From cd4838f692718873e0a8e7838371aa0839493400 Mon Sep 17 00:00:00 2001 From: apex Date: Sun, 29 Jun 2025 04:21:52 +0000 Subject: [PATCH] Update README.md --- README.md | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/README.md b/README.md index 0776474..7109fc5 100644 --- a/README.md +++ b/README.md @@ -1,7 +1,7 @@ # valDict A very simple computer readable online dictionary based entirely off of data from **[Wiktionary](https://en.wiktionary.org/wiki/Wiktionary:Main_Page)** that uses data made with **[Wiktextract](https://github.com/tatuylonen/wiktextract)**. -valDict can be served using the built in react server that comes with this repo, or your choice of web server. +valDict can be served using your choice of web server. ## Building @@ -15,7 +15,7 @@ tune2fs -O dir_index /dev/DEVICE ``` -IN ADDITION, you need like... ALOT of Inodes like 10 million, i highly recommend making a DEDICATED ext4 partition JUST FOR this dictionary that has an outrageous inode count +IN ADDITION, you need like... ALOT of Inodes like 10 million, i highly recommend making a DEDICATED ext4 partition JUST FOR this dictionary that has an outrageous inode count (remember to set the large_dir and dir_index in the new fs) personally i recommend a 40-45GB that is 45,000,000,000 bytes / 10,000,000 goal inodes = 4,500 bytes per inode (waow that is crazy) this is a decently efficent way to patition the format while still leaving a very decent amount of space left for extra files, or whatever future expansion may be required making a filesystem like this would probably require first making an ext4 partition with fdisk, then using mkfs.ext4 -i 3200 to achieve the 3200 bytes per inode ratio...