That darned buy-or-build decision
Well, my blog is back, finally. It took a couple of months for me to fix it, but actually only about ten minutes.
It went like this.
I used to be a huge fan of TWiki, which was sort of an open source, enterprise-class wiki system with a nice plug-in architecture and the ability to set some really fine-grained access control. On the other hand, it had some security issues and about a year ago, some things happened, and there was a split in the developer community, and bottom line all the cool kids are working with Foswiki now.
Foswiki started with the TWiki code base but since then has really improved the security situation and now has a very stable production release. I love it, and I use it all the time to host client documentation and the like.
Now of course it made sense to host my blog on my own Foswiki server. Foswiki has a blog add-on, and there’s a comment plugin, and with a flip of a few configuration switches I had it more or less working. There was some tinkering around the edges with a bunch of template files, but… okay, it took a day or two. Still not too bad. And I had to fix up a thing or two in the underlying Perl code. No big deal.
Then the spammers came.
Spam comments arrived, by the hour, in dozens. They were obviously quite repetitive, and it wasn’t too hard at first to write a little Perl script to delete them quickly. That got overwhelming though, so I hacked the comments plugin to save everything in a review queue. But then I had to pore through that, which took something like an hour a day, to find the legitimate comments.
And I could have downloaded and enabled the antispam plugin, but… oh look, do I want to add blog software maintenance to everything else I do?
The question was answered for me–as I find, about two months later, still not getting around to it. I’d simply taken the blog offline completely because comment spam was killing the server.
This past week…
…drsweetie asked me to set up a blog for a developmental writing class this semester. I poked around wordpress.com a little bit, and decided it would work well enough for dr’s purposes. About half an hour later, I had a login and a blog and a theme and a URL. Easy! The students are working with it right now with no worries.
And no spam. WordPress blocks it automatically.
Have you ever had that happen in your own projects? You get all puffed up to make something work, do your own integration, do it right, make it perfect, make it your own? And then it drags out for months longer than just grabbing something off the shelf would have?
You know, I might get around to re-implementing that Foswiki solution, when the blog add-on matures a little more and I understand its templating functionality. When I understand the antispam mechanism too.
But for now? It doesn’t count as a “buy” decision exactly, because hosting on wordpress.com is free, but starting with the end in mind leads to something quick and reliable that can be deployed right away–instead of something with unlimited options that takes a while–time I don’t have!–to get right.
I felt bad about not eating my own dogfood on this decision. I like Foswiki, I use it all the time, and I encourage others to use it. It seemed like a little betrayal to switch to something so quick and easy and prebuilt. Then I noticed: http://blog.foswiki.org/ is WordPress! Okay, I’m off the hook for now.
What are your experiences with “buy versus build” decisions?
When have you regretted implementing your own solutions when something off-the-shelf would have worked? Or have you had the opposite experience?