sol

[sol @thndrwsp]

@morganlinton The piece on tooling coming anytime soon? Would be very helpful to understand how to do this properly instead of semi manual approach

[sol @thndrwsp]

@thekitze @sama Back in the old days, I used Notion *exactly* for this haha

[sol @thndrwsp]

1) Wasn't /goal presented at I/O already? And it's probably the worst implementation among agentic IDEs. Same problem as before - model doesn't try to think, rushes with shitty solutions to you ask. Thought that /goal is here to explicitly help with long running hands off tasks, but it's virtually useless in Antigravity compared to Codex/Claude and even @opencode plug-in
2) Reworked model limits have been presented like a week or two ago, ans repeatedly being artificially hyped again again. Here they do it again
3) Workspaces being renamed to Projects is considered an important enough update?

And several other points.

What's going on with @antigravity. Used to love it half a year ago. It and gemini for coding are barely usable now. To the point that if I do use them instead of the codex, I feel extremely frustrated and not only less productive but unproductive.

The speed on 3.5 Flash could be a huge differentor, even at this state has some useful applications, but the model itself is extremely bad at coding. Out of principle, spent like 10 prompts trying to align 2 cards on my web app - wasn't able to do it.

In any case, feels that folks at Antigravity try to hype up the product and artifically create an image of high velocity development of it. But as with benchmaxxing for Gemini - it's now obvious, disingenuous, and tiring.

Love Gemini for general knowledge work, was weirdly partial to Gemini in general past few years, was advocating Antigravity 1.0 a lot - because it was genuinely good, fresh, and felt like a new and correct way of interacting with agents (now proven by Codex, T3, and others, who all feel like an evolved and improved Antigravity 1.0).

But at this point trust is evaporating when it comes to Gemini/Antigravity for dev work.

But still the best for knowledge works and Assistant-like usecases.

[Google Antigravity @antigravity]

Mission Report is out! Check out the latest product highlights and updates.

[sol @thndrwsp]

When on Android tho

[Flighty @Flighty]

Introducing Connection Assistant 2.0!

🧭 Every step + its typical time
🛂 Every checkpoint: customs, security, bag recheck
🚇 Every train, tram, bus
🌎 Every airport & airline
🪪 Custom to your citizenship & seat class

All-new clarity in one of flying's most stressful moments.

[sol @thndrwsp]

@Anton__Chigurh_ @bscholl @altcap @boomsupersonic I understand that technically might be inflationary but the scale is too small and the direction is long term investments, not consumption/spending. In fact, short to mid term this might be deflationary, no?

[sol @thndrwsp]

The support (human) got back to me finally, it's on their side indeed that the banked resets were not issued to my account. The responsible team will review further and will get back with the conclusion whether I'll be granted the resets or not.

I take it that either banked resets are actually *not* for everyone (despite what @thsottiaux said) or that there is something going on with their system. In any case, very weird.

[sol @thndrwsp]

We need Codex app on Linux ASAP, more or less the only reason to have macos as a co-main machine at this point.

[sol @thndrwsp]

@AndrewWarner Can it learn you vocabulary from corrections and on-screen context?

[Sheel Mohnot @pitdesi]

Truly deranged to tell people not to take free money the government is giving you bc of the name on the account

I suspect the name doesn’t last that long anyway

[sol @thndrwsp]

@micsolana This list itself is enough to suspect that the methodology/raw coefficient does not befit the task.
If we assume that Gini shows that there is something wrong with how people actually live in the country, we wouldn't see UAE, KSA, USA and Sweden this high.

[sol @thndrwsp]

@YashHustle_22 I don't even understand how I was able to use it before. Crazy difference.
But Gemini is still great, maybe the best, for non-coding knowledge work.

[sol @thndrwsp]

Day 6 since this and now banked resets at all. No banked resets for any previous distributions as well. I'm going to continue whining about it.
No word from support either.

[Tibo @thsottiaux]

Codex usage limits will be fully reset again in the next hour and we will credit one additional reset into your bank for your own usage over the next 24 hours.

We investigated reports that Codex usage was being consumed faster than expected. There wasn't one central issue, but a few smaller problems compounded for some users.

Here's what we found and changed:
- Actual usage: Auto-review had become more proactive, another change was triggering more subagent work, and background suggestions could run twice or retry too frequently after failures. We reverted the changes and fixed suggestion scheduling, duplicate generation, and retry behavior. This should reduce unnecessary background token consumption while preserving the work users explicitly request.
- Usage reporting: Auto-review was incorrectly appearing as GPT‑5.4 usage, and failed or rate-limited requests were still shown as turns. Auto-review now appears as its own category, and only successful requests count toward the turn graphs. Rate-limited requests were never charged, but they were being displayed incorrectly.
- Immediate relief: We reset usage limits while rolling out the fixes, then shipped hotfixes across the CLI, desktop app, and usage backend.
- What to expect: New usage data should be clearer and actual consumption should be lower. Historical charts may still show auto-review under GPT‑5.4 because older turn data was not relabeled. Features that intentionally perform more work; such as /goal, subagents, and higher reasoning levels will still naturally use more capacity.

All fixes are now deployed, and we've added more detailed monitoring so we can detect background-usage regressions sooner. We'll continue watching the results closely.

Thank you for building and doing all sorts of things with Codex.

[sol @thndrwsp]

@rodydavis @ritu_twts @antigravity @shengzheyao @_anshulr @_mohansolo This is a bait post, they do these every so often

[sol @thndrwsp]

@thdxr What if they are...

[sol @thndrwsp]

>The Americans have succeeded as against the Europeans and the Japanese because they have m more extremes of random behavior

Bars, God damn

[rinko @mrinko]

lee kuan yew on american culture

[sol @thndrwsp]

Such an amazing and fun aspect of LLMs for non-developers. Very grateful that this is happening now and that the progress is this fast. Amazing times!

And huge thanks to @theo @ThePrimeagen @morganlinton @yacineMTB, inspired me to start doing this.

Probably shit quality, but this is so much fun and led me to start learning how thing actually work as well.

[sol @thndrwsp]

I'm not a developer. Have some *very* basic understanding of dev work, some product sense, have seen a lot of products as a VC as well.

But I hate SaaS products that could've easily been a one time purchase. Also, love FOSS but oftentimes it's ugly.

LLMs are coming in clutch.

[sol @thndrwsp]

For exampme, wanted to have Granola for a long while, significant reason to be on macOS for that type of work. Not anymore. Easy to build replacement. Not perfect, but functional enough, and tailored to my needs too! Plus it runs on local models!