Cam Profile
Cam

@Cameron_C2

Followers
468
Following
315
Media
28
Statuses
227

i write code @voidzerodev

Joined October 2015
Don't wanna be here? Send us removal request.
@boshen_c
Boshen
5 days
Oxc parser gets 2 - 7% faster. https://t.co/LcUkha829c
5
13
154
@Cameron_C2
Cam
6 days
While diagnostics are helpful for devs, as agents becoming more and more common, its important for them to get detailed feedback, including how to solve the problem, not just what the problem is. Matt highlights it in his talk on DX to Agent Experience here:
0
1
9
@Cameron_C2
Cam
6 days
delightful diagnostics
2
2
28
@Cameron_C2
Cam
9 days
Oxlint v1.26.0 is out! πŸš€ β†’ new rule: typescript/no-deprecated β†’ better diagnostics for invalid typescript configurations β†’ JS plugins' settings access β†’ lots of bug fixes β†’ docs improvement for config options Full notes below: ⬇️
5
9
142
@Cameron_C2
Cam
11 days
AI can't one-shot complex features where correctness matters. It didn't "do the job"β€”it amplified my speed.
1
0
0
@Cameron_C2
Cam
11 days
This let AI maintain momentum on the easy stuff, while I dived into the nuanced cases. Ensuring neither of us were stuck! πŸ“ˆ
1
0
0
@Cameron_C2
Cam
11 days
Near the end, subtle failures stumped it. We switched approaches; time to pair: - "What can you handle solo?" - "What needs my help?" Teamwork makes the dream work. πŸ€πŸ€–
1
0
0
@Cameron_C2
Cam
11 days
AI needs tight iteration loops, just like us. Fix everything at once? Fixes nothing. "He who chases two rabbits catches neither" πŸ‡πŸ‡
1
0
0
@Cameron_C2
Cam
11 days
I asked it to run in a loop: 1. Run all tests 2. Pick one failure 3. Isolate & fix it 5. Re-run all & repeat This gave it a structured feedback loop, helping progress resume.
1
0
0
@Cameron_C2
Cam
11 days
Tests hit ~60% passing, then AI stalled - spinning wheels endlessly. So I mimicked human devs: FOCUS. πŸ”
1
0
0
@Cameron_C2
Cam
11 days
Scrapped AI's work and scripted a manual port. Now, everything's verified and deterministicβ€”no more "hope AI got it right." For precision, human oversight wins. πŸ‘¨β€πŸ’»
1
0
0
@Cameron_C2
Cam
11 days
With tons of tests, AI kept: - Skipping some - Stopping midway - Making subtle errors Correctness is key in oxlint, this meant there were hard to spot real bugs, and missing tests. Comparing hundreds manually? Torture. β›”
1
0
0
@Cameron_C2
Cam
11 days
Two key parts to any new lint rule: - Test cases (verification) βœ… - Implementation (logic) πŸ› οΈ Oxc has great infra for porting rules, but this one's type-aware. I asked AI to import tests from the original. Simple? Nopeβ€”total pain. 😩
1
0
0
@Cameron_C2
Cam
11 days
AI struggled hard implementing typescript/no-deprecated in oxlintβ€”with Copilot and Claude hitting ~60% test coverage and quitting. πŸ€–πŸ’₯ Here's what went wrong, my fixes, and lessons learned.
@Cameron_C2
Cam
11 days
In the next version of oxlint, support for the typescript/no-deprecated rule is implemented
1
0
6
@voidzerodev
VoidZero
12 days
What's new in ViteLand? Our October 2025 recap is out ✨ Vite+ πŸ“£ Series A announcement πŸ‘‰ @vite_js 7.2 beta πŸ’» @vitest_dev 4 with stable Browser Mode βš“οΈ Oxlint JS Plugins & Oxfmt pre-alpha πŸŽ™ Updates from the Community πŸ‘€ and more https://t.co/kkJuwPuOMO
Tweet card summary image
voidzero.dev
Our October 2025 recap includes VoidZero's $12.5M Series A, Vite+, Vite 7.2 Beta, Vitest 4, Rolldown and Oxc updates, upcoming events, and community highlights.
1
15
85
@Cameron_C2
Cam
11 days
In the next version of oxlint, support for the typescript/no-deprecated rule is implemented
1
3
29