This latest rewrite has been coming for a while, but things haven't lined up until now to start making progress on it. The current incarnation of the engine runs this website in addition to several others, and it works fine, but as with all programming projects, it becomes clear how things could be better and that they should be so. The new rewrite will be initially written in Python, taking the best from the current PHP and Objective-C versions, which are the two most mature versions currently in service. Each has gotten tweaks over the last couple of years that ended up being unique to them, but which all versions should benefit from. [Read more…]

Somehow beer drinking led to expanding the small list of books that I'm currently reading into its own expanded page. Not a big whoop, but could be interesting. Should I attach a discussion page to each book? Might do that soon. At any rate, it includes comments on each book, so could be trivially transformed into a basic book discussion page. Perhaps tomorrow. It's true that I need to clean up the alignment on the page. The Unibroue has me happy and sleepy. (Such a good Canadian brewery.)

Update : The books now all their own forum pages, as with this example.

Maybe because my custom blog system flies (way) under the radar of spambot standard practices, but I've been checking each item flagged as spam by my spam checker for the last couple of months and finally going to have it no longer report back to me about it, since it has a 100% success record on distinguishing spam from real submissions during that time. As a side note, if I ever met someone who participated in spamming, I'd certainly try to deck them.

Just wondering if this could help solve spamming problems, but what if webpages didn't have form tags at all when they loaded, and only got them at a user's request to post something. Dynamically, using AJAX, so the form tag is never available on a generic page load. I would presume that spambots would ignore the page as just a static webpage. I'm amazed at the level of spam that hits any page with a form on it. It might not help for pages which have already become known to the bots, which don't seem to re-scan pages that they already know respond to form posts. So, any chance this could work?

Ok, so it seems that browsers from 1995 don't view this site well at all. Mosiac 1.0.3, from 1994, doesn't even know what to do with URLs that don't start with "www," so can't even see these blog pages. (It seems in the later half of '95 they started to deal with them.) [Read more…]

Since the server went from Panther to Tiger, and much of the core stuff got upgraded as well (PHP 4 -> 5, MySQL 3.x -> 4.x, Apache 1.x -> 2.x, Python stuff, fresh new TeX install, etc etc), many tiny things have been acting weird and I've slowly been finding them and cleaning the place up. In particular the RSS feeds got messy because the date formatting between PHP and MySQL changed, and the auto-TeX rendering still has a missing link in the chain somewhere. Just takes time to find it all. [Read more…]

Ok, not exactly direct, but close. Up until this blog entry, I've been using the CocoaMySQL-SBG* program to directly enter the blog items in order to concentrate on pulling data out of the database. So today I added a page to both put in new entries and edit existing ones. [Read more…]