Brad Wilson
Brad Wilson was not too impressed with my homemade blog system. He’s basically saying that I sowed the seeds of my own failure by hobbling myself with inappropriate tools (XSLT and Make) and impractical constraints (static-only, portable cross-platform, no code).
He’s right, I think. It probably would be much easier to just write some code. But even though I proved it to myself, I still find it disappointing that XML and the webaren’t to a point where we can count on low-tech “glue” to build something so seemingly simple as a weblog. The whole point of XML is to lower the bar for people who want to build systems that work together, and you would think that a weblog would be a model use case.
The problem with code, as I see it, is that there aren’t many people who can write code. The ability to understand even the basics of programming seems to be an innate characteristic that some people have, and most others don’t.And in most cases, if a person doesn’t “have it”, they aren’t going to “get it”, no matter how much training they get. I’veseenplenty of people who sat through all of the relevant CS courses and give all indications of knowing the material and concepts flawlessly, but just don’t get it when it comes time to write some really simple code. And many of the best programmers I know have come from backgrounds that had nothing to do with computers, and when they started to write code it just clicked for them.
This is not to saythat the cognitive abilities required of programming are somehow “superior”. In fact, this is where many of the Linux advocates go wrong. To say that providing source code is the same as providing freedom is tantamount to saying that people who can’t or don’t write code are people who are inferior and don’t matter. This philosophy assumes that anyone with half a brain can write code, and if someone for some reason can’t, they should just employ the friendly neighborhood coder.
My personal philosophy is that the highest virtue a coder can have is to always strive to make coders irrelevant. Writing code is not an end in itself, it is a means to an end, and an incrediblyexpensive one at that. The more ends we can achieve without writing new code, the better society is. The more ends we make dependent on writing new code, the worse society is.
Of course, I can write a blog without writing new code, so in the case of blogging alone my disappointment would seem unfounded. But I think of blogging as just one application of some really generalized concepts — separating content from presentation, aggregating metadata, and publishing to the web. It should be possible to perform all of these tasks with a simple set of tools that are general-purpose and not specific to blogging. Or to put it another way, if you can’t even make a blog without writing procedural code, XML has a long way to go.
Brad says another thing that one of my co-workers, Dare Obasanjo, also challenged me on. He’s wondering why the obsession with static-only sites. It used to be that ISP hosting was significantly cheaper with static-only, or else you ended up having to write code to use whichever stale old version ofPerl or PHPyour ISP installed. On Windows hosting, it used to be all but impossible to get your own component libraries installed, since legacy COM required registration of components in the machine’s registry. Most providers still do not offer ASP.NET or JSP in their default entry-level packages, but things have gotten much better. So I suppose it is a bit silly for me to see dynamic content as still being a show-stopper barrier to entry.
But I still have some gut-level aversion to dynamic code for blogs. Inmy mind, a blog is nothing more than a metadata stream, with some default visual formatting applied by the author. Heck, blogs shouldn’t even require HTTP, as far as I’m concerned. Every single thing about the blog, except for the actual content of the posts, should be configurable by the user without requiring the author to do anything at all.