Supan Adit Pratama

Open Source Lover

[Peter Dedene @dedene]

POV: you're still using GitHub Copilot after June 1st, 2026

[Ray🫧 @ravikiran_dev7]

NEVER HIRE A VIBE CODER AGAIN !!!

[Supan Adit Pratama @supanadit]

@NoMansSky @hellogames This posted in May + Bee ? Maybee ?

[Supan Adit Pratama @supanadit]

@LottoLabs Kimi K2.6 Higher than MiniMax M2.7. I won't survive till end of the week

[Supan Adit Pratama @supanadit]

@LottoLabs I would love to use Kimi but my weekly limit quickly filling up although using model L2 like MiniMax M2.7

[Supan Adit Pratama @supanadit]

@MiniMax_AI @MiniMax_AI International users are paying a 3x tax for laggy routing.

Global: $40 (~700k IDR) ➔ 260ms latency
Domestic: ¥98 (~250k IDR) ➔ 24ms SG route

Hard to justify paying more for a worse connection just because we lack a CN phone

[Supan Adit Pratama @supanadit]

@MiniMax_AI From Indonesia, global https://t.co/4nt1FkwS3z has awful latency (~260ms + packet loss). https://t.co/mQHl8LYdHl resolves to Singapore at a crisp ~24ms, but registration requires a CN phone.

Any plans to fix global SEA routing or allow international signups?

[Supan Adit Pratama @supanadit]

@Samaytwt Turbo C

[Supan Adit Pratama @supanadit]

@HumansNoContext 🦅"Hey, thank you, anyway i like your shoe...."

[Supan Adit Pratama @supanadit]

@tonysimons_ @MichaelGannotti @ollama Me too... weekly bar increase more than 5hr...

[Supan Adit Pratama @supanadit]

@ramxcodes Recently Ollama Cloud unstable due too extremely slow for some model, but i guess no monthly cap. I hope they fix slowness as quick as possible

[Supan Adit Pratama @supanadit]

@ramxcodes I switch to Ollama Cloud :) My limit reach 100% for the rest of subscriptions

[Supan Adit Pratama @supanadit]

@ollama @FreedomGoat_ This is the full picture of the API response time

[Supan Adit Pratama @supanadit]

@ollama @FreedomGoat_ This post from Apr 21 ? Today is May 5, few days ago I subscribe to Ollama Pro, I thought the speed would be better than OpenCode Go... but I was wrong, it's slower...

[Supan Adit Pratama @supanadit]

@ollama @FreedomGoat_ I use OpenCode, the model mostly Qwen 2.5 both plan and build mode. I want to use Deepseek V4 Pro or GLM 5.1 in Plan Mode, then switch to fast model like MiniMax M2.5 or Deepseek V4 Flash or even Qwen 2.5. But I can't. it's deadly slow for code. The only reasonabily speed is Qwen

[Supan Adit Pratama @supanadit]

@divina__nemesis Agree, i just subscribe last week, the worst thing is i need to wait 1 - 2 minutes for each /v1/chat/completions.... IF not 502 error

[Supan Adit Pratama @supanadit]

Go checkout the latest update at https://t.co/6wmyQj5tfZ

[Supan Adit Pratama @supanadit]

phpv just got bigger: PHP 8.5 support, phpv extensions to browse compatible extensions, rebuild command to add extensions without full recompile, phar manager for composer/PECL, .phpvrc & composer.json auto-switch, Zig compiler for legacy PHP, and a lean minimal install.

[Supan Adit Pratama @supanadit]

Stop fighting with PHP versions. 🐘

I built PHPV to end the installation struggle. Similar to NVM, Rbenv and Pyenv.

🚀 Manage PHP like it's 2026: https://t.co/6wmyQj5tfZ
#PHP #OpenSource #DevOps #Golang