Saturday, November 10, 2001
The Quick and Dirty B-Movie Plot Generator
My friend JeffC sent me a link to They Fight Crime, a CGI script that prints out a plot.
I checked the page, and it's all in JavaScript. Pretty easy. It looks easy. So I take the text (it's there for the taking) and write my own version (in C of course) that can be easily extended (all the text are in files) and I've even included character names. Mine also has more options than just fighting crime.
I do need to fix the pronouns in a few cases but I'll save that for later.
Update on Friday, July 6th, 2007
Oh, it only took a few years to fix the problem.
A google spiders
In checking the log files for this site I've notived that Google has finally found it and has spent the past few days spidering through it.
There are a few thousand links for it to follow (out of what? A million potential URLs on this site? I know the Electric King James has over fifteen million URLs). For instance, there are three just for the years, 12 each for each year (okay, so there's only 11 for this year, but close enough) so that's now 39 URLs. Each day (for those days that have an entry) have at least one entry and while I may have skipped a day or two here and there, let's say there's an averave of 300 per year, so that's over 900 there. And if you assume an average of two entries per day (remember, you can retrieve the entire day, or just an entry) that's another 600 per year or 1,800 so we're now up to nearly 3,000 URLs that Google has to crawl through (with lots of duplication).
robots.txt for bible.conman.org
#----------------------------- # Go away---we don't want you # to endlessly spider this # site. #----------------------------- User-agent: * Disallow: /
There's a reason I don't allow web robots/spiders to the Electric King James—it would take way too long to index the site (if indeed, the spider in question was even aware of all the possible URLs) and my machine isn't all that powerful to begin with (it being a 33MHz 486 and all). But I feel that there is a research problem lurking here that some interprising Masters or Ph.D. candidate could tackle: how best to spider a site that allows multiple views per document.
