ojr

6 ms·

31.
by ojr 9mo ago
I built my own ai coding agent and do vector search and embeddings locally https://slidebits.com/isogen
32.
by ojr 9mo ago
I don't get this sentiment, regressions still exist, you can't just prompt them away and a programmer will spend 10x more time fixing regressions, bug fixing and improvements than scaffolding in most projects that people pay for.
33.
by ojr 9mo ago
I found out that Claude Code and OpenCode doesn't do vector search and embeddings but relies on grep to power their agents. Most people are satisfied with this even though I think vector search approach is way better and saves on token
34.
by ojr 9mo ago
https://olaji.de my first name is Olajide https://olaji.de/portfolio
35.
by ojr 9mo ago
Valorant, Fortnite, Call of Duty, Apex Legends, Destiny 2 and Rainbox Six Siege are not supported on Linux, how much of those 30 games of 1000 are some of the most played games? I already name 20% of them.
36.
by ojr 9mo ago
Windows still have the gamers. A lot of anti-cheat system completely block out Linux users. The Year of the Linux Desktop will still be a meme at the end of this year as well.
37.
by ojr 9mo ago
I spent so much time getting the mocks right with AI tests and the tests could not be one shotted or done by an inexperienced intern. Certainly don't have the budget to through Claude Opus on it, I'll give it some time though mayb
38.
by ojr 10mo ago
99% of the US population have access to 4G, caring about 3G is a wasted effort. The point I’m trying to make about tech stack is users don’t care if you used Gemini, ChatGPT or Claude to generate code. As someone who hasn’t converted to SSR
39.
by ojr 10mo ago
In the next year developers need to realize normal people do not care about the tech stack or the tools used, there are far too many written thoughts and opinions and not enough polished deployed projects. From an industry standpoint it’s b
40.
by ojr 10mo ago
I didn't have reliable structured outputs until I switched to Gemini 2.5
41.
by ojr 10mo ago
https://slidebits.com/isogen An AI coding tool desktop application written in Rust and Javascript. Cursor, Windsurf and etc uses too much memory on my machine. As an engineer it is important that the tools I use daily are p
42.
by ojr 10mo ago
Cursor uses too much memory, hallucinates on multi file rewrites, and has a difficult accept/reject diff user interface that makes git a nightmare to use. I built my own coding tool in Rust because of VSCode-based forks Electron app fl
43.
by ojr 10mo ago
I would have never heard of Studio Ghibli along with several others if it wasn't in the training data. The result? More awareness and relevance, more visit to their museums, more viewing streams, and more snarky discussion comments sur
44.
by ojr 10mo ago
the people making AI moderation tools are also good, and in my opinion more skilled than the abusers
45.
by ojr 10mo ago
being pedantic doesn’t change the fact of what I said. If I said Obama is a better a leader than Julius Caesar, would you reply well actually, Julius Caesar is dead.
46.
by ojr 10mo ago
Disney witnessed AI boosting Studio Ghibli’s cultural impact and relevance and decided to invest big in OpenAI, not surprising.
47.
by ojr 10mo ago
Tiktok has AI moderation tools that you are highly underestimating.
48.
by ojr 10mo ago
> So in essence some concepts strictly exist to protect against malicious actors, while others protect against random conditions or good faith actors. Guess what both fall under umbrella of cybersecurity. Please consult with your favorit
49.
Show HN: Isogen – Lightweight AI Coding Tool (Rust and JavaScript, <50MB, BYOK)(slidebits.com)
1 points by ojr 10mo ago|0 comments
50.
by ojr 10mo ago
Security is built out of generic programming concepts, rate limiting, input validation, authorization, encryption, etc. When you apply these programming concepts like atomic operations to something that control moneys it becomes security en
51.
by ojr 10mo ago
I didn’t know what atomic operations, important concept in cybersecurity until working with crypto and blockchain, prompting AI is a skill, there’s an obvious context engineering skill gap I possess. I built a desktop application in Rust in
52.
by ojr 10mo ago
Crypto helped me step up my cybersecurity skills forever, when someone was exploiting a race condition in my AI app, I only knew how to prompt AI to do atomic operations because of popular smart contract hacks. I wish I got funded to make a
53.
by ojr 10mo ago
My first text editor for PHP was Komodo Edit, it was super slow and everyone jump shipped to Sublime Text, then VSCode was slower than Sublime but had incredible industry supported extensions like Git and Prettier, programming didn't p
54.
by ojr 10mo ago
As long as the dollar remains the reserve currency of the world and US retains its hegemony, a lot of the finances will work itself out, the only threat to the US empire crumbling is by losing a major war or extreme civil unrest and that th
55.
by ojr 10mo ago
all I need is one SaaS to work out, I've built five of them this year, the fifth one is the one though, I promise.
56.
by ojr 11mo ago
I always thought years of experience in a language was a silly job requirement. LLMs allow me to write Rust code as a total Rust beginner and allows me to create a valuable SaaS while most experienced Rust developer never built anything tha
57.
by ojr 11mo ago
compare the worldwide latency, I released an app in the App Store, I got users from Japan to Saudi Arabia to the United States. AWS basically guarantees to reach anyone who speaks English low latency.
58.
by ojr 11mo ago
that is what they do in the software industry, before it was let me catch you off guard with asking how to reverse a linked list, now its leetcode questions that are so hard that you need to know and study them weekly, and prep for a year,
59.
by ojr 11mo ago
the orchestration layer is the moat, ask any LLM and they will give paragraphs explaining why this is...
60.
by ojr 11mo ago
5% prompt engineering, 95 % orchestration and no you can not vibe code your way and clone my apps, I have paid subscriptions why aren't you doing it then? Oh because models degrade severely over 500 lines. LLMs is the new AJAX. AJAX ma

More ›