2025
Finding the Size of an LRU Cache Fast
Consider the question “Given an empty least-recently-used cache, find the size of the cache if you can to tell when a search is a miss?” The straightforward solution to this is easy, load a new cache block, and then look through all of the old cache blocks to make sure that none of them unloaded. Repeat that until you find one of the previous blocks unloaded, and there you go. Assuming that checking the cache is in constant time, that solution is \(\Theta(n^2)\). However, we don’t have to look through all of the previous blocks to check if one deloaded, right? Is a \(\Theta(n)\) solution possible?
[Read More]2023
The End of Keitairon (For Now)
Over the past 3 weeks, I’ve been with my relatives (both my cousins and my aunts, though my aunts are technically younger than me) so I have not been able to work on Keitairon that much, but that hasn’t stopped some more important developments from occurring.
[Read More]Keitairon Week 3
Instead of doing a day-to-day breakdown, this week was mostly two main challenges: creating a website and kanji. As such I will break it down like so.
[Read More]Keitairon Week 2
Day 1:
After creating derivations the next thing to move on to was special verbs. Japanese has a small number of irregular verbs, and luckily all of them (except for 2) only have a few changes that need to be made. At first, my approach was to create special conjugations tables for each irregular verb, but then I realized that all of the irregular verbs follow either the Ichidan or Godan verb conjugations, with a few exceptions, and I could make special rules for those exceptions.
[Read More]Making Keitairon, a Japanese Morphological Analyzer
Day 1:
Today I have decided to start a new project called Keitairon. The name comes from the Japanese word for morphology (形態論) and my main goal is to create a program that can describe the different morphologies of a verb in higher detail. This comes from the fact that [jisho.org], while it can recognize some of the morphologies, struggles with others. Along with this, I hope that this will be able to be reversed and that all possible conjugations of a verb can be shown. (Maybe this can also extend to nouns, like with 無し).
[Read More]Can We Procedurally Generate Kanji?
I was recently thinking about how one could programmatically generate an infinite amount of Kanji. Kanji itself is made out of radicals, and it may be trivial to isolate radicals and use a random selection of them to generate a new kanji.
[Read More]2022
Will There Ever be New Kanji?
In Japanese, kanji (漢字) are characters borrowed or adapted from the Chinese alphabet, and over time Japanese has developed its own set of unique kanji. However, as the digital age has come about and fonts become widespread, will the evolution of kanji stop?
[Read More]Fek: Breaking Down a Language
Fek as a language was created organically through an online chat channel. The restrictions were that people were not allowed to speak any existing language, forcing them to communicate in a new language. The language that formed died sometime later, and I have gone and documented what I know of Fek here.
[Read More]The Glass Number and Interesting Open Forms
For a long time, my friend, Advay Patil, and I have been investigating linear recurrences, and in this blog post, I will briefly detail extracts from our research. You can find a complete paper here.
[Read More]