Thu, 30 Mar 2006
Blog Posts Escape from Lab
To prevent unfinished articles from getting out before they are ready, I have a Blosxom plugin that suppresses any article whose name is title.blog if it is accompanied by a title.notyet file. I can put the draft of an article in the .notyet file and then move it to .blog when it's ready to appear, or I can put the draft in .blog with an empty or symbolic-link .notyet alongside it, and then remove the .notyet file when the article is ready.
I also use this for articles that will never be ready. For example, a while back I got the idea that it might be funny if I were to poſt ſome articles with long medial letter 'ſ' as one ſees in Baroque printing and writing. To try out what this would look like I copied one of my articles---the one on Baroque writing ſtyle ſeemed appropriate---and changed the filename from baroque-style.blog to baroque-ftyle.blog. Then I created baroque-ftyle.notyet ſo that nobody would ever ſee my ſtrange experiment.
But somehow in the big shakeup last night, some of the notyet files got lost, and so this post and one other escaped from my laboratory into the outside world. As I explained in an earlier post, I can remove these from my web site, but aggregators like Bloglines.com will continue to display my error. it's tempting to blame this on Blosxom, or on CVS, or on the sysadmin, or something like that, but I think the most likely explanation for this one is that I screwed up all by myself.
The other escapee was an article about the optional second argument in Perl's built-in bless function, which in practice is never omitted. This article was so dull that I abandoned it in the middle, and instead addressed the issue in passing in an article on a different subject.
So if your aggregator is displaying an unfinished and boring article about one-argument bless, or a puzzling article that seems like a repeat of an earlier one, but with all the s's replaced with ſ's, that is why.
These errors might be interesting as meta-information about how I work. I had hoped not to discuss any such meta-issues here, but circumstances seem to have forced me to do it at least a little. One thing I think might be interesting is what my draft articles look like. Some people rough out an article first, then go back and fix it it. I don't do that. I write one paragraph, and then when it's ready, I write another paragraph. My rough drafts look almost the same as my finished product, right up to the point at which they stop abruptly, sometimes in the middle of a sentence.
Another thing you might infer from these errors is that I have a lot of junk sitting around that is probably never going to be used for anything. Since I started the blog, I've written about 85,000 words of articles that were released, and about 20,000 words of .notyet. Some of the .notyet stuff will eventually see the light of day, of course. For example, I have about 2500 words of addenda to this month's posts that are scheduled for release tomorrow, 1000 words about this month's Google queries and the nature of "authority" on the Web, 1000 words about the structure of the real numbers, 1300 words about the Grelling-Nelson paradox, and so on.
But I alſo have that 1100-word experiment about what happens to an article when you use long medial s's everywhere. (Can you believe I actually conſidered doing this in every one of my poſts? It's tempting, but just a little too idioſyncratic, even for me.) I have 500 words about why to attend a colloquium, how to convince everyone there that you're a genius, and what's wrong with education in general. I have 350 words that were at the front of my article about the 20 most important tools that explained in detail why most criticism of Forbes' list would be unfair; it wasted a whole page at the beginning of that article, so I chopped it out, but I couldn't bear to throw it away. I have two thirds of a 3000-word article written about why my brain is so unusual and how I've coped with its peculiar limitations. That one won't come out unless I can convince myself that anyone else will find it more than about ten percent as interesting as I find it.