26, Ex-SWE → Solopreneur 📈Working on #Canvograph ($0 MRR) 🗺️Submitted #Luxiom for Appstore ($0 MRR) ✉️Blocked @Minmailist due to CASA
Terminal
Joined October 2025
- Tweets6.3K
- Following602
- Followers804
- Likes10.9K
Pinned Tweet
It's a beautiful Saturday Morning and We breached 800🙌🔥🔥
Thank you for you support everyone🙌
I'm also working on my content strategy!! It's still in constant flux but end of the day, rule of cool thats it
#Tabbular → I have like 7000 tabs spanning over 6 years😭. Codex picks 5 random tabs everyday and see what we can clean up and revisit. And it's been enlightening
#Canvograph and #Luxiom → My chrome extension and my ios app(stuck in app review), I'll share my distribution plan and process as a noob. Right now, my last update was last week because I didn't focus on this, will get back
Interesting stuff on X/Github/Reddit → There are many interesting and amazing articles and posts. Recently, I shared and summarized an article from Yihui which was amazing, how he distributed and planned Skillry. Overall, this type of content is super useful for everybody.
Coding, System Design and AI → Haven't shared anything yet, but I do have plans. I'm reading Illustrated Guide to AI agents and also finished free code camp's "Senior System Design" Youtube video
Yapp → Just yapp about random stuff, Apple (Macbook btw) and also quirky stuff which I generate with Argon (shared some fun feature ideas I had about Macbook).
I like anime and manga (I miss bleach btw😭)
But yeah, sometimes I might yap about stuff so @X please dont nuke my reach😂
Tabbular issue #6 teapawt.org/devdiary0x/status/2116…
towardsdatascience.com/gener…
An article about GAN, it was the rage back in 2020. People used it for image generation. Didn't read it, might pick it up someday but not now
explainthatstuff.com/howcomp…
Bookmarked it loong time back. Fundamentals about how things work, it's not bad, it does have interesting info about everything (science, tech, etc). But no time to read it😂
DevDiary retweeted
"temperature = 0 means the LLM gives the same output every time"
not necessarily
- temperature 0 makes decoding greedy, picking the highest-probability token
- GPU floating-point operations aren't always perfectly deterministic
- batching and parallel execution can slightly change numerical results
- even tiny differences can flip the next token when probabilities are close
and once one token changes, the entire response can diverge. deterministic decoding ≠ deterministic inference
movable-type.co.uk/scripts/s…
A nice implementation of SHA-256, damn I got hooked and super interested, I was reading this and writing out code and logic by hand😂
Still halfway through, will be done in 1-2 hrs, its too much fun dammit
docs.oracle.com/cd/E19957-01…
A super dense guide about FP arithmetic, might pick it up some other day. It's really fascinating to see how software and hardware handle floating point numbers
Ughhh Gemini Argon's usage is soooo low
Pro gets over in 10 prompts
They keep pushing ultra's 5x but its just 50 prompts ig, will get over in 20 mins(chatty session)
Hope Carbon has better usage
DevDiary retweeted
the craziest thread on X right now is THIS one
@pancakegtm is going for the tallest pancake stack record, live.
photos from people in the room are in the replies 👇
(best photo gets $1k btw)
Illustrated guide to AI agents is seriously really good
Ik stuff, but after reading this and playing around with Gemma 4b I feel more in control
Plus I have a vibe coding interview this weekend, understanding the tool + being the guy in control is what helps in actual enterprises
DevDiary retweeted
Day 25/
We are doubling all usage limits for all Pro 100/200/500 tiers and Business Premium, effective today.
We heard your kind feedback and won’t be reducing usage limits for Pro 200, so your Work and Codex usage will stay the same. We’re also going to continue offering extended limits for dots for now, and usage of your dot will still continue to be separate from your default usage.
And we'll of course also keep shipping better models and other efficiency improvements that increase your effective usage over time.