@santanu_sinha Absolutely agree. It's also a lack of vision. It's easy to copy. And sure, people are used to using it... but what a lame reason. We have the power to set a new standard while also encouraging people to build things.
@esrtweet Agreed. It's lazy and a shame. No original thought, and just ripping software. This seriously discourages people to actually put time into making new software, because some asshole will go and try to decompile it.
@DavidOndrej1 Honestly, kind of garbage. He couldn't rethink anything? Instead, just rips it, lazy and horrible for the future of people building products. Basically pirating.
@melvindvivas Claude is crushing it. I'm a power user of both, and have had 3x subs for both Codex Pro and Claude Max. Right now, it's no comparison. If you really are curious, Claude is the best bang for your buck.
@Codexresets_ I just canceled 3 of my x20 subscriptions, Sol and Astra were simply getting stomped by Opus 5.5. Maybe i'll return, but it's not even close currently.
@CoderSteven@MatthewBerman Totally haha! I was trying this, and at least for me, Sol seems to be working way to slow even on Fast. I've gone back and forth over the last few months, and I think a big part for me is Opus 5.5 + Subagent handling feels fluid and rubust.
@0xpasan Not the cheapest, GPT 6.1 Sol is currently crushing. Using it atm as a robust heavy Subagent orchestration environment, and I am very impressed. Just built a whole new "Ticket" system around the CLI and it's nutty.
For context I've had 16 Agents on Fast for hours.