So if I’m going to move this forward, it’s by writing a real backend system, or coming up with another approach.
That’s the fun of doing this early prototyping; I spent a week or so of evenings exploring this. It taught me a lot about what worked with this data set as well as what didn’t, and very little real time doing the tests. It would have really sucked if I’d spent a month or so on wireframes and page design and chrome and the look and feel, only to get a week into actual implementation and hitting that “oops, we have a problem” point.
So now this one goes back on the back burner while I decide what my next steps are, or until I decide I want to go ahead and build the back end. It’s actually a fun project, but I’m not yet convinced I want to invest the time — because remember, it’s not just writing the code, it’s administering and maintaining the thing once it launches, too. Is it worth it?
Probably. But I have to think that one threw a bit…