From 70e00a426b4e3a428025552e570d8dca3555189c Mon Sep 17 00:00:00 2001
From: Himadri Bhattacharjee <107522312+lavafroth@users.noreply.github.com>
Date: Wed, 24 Sep 2025 10:46:09 +0530
Subject: [PATCH] feat: update jdupes article

---
 content/post/liberating-14GiB-of-space.md | 72 +++++++++++------------
 1 file changed, 33 insertions(+), 39 deletions(-)

diff --git a/content/post/liberating-14GiB-of-space.md b/content/post/liberating-14GiB-of-space.md
index 1d91b4dc..d969e60d 100644
--- a/content/post/liberating-14GiB-of-space.md
+++ b/content/post/liberating-14GiB-of-space.md
@@ -14,73 +14,67 @@ The idea is simple:
 
 Since the mileage for second step might vary from person to person, I'll elaborate on the first step.
   
-I chose [jdupes](https://github.com/jbruchon/jdupes) as my weapon of choice for finding and removing the duplicates.
-It's free and open-source and is cross platform.
+I chose [jdupes](https://codeberg.org/jbruchon/jdupes) for deleting the duplicates because it's open-source and is cross platform.
 
 For a given folder we would run the following to wipe the duplicates:
 
 ```powershell
-jdupes -rdNz .
+jdupes --recurse --delete --no-prompt --zero-match .
 ```
 
-Let me explain the flags:
-
-Flag|Explanation
--|-
-`r`|Find duplicates recursively
-`d`|Delete duplicates
-`N`|No-prompt: when used with the `d` flag, it keeps the first file and removes all the others in a collection of duplicates 
-`z`|Consider zero length files to be duplicates
-
-The `.` here means the current directory.
-
-Please read the tool's help page for more granular control during the cleanup.
+This will recursively delete all the duplicates except the source file without prompting for a confirmation.
+It will also consider zero length files to be duplicates.
  
-The computer in question runs Microsoft Windows and there's a thing
-common in almost all Windows setups, *drives*.
+The target OS is Windows, which has the glaring problem of *drives*.
 
-This was a glaring issue. There could be files that are unique in a given drive but are actually duplicates
+There could be files unique in a given drive but are actually duplicates
 in the inter-drive space. There are two ways to combat this.
 
-First method:
-- Run jdupes on a drive to free some space
-- Move some data from other drives into the current drive to fill it up again
-- Repeat
+### Rinse, move, repeat
+  - Run jdupes on a drive to free some space
+  - Move some data from other drives into the current drive to fill it up again
+  - Repeat
 
 This, obviously, is a terrible idea beacause we have the overhead cost of moving the files after each run
 as well as the fact that we have to run jdupes exhaustively for many iterations.
 
-Second (and probably the more elegant) method:
-- From the space of drives to be cleaned, pick a random drive (parent)
-- Hardlink all the other drives from the space into the drive we picked previously (children)
-- Run jdupes
+### Single pass with hardlinks
+  - Pick a random drive as the parent node
+  - Hardlink all the other drives to the parent
+  - Run jdupes
 
-This method only requires us to run jdupes once.
+Consider the following scenario
 
-Assuming we have picked the `A` drive as the parent and the `E` drive is one of the children,
-we would run the following powershell command to hardlink `E` drive to a folder called `Edrive` in `A`.
+- Parent drive: A
+  - Child drive: E
+  - Child drive: B
+  - etc.
+
+To create the hardlink of `E` in `A`, we would run the following in powershell.
  
 ```powershell
 New-Item -ItemType HardLink -Path A:\Edrive -Value E:\
 ```
 
-We would repeat this for all the children drives, modifying the command
-ever so slightly to meet our needs.
+We repeat this for the rest of the drives like drive `B` where
+- `Edrive` becomes `Bdrive`
+- `E:` becomes `B:`
 
-This implies that when we run jdupes from the root of the `A` drive, it would
-traverse the hardlinks and find duplicates in the inter-drive space.
+When we run jdupes from the `A` drive with
 
-Next, we'd go to the root of `A` drive and run jdupes.
-```powershell
-A:
-jdupes -rdNz .
+```sh
+cd A:
+jdupes --recurse --delete --no-prompt --zero-match .
 ```
 
-Finally we remove the hardlinks:
+it traverses the hardlinks and removes duplicates in all the linked drives.
+
+Finally we can remove the hardlinks
+
 ```powershell
 rm A:\Edrive
 ```
 
-> Note: Do not run jdupes at `SYSTEMROOT` (`C:` drive for most people)
+> Note: Do not run jdupes at `SYSTEMROOT` (the root of `C:` drive for most people)
 as there are legitimate duplicates which, if deleted, can brick a system. I'd recommend
 running jdupes in individual directories like _Music_, _Documents_, etc.