Finally on my way to Wuhu by fast train.

From my research, there’s not much to do here. But a city without much to see also has its special draw since it’s a good chance to get away from the familiar. I plan to spend time working on programming projects, applying to jobs, as well as wandering the city and seeing what Wuhu has to offer.

This is my first time in Anhui Provence, even though it’s extremely close to Nanjing. I’ll have to see what sort of food and local 特产 they have around here.

After arriving though, it turns out that everything is closed. I had trouble finding food for dinner but eventually found a chain dumpling place. I biked around a bit and couldn’t find much interesting besides a food street (凤凰美食街). I’m probably going to stick around the hostel and work on my chantrending project or labyrinth-site. If I miss home, there’s tons of trains going back to Nanjing and it would be simple to get a ticket home whenever I want.

The hostel is nice and simple. It’s a converted 18th story apartment with two rooms of bunk beds. I’m the only one here. The two employees have been sitting in the main room watching Jiangsu Broadcasting dramas, smoking, and eating delivered food.

Chantrending

After sleeping on it, I realized that it was pointless to try cramming all this data into Redis, so I’m going to go with a more familiar PostgreSQL with the following structure. Each board would have the following tables:

• Thread Table: Thread number as key. Array of post numbers as a field.
• Post Table: Post number as key. Date, string comment, string subject, and json blob of all metadata besides image as data.
• Snapshot Table: Unix time stamp, Term, and frequency.

I’m also thinking about having references (quoting other posts) included in this so I can assign more “hotness” to terms that are in posts that are frequently quoted. This could be done by adding a field to the post table that is an array of references, but postgres has no good ways to do sets and I’m not keen on having to search the array for duplicate comments. It might come to json unmarshalling-remarshalling madness. IDK yet.

PostgreSQL 11 can handle more than 2048TB so I’m not worried about the limitations of the software, but more with how much I can store. Might have to take a look at AWS at this point. It’ll help to get an understanding how of much space this will take up.

Another option is Neo4j which is a series of nodes that link to each other with relationships. This has interesting implications such as having a thread which posts are part of, terms that are part of certain posts, and posts that reply to other posts. This could be used to further analyze things!

Website

There’s now a nifty “Previous” and “Next” button for checking out more articles at the bottom of each page! This was really easy to do with hugo’s page functions and how hugo server rebuilds the site and refreshes the page very time somethings is updated.

Things I’m Liking

• Biography of Terry A. Davis: Terry Davis (RIP) is an interesting part of internet history. He’s simultaneously a mad genius and what the internet might call a lolcow. This is a comprehensive history of his life, or as comprehensive as someone can get from information from the internet. It chronicles his life from his birth in Wisconsin to his times of homelessness and lawlessness and then to his death. I’ve used TempleOS before and it’s best described as “outsider art.” It’s not stable and not really usable, but is fascinating. You can see this article which is a “constructive look at TempleOS for a better overview.

Misc

Last night I had a dream that I was clipping my nails, and they kept being uneven so I’d have to clip further and further down until there was no nail left.

Ublock Origin keeps getting disabled randomly in Firefox, and I only realize it when there’s ads all over and pages load slowly. It’s fixed by disabling and reenabling the add-on. I’ll dive into the firefox logs next time I notice this happening.