My O'Reilly Network weblog reader is complete crap, I discovered, because, now that I am using it, I realize that the number assigned to the weblog is only accidently related to its release date.
To write a weblog, I create an entry, and the O'Reilly Network system records this entry as a "draft". I can work on it all that I like and nobody will see it until I mark it as "final", when it then shows up on the public website.
So, if I want to read all of January 15's new (public) entries, I have to download those marked { 4189, 4194, 4195, 4196, 4198 }. Looking back at January 14's entries, I find { 4151, 4102, 4184, 4185, 4187, 4188 }. Not only are they not in consecutive (my program can deal with gaps), but they are not in order either (which craps up the whole thing).
However, I do not consider this some sort of flaw---I like the way they have made things. It is just more of a challenge to read offline.
I could grab the RSS feed, but that only lists the articles they show on the main page. If I cannot read the weblogs for a week, I am going to miss some. I can choose a monthly archive, but that has no RSS feed. I do not really need an RSS feed because I can just use HTML::SimpleLinkExtor and a regular expression to pull out the URLs that I want.
No matter what I do, I now need to keep track of everything that I have read, not just the last thing that I read, and if I am going to do that, I might as well go back to the first entries, in 2000, where it was just Tim O'Reilly remarking on whatever cool thing he saw that week.
