roy

A few days ago, I posted an entry about how hot Tom Welling was. A little while later, Tatsu posted this ... VITRIOL.

Roy, there is no way that Tom loves you more than he loves me. He looks RIGHT AT ME every week. I mean he knows that I DVR Smallville so that I can catch it even when I teach classes (Spiritual or ass kicking) and he loves me for it. You are a faithless bitchman. I am his true obsession and he is mine.

YOU AIN'T GOING THERE. NUH UH... YOU AIN'T GOING THERE!!!!!!!!

I decided that the best way to deal with this was to post some pictures ... me and Tom have been good buddies for a while now (kinda like Alexander and Hephaestion!) ... Tom, I'm sorry I have to post these pictures and ruin your privacy, but I cannot stand by and let these allegations ruin my life.

This is right before our first kiss

Well here we are just hanging out

We were having a picnic ... look how he has his arms around me lovingly!!! LOVINGLY!!!

This is us chilling out in chem lab. He's superman so he doesn't need eye goggles; I err on the side of caution.

I hope this dispels all rumors and myths regarding me and Tom Welling.

. . .

Because I know this entry will be read heavily, I will also throw in useless Tabulas work-related stuff so that no one is angry that I've posted a "useless" entry ... muhahaha.

One of the hardest parts of the templating engine part of Tabulas (the output; what your webpage displays) is in (1) remembering all the data that has to be called to output your data. If you look at the complexities of simply templating alone ... it's quite complex. Then I also have the fun task of dealing with entry templating as well.

Previously, I had compartamentalized a lot of functions iinstead of having 'super functions' that did everything. This makes sense ... but one of the problems was that I was being very very inefficient with hitting the DB. One of the technical goals for the new Tabulas is to minimize hitting the DB on the output engine ... I would estimate that on any given Tabulas homepage, there were roughly 50-60 SQL queries (a lot of them totally inefficient as well).

So now all functions call from the same set of "memory-cached" functions ... for example, no longer do I lazily just hit the SQL if I need to grab someone's username and what PDS they're on ... now I look in memory to see if that data has already been extracted from the DB; if it's already in memory, I just grab it there.

Totally more efficient. And now that I'm creating that abstraction layer between DB and memory, all the different libraries (commenting, entry, templating) are all using these in-memory caches ...