I made riceable TUI client for Whatsapp
https://redd.it/1n7ftwb
@r_linux

terminal multiplexor scrolling

hi everyone,

I have been using tmux for a while but have never gotten mouse scrolling to work. I know I can use the keyboard, but I'd like to be able to use both. I understand that set -g mouse on is meant to make this work but it doesn't.

is there a multiplexor out there where this just works?

https://redd.it/1n7dpuk
@r_linux

Built an “Everything”-like instant file search tool for Linux Btrfs. I would love the feedbacks & contributions!!

I’m a first-year CSE student who was finding a file search tool and found nothing close to "everything" and I’ve always admired how **“Everything” on Windows** can search files almost instantly, but on Linux I found `find` too slow and `locate` often out of date. So I asked myself , "why not make one own" .

I ended up building a **CLI tool for Btrfs** that:

* Reads **Btrfs metadata directly** instead of crawling directories.
* Uses **inotify** for real-time updates to the database.
* **Prewarms cache** so searches feel nearly instant (I’m getting \~1–60ms lookups).
* Is **easy to install** – clone the repo, run some noscripts , and you’re good to go.
* Currently CLI-only but I’d like to add a GUI later. even a flow launcher type UI in future.

This is my first serious project that feels “real” (compared to my old noscripts), so I’d love:

1. Honest feedback on performance and usability.
2. Suggestions for new features or improvements.
3. Contributions from anyone who loves file systems or Python!

GitHub repo: [https://github.com/Lord-Deepankar/Coding/tree/main/btrfs-lightning-search](https://github.com/Lord-Deepankar/Coding/tree/main/btrfs-lightning-search)

CHECK THE "NEW UPDATE" SECTION IN [README.md](http://README.md) , IT HAS THE MORE OPTIMIZED FILE SEARCHER TOOL. WHICH GIVES 1-60ms lookups , VERSION TAG v1.0.1 !!!!!!!!

The github release section has .tar and zip files of the same, but they have the old search program , so that's a bit slow, 60-200ms , i'll release a new package soon with new search program.

I know I’m still at the start of my journey, and there are way smarter devs out here who are crazy talented, but I’m excited to share this and hopefully get some advice to make it better. Thanks for reading!

Comparison Table:

|Feature|`find`|`locate`|`Everything` (Windows)|**Your Tool** (Linux Btrfs)|
|:-|:-|:-|:-|:-|
||||||
|**Search Speed**|Slow (disk I/O every time)|Fast (uses prebuilt DB)|Instant (<10ms)|Instant (1–60ms after cache warm-up)|
|**Index Type**|None (walks directory tree)|Database updated periodically|NTFS Master File Table (MFT)|Btrfs metadata table + in-memory DB|
|**Real-time Updates**|❌ No|❌ No|✅ Yes|✅ Yes (via inotify)|
|**Freshness**|Always up-to-date (but slow)|Can be outdated (daily updates)|Always up-to-date|Always up-to-date|
|**Disk Usage**|Low (no index)|Moderate (database file)|Low|Low (optimized DB)|
|**Dependencies**|None|`mlocateplocate` or|Windows only|Python, SQLite, Btrfs system|
|**Ease of Use**|CLI only|CLI only|GUI|CLI (GUI planned)|
|**Platform**|Linux/Unix|Linux/Unix|Windows|Linux (Btrfs only for now)|

https://redd.it/1n7lkon
@r_linux

I created a flat, pastel-colored icon theme for Linux called Mignon!
https://redd.it/1n7nmrz
@r_linux

finding and disabling damaged ram addresses

my dad gave me his old ThinkPad that he stopped using because broke and some of the ram addresses cause the OS to crash, he told me that Linux has a way to disable those addresses if I can find them, so if I could get a little help with finding them that would be great.

the ram in question is 16gbs of soldered ram, there may be an easier way of just completely blocking all the soldered ram instead of just segments and that would work great for me too.

all my ram tests I've run say it's fine though so my dad might have been wrong about what caused the issue. it's caused one blue screen in windows, don't remember the code though and so far Ubuntu 25.04 has stayed stable.

https://redd.it/1n7qmko
@r_linux

Yummy ,I am gonna slurp up boot logs
https://redd.it/1n7rrgm
@r_linux

rtask 0.91-beta - select 1-N cpu(s) from cpu topology to run a linux command or pin a process

I have 2 Minisforum MS-01 servers that use Intel CPU's with big.LITTLE architecture comprising of performance P-core and energy efficent E-core CPU cores. Both run fedora linux 42.


They run a bespoke image database with various plug-ins to social media channels
and I noticed that selecting an image, resizing said image and generating a caption
text was taking anywhere from 4 to 14 seconds. Our billing system also had large
variations in how long it took to run a query and generate report (6 to 12 seconds).


Found time and took a look at what was causing such variations in runtimes.

For my set of applications it came down to:

1. the overhead of scheduling between p-core or e-core cpu's

2. a big pool of p-core cpu's also caused scheduling issues


With that in mind I created a little utility to easily:

1. list cpu topology and list which cpu's are p-core and e-core

2. manually specify 1-N cpu's to use to run a command or aleady running process

3. automatically generate a list of cpu's based on socket, numa, core and cpu

4. allow realtime scheduling and fast I/O priority scheduling


Using the rtask utility I was able to get faster and more consistent runtimes:

1. select+resize image with caption text: 1.5 vs. 4-14 seconds

2. generating our standard billing report: 0.6 vs. 6-12 seconds


Download: [https://lightaffaire.com/code/linux/rtask](https://lightaffaire.com/code/linux/rtask) (+ chmod 755 rtask)

```
$ rtask --help

Usage: rtask [options]
--pid process pin process
--run command run command
--time-it time the --run command

--realtime set real-time scheduling (can starve system)
--fast-io set if --run/--pid is I/O-bound (disk heavy)

manually assign cpu list (--list-cpu):
--cpu-list list rtask --cpu-list [1,2,N|1-N]

automatically generate cpu list:
--cpu-socket num cpu socket (default: 0)
--cpu-numa num cpu numa (default: 0)
--cpu-core num cpu type (default: .*)
--cpu-type text cpu type [p-core|e-core] (default: p-core)
--num-cpu num number of --cpu-type cpu's to assign (default: 4)
--all-p-core assign all p-core cpu's to --run|--pid
--all-e-core assign all e-core cpu's to --run|--pid
--randomize randomize cpu list

list cpu/scheduler info:
--list-cpu list cpu p-core and e-core layout
--list-raw list cpu raw values [maxmhz,mhz,socket,numa,core,cpu]
--list-topology list topology tree [socket->numa->core->cpu]
--list-scheduler list kernel scheduler

--system-info system info
--help help

Examples:
$ rtask --list-cpu

$ rtask --list-topology

$ rtask --list-scheduler

automatically select 4 p-core cpu's and run the command
$ rtask --run "COMMAND"

manually select 2 p-core cpu's and time the command
$ rtask --time-it --cpu-list 1,2 --run "COMMAND"

automatically select 2 random e-core cpu's and run the command
$ rtask --cpu-type e-core --random --num-cpu 2 --run "COMMAND"

automatically select all e-core cpu's for the running process
$ rtask --all-e-core --pid PID

fastest set of options to run the command
$ rtask --all-p-core --realtime --fast-io --run "COMMAND"
```

Lets check the number and speed of P-core and E-core cpu's on a MS-01:

```
rtask --list-cpu
13th Gen Intel(R) Core(TM) i9-13900H

P-core 5400Mhz
socket:0 node:0 Core:2 CPU:4
socket:0 node:0 Core:2 CPU:5
socket:0 node:0 Core:4 CPU:8
socket:0 node:0 Core:4 CPU:9

rtask --cpu-list 4,5,8,9

P-core 5200Mhz
socket:0 node:0 Core:0 CPU:0
socket:0 node:0 Core:0 CPU:1
socket:0 node:0 Core:1 CPU:2
socket:0 node:0 Core:1 CPU:3
socket:0 node:0 Core:3 CPU:6
socket:0 node:0 Core:3 CPU:7
socket:0 node:0 Core:5 CPU:10
socket:0 node:0 Core:5 CPU:11

rtask --cpu-list 0,1,2,3,6,7,10,11

E-core 4100Mhz
socket:0 node:0 Core:6 CPU:12

socket:0 node:0 Core:7 CPU:13
socket:0 node:0 Core:8 CPU:14
socket:0 node:0 Core:9 CPU:15
socket:0 node:0 Core:10 CPU:16
socket:0 node:0 Core:11 CPU:17
socket:0 node:0 Core:12 CPU:18
socket:0 node:0 Core:13 CPU:19

rtask --cpu-list 12,13,14,15,16,17,18,19
```

Now lets time a noscript that looks up whether an IP belongs to an OK or SPAM ASN:

```
$ time check-asn-ip 31.222.220.28

31.222.220.28 GB, England, E1W London
31-222-220-28.static.aquiss.com
asn+org: AS215066 Aquiss
inetnum: 31.222.220.0/24
netname: AQUISS-BROADBAND

OK: 31.222.220.28


real 0m7.553s
user 0m1.652s
sys 0m6.613s
```

And now the same noscript that uses by default 4 P-cores:

```
$ time rtask --run "check-asn-ip 31.222.220.28"

31.222.220.28 GB, England, E1W London
31-222-220-28.static.aquiss.com
asn+org: AS215066 Aquiss
inetnum: 31.222.220.0/24
netname: AQUISS-BROADBAND

OK: 31.222.220.28


real 0m1.275s
user 0m0.720s
sys 0m0.575s

```

Result: 1.275s vs. 7.553s

Download: [https://lightaffaire.com/code/linux/rtask](https://lightaffaire.com/code/linux/rtask) (+ chmod 755 rtask)


Always interested in constructive feedback either here or via Email [code@lightaffaire.com](mailto:code@lightaffaire.com)


Iain



https://redd.it/1n7wvll
@r_linux

September 4, 1998: Google was founded by Larry Page and Sergey Brin

On this day in 1998, Google was founded by Larry Page and Sergey Brin in their friend Susan Wojcicki's garage in Menlo Park, California.

Over these years, it has become not just a search engine, but a real lifesaver for all IT professionals. How many times has its search bar rescued developers in difficult situations - from finding solutions to mysterious errors to learning new technologies.

Google has turned the phrase “just Google it” into a universal answer to any technical question. Stack Overflow, GitHub, documentation - all of this became more accessible thanks to its algorithms. It has helped millions of developers find that exact piece of code that solves the problem at 2 AM before a deadline.

Thank you, Google, for making developers’ and sysadmins’ work easier and learning more accessible. Over 27 years, it has become an integral part of IT culture.

Happy Anniversary, Google! May it keep indexing the world of knowledge! 🚀​​​​​​​​​​​​​​​​

https://redd.it/1n81fwu
@r_linux

Thinking about installing arch for laptop

I have been using fedora kde for some time and was thinking that arch being one of the most minimal distros would help me get even better battery life. So tried it on a vm and let me tell you, apart from it being cli based, I haven't seen a smoother installation process. You connect to wifi if installing on a device but on a vm just type archinstall and you get the setup screen and after selection various stuff, it just installs without any problem.

Now my main concerns before committing to it are
1. I have heard that it being a rolling release model means you have to keep up with the news for some breaking changes that might cause your workflow to change. So I was thinking can't I just postpone the updating to like doing it once a month and figuring out if something broke and take care of it then and there. I mean is this possible, can you make cumulative update of the rolling updates or would you have to install all the updates that came along the way?
2. Is it really that bad that you have to follow its news for breaking changes?
3. Is it really hard to maintain it in the long term and are the problems that occur simple ones that can be repaired in a couple of minutes or do you have to go deep like spending time reading documentations and such to figure out what the heck is going on?

Would love to hear your thoughts. I an not a complete beginner and can solve problems with slight guidance, try to give recommendations from that perspective.

I mainly use electron based apps, like browsers, obsidian, vs code, etc. and would just like to squeeze as much battery as possible without compromising functionality, and I am not going with xfce or any of the lighter stuff. Have grown comfortable with kde.

https://redd.it/1n83paz
@r_linux

[OC] Linux Beginner Glossary
https://brogolem35.github.io/linux-beginner-glossary/

https://redd.it/1n84hg7
@r_linux

What is so bloated about GNOME?

For some reason, I see people saying that GNOME uses half of the memory even if you are doing nothing on your computer. I even come across people that say it’s as bloated as Windows 11 despite all of the telemetry on GNOME is opt in. I wonder how much actually bloatware does GNOME have and why people say KDE Plasma is much less bloated?

https://redd.it/1n85s0c
@r_linux

Teenagers first proper laptop - high school use - advice on hardware and software please

Good morning - and thank you in advance to anyone who takes the time to read this (let alone comment)

The situation: My 14yr old needs a laptop for his school work and so I’m looking to get something that is a balance of the usual; utility, robustness, future optionality (will he be a CS major? A graphics artist? Who knows), and so I’m looking for advice on a few things:
1. Hardware - which might not be the thing folks in this forum focus on, but I’m betting some of you have opinions: things like CPU, RAM, HDD vs SSD, screen resolution etc.
2. Operating system - This is why I am posting here. I used Windows laptops for most of the last 20 years - so I’m familiar with it, and this will be my default option (vs. Mac). I now have a Mac Air laptop which is fine for what I do, but I much prefer excel on a windows machine due to shortcuts. (Also my kid would bend that MacAir within 24 hours with how he just bounces around in the world.)

I want to avoid bloat-ware so the chromebooks and google OS stuff worries me (and I know windows has plenty of this also… I guess I’m just more familiar with it so am able to navigate it better) and this got me down a rabbithole for Linux. So here i am.


I’m old enough to have been through school and university without owning a computer (the rich kid at Uni had a 386..!) so I could be missing some requirements here but I see his needs to be fairly basic.
1. Documents, presentations, spreadsheets, and likely the ability to collaborate with project team-mates.
2. Technical writing features (mathematical formulae such as integration and differentiation)
3. Filing systems
4. Communication: emails, instant messenger
5. Art / drawing / picture editing (he likes to draw)
6 other..? I dont know of any needs for things like CAD or virtual machines in high school… maybe a younger person could help guide me here as to what might be on the curriculum.

I’d like my son to have a bit more knowledge in the underlying tech and architecture of “how things work” so that he’s better able to maintain (or modify) his equipment to suit his needs as they evolve. And I’d like to avoid him joining the ranks of the “less tech savvy” that seems to be growing amongst the younger folks due to apps just working out of the box (basically I’d like him to learn something his school may not teach him, and as a dad I want him to be independent and self-sufficient - do people still de-frag their drive to free up space?)

I’m honestly not 100% sure I’ve asked the right questions - I genuinely feel like a dinosaur - but I hope I’ve conveyed the sentiment. Any and all guidance is welcome.
Thanks again.













https://redd.it/1n88i6c
@r_linux

IF YOU PLAN ON DUAL BOOTING MAKE SURE TO KNOW YOUR BITLOCKER KEY I ALMOST LOST EVERYTHING BECAUSE BITLOCKER LOCKED ME OUT AFTER PARTITIONING MY DRIVE

As a warning to all windows users if you plan on dual booting do not lose your key write it on sticky notes save it on a flash drive tattoo it directly onto your skin everything unless your keen on nuking your drive and starting over

https://redd.it/1n8eqs1
@r_linux

Storage system like Ceph (policy based data placement) but for local storage (like ZFS)

I would love to have a storage system that I can throw storage, hdds, ssds etc... at and have a set of policies defined that ensure data is placed where needed to accomodate those policies.

For example a policy that requires 2 replicas, performance such as read throughput minimum (10MBs) and a write throughput (500MBs). Which would tend to indicate cold storage on HDDs, inbound write buffer to SSDs/NVME with writeback to HDDs.

Another policy could be IOPs based that would tend to excluded HDDs or require striping across many HDDs or maybe a policy that says recent data does not need replicas but once its 10days old it does (and maybe hands off to another policy) to accomodate scratch areas that must be fast but less likely to be needed when unused so could write back to HDDs for example.

Another policy concept could be a based on access patterns such as 'if 500MB of data is read from a particular directory, preload the entire directory to fast storage'

Or maybe something like requiring at least 2 replicas but if there are lots of HDDs with capacity available system can replicate 10x to speculatively improve read performance (can read from all/any of the 10 replicas). If capacity drops below some threashold replicas can be reclaimed.

In other words I want CRUSH but soemthing ZFS-like (local filesystem). Define rules/polcies, throw hardware (HDD, SSD, NVMEe) into a pool of capacity, iops, throughput and let the system dynamically figure out how best align those requirements. I'm also in a place where my awesome, power hungry, cluster running Ceph is turning into a single threadripper server which means I'm losing all the awesomeness that is Ceph/CRUSH by converting all my storage to ZFS.

https://redd.it/1n8fd0p
@r_linux

Input Remapper causing stuttering in video games

Realized today that Input-Remapper, which I've been using to put keyboard-inputs (Alt and Shift) on my mouse-buttons, is causing games to lag and stutter when I press those mouse-buttons.

I'm not entirely sure why this is, and I would like to either fix it somehow, or switch to a program where this isn't a problem.

One thing I'm wondering about is if it's related to it having to go through Wine, and if it's possible to then bypass Input-Remapper in some type of launch-argument (like a more complex version of launching it with a specific keyboard-layout).

https://redd.it/1n8jmhq
@r_linux

I'm running KDE with the least obtrusive taskbar possible --> Should I be using a different WM?
https://redd.it/1n8kf3v
@r_linux

Manjaro KDE vs Cachy OS KDE, the good and the not so good

Hello,

After using Manjaro for a few months I got into really optimizing the OS for responsiveness which to me relates to many things but also implies low RAM usage when idle, fast boot time and debloated programs and services running in the background. I would add a dash of clean GUI set up where everything I use regularly was easy to view or access, if not at a glance then after 1 or 2 clicks at most.

https://www.reddit.com/r/kde/comments/1n4876s/with\_small\_tweaks\_manjaro\_kde\_idles\_at\_916mb\_of/?utm\_source=share&utm\_medium=web3x&utm\_name=web3xcss&utm\_term=1&utm\_content=share\_button

After recently making a post on my optimization process I received several messages criticizing Manjaro and that I should have chosen something else, CachyOS being one of the known and popular alternatives that use KDE as one of their main desktop environments. After hesitating for a bit I gave it a try and I was not pleased by what I found, contrary to popular belief CachyOS lacks polish and is less usable and stable than Manjaro. Let me break it down.

Boot time. After running the same optization on both distros, namely using the contents /etc/xdg/autostart as reference point to find out if there are programs or KDE features I don't need that I can uninstall, editing GRUB timeout and changing the value to 0, installing the latest available kernel version, editing Background services using the KDE window with the same name and turning off and disabling services I don't need and after editing Configure System Tray and disabling widgets from starting automatically at boot that I don't need the results were....crash. CachyOS stuttered, crashed the plasmashell once, after forced reboot it did the same but this time it reverted back to the log in screen and proceeded to freeze after entering the password and attempting to log in and after a second forced reboot, it froze once again while using Configure System tray, but this time it managed to restart the shell after a few seconds. Was this a KDE problem? Was it a kernel problem? Was it something else like zram which is automatically preconfigured for CachyOS and very aggressive? Maybe, idk, but the experience was not pleasant. Also after one of those reboots it also failed to enable the keyboard after booting.

Getting back on track, boot times. After finishing optimizations the best result I got on Cachy OS was 18s while on Manjaro with the 6.17 kernel I managed 13.2s

https://imgur.com/a/yA5WwMi

https://imgur.com/a/xddFOU1

There is not much to say here other, other than lowering the grub timeout, the same bios (UIEFI) settings were used and yet the results are quite different. Almost 50% more time needed for Cachy OS using the same boot loader, namely GRUB.

Ram usage while idling on the desktop after fresh restart, which to me communicates how bloated the system is compared to how optimize it could be with a common sense setup that provides all the needed functions while not being as bare bones as the command line stans would desire. Here again I noticed a gap with CachyOS being more RAM hungry after simillar optimizations

CachyOS

https://imgur.com/a/jX1xTrJ

Manjaro

https://imgur.com/a/6Gpv6xx

Of note here is the pretty aggressive use of zram which on one side does appear to make the GUI more responsive but also has introduced the chance for freezes and stutters which I would qulify as a fail for a normal, daily use OS and Manjaro did not really feel slow ever, in fact the most impactful setting one can make for KDE to make the GUI appear to work quickly and be responsive is to go to System Settings>Quick settings or General Behavior (depending on the distro the wording for this category might be different

despite all being Plasma) and finding the "Animation speed" slider and setting it to the fastest value. Here's a lsblk from Cachy displaying the use of zram which was configured by the installer, I had no input.

https://imgur.com/a/Rd2YpsU

Lastly, though a bit unrelated to performance and more related to usability, the Package Manager for CachyOS is far less intuitive to use having a very simplistic GUI compared to Manjaro which offers by contrast an easy way to browse installed packages and install/uninstall them at leisure. Not so clear or usable on the CachyOS side though I give it props for listing the repo packages in a list which makes it a lot more usable for casual use than requiring to know the console command to install them. This is a feature that Manjaro should copy.

CachyOS (crappy) GUI for packages

https://imgur.com/a/cxQumhG

CachyOS (useful repo list of packages) GUI in the package manager

https://imgur.com/a/07vSBGQ

Overall I am not impressed with CachyOS compared to Manajaro for daily use, far less stable and easy to use for casual PC users, especially those migrating from Windows. I give it props for the responsive GUI, likely a combination of aggressive zram config and fabled CachyOS optimized and kissed approved kernel but this trades off stability and leaves a bad first impression. These, imo should be user enabled features post install and not configured automatically from the start. Mediocre boot time, dodgy GUI decisions and overly enthusiastic optimizations, frankly speaking fanboys need to take a sit and be more humble, Manjaro in my findings is far more casual PC user friendly and better set up for first time Linux users. Use CachyOS at your peril, better dual boot with a stable distro, it doesn't boot fast anyway so no need to cry about it with multiboot.

https://redd.it/1n8j0pi
@r_linux

Lex Fridman quoted Linux when talking with M$ famous Dave Plummer
https://redd.it/1n8o22i
@r_linux

Database Subsetting and Relational Data Browsing Tool.
https://github.com/Wisser/Jailer

https://redd.it/1n8o8bu
@r_linux