Rendered at 17:07:04 GMT+0000 (Coordinated Universal Time) with Cloudflare Workers.
whimblepop 22 hours ago [-]
There aren't any documented settings to disable this, but there is a feature flag it currently sits behind. You can flip it in your ~/.codex/config.toml[1]:
[features]
daemon_auto_start = false
There are also some files that it writes to in ~/.codex at startup: pid files, lock files, things like that. You can delete them and replace them with symlinks to /dev/null, then mark those unwritable with `chattr +i` or similar. Presumably then the background daemon will crash on startup, if the feature flag ever goes away.
Yes this should be turned off by default. The daemon is useful when building apps on top of the daemon server, e.g. pi or pi-durable custom harnesses.
whimblepop 22 hours ago [-]
What makes ACP deficient for headless uses of LLM agents? Maybe fixing that or standardizing on another protocol designed for pluggable LLM agents (I guess ACP is more oriented towards pair programming in an IDE, at least originally) is a better idea imo.
sy135673 10 hours ago [-]
[flagged]
vunderba 22 hours ago [-]
Yep, the real issue is that it was something they didn’t have before, and then with Codex 0.157.0 enabled, they installed this persistent nonsense [1] that survives after the last instance of Codex has exited on your machine, which is just gross.
For whatever reason, the daemon has been refusing to start and Codex has been grumbling to me. I was wondering where the daemon had come from, and thought maybe I missed something previously and that it had just broken.
starcast2026 16 hours ago [-]
Hope they don't remove this flag
zulban 4 hours ago [-]
And then it changes a month later. Then new (bad) defaults. And again and again. Good luck with that.
icantevenhold 22 hours ago [-]
It also fills up my HD with old versions.
I investigated and it took up something like 12 gigabytes in total on my system; which seems insane to me, it’s just a harness?
People moan about how badly made Claude code is but Codex seems much worse on every metric
InsideOutSanta 22 hours ago [-]
All of the harnesses I've used are terrible at cleaning up after themselves. They leave all kind of clutter behind, to the point where on my dev VM, I now have an LLM run once a day to just look through the disk and delete everything that looks like it's leftover garbage.
It's like fighting fire by employing an arsonist.
cstever 17 hours ago [-]
Running an LLM to delete files seems like overkill to me. Couldn't a cron job do the same thing? Or is the clutter just put into random places, which, if true, seems like bad software, which then leads me to consider the output of LLMs in general, which then leads me to wonder why are we doing this?
But, I'm just wondering.
dannyw 16 hours ago [-]
Oh it makes all sorts of temporary files and directories, sometimes in the right places as per UNIX filesystem conventions, but often not.
hiitsmyaccount 15 hours ago [-]
You could use snapshots or whatnot to reset the state whenever you like. Using an LLM to clean up files often seems like overkill to me.
ipaddr 12 hours ago [-]
Wouldn't a snapshot delete everything else you saved?
groby_b 11 hours ago [-]
You keep your source tree on a separate file system with version control, one hopes.
So you only snapshot the rest. (And since your source tree is under version control, finding stray garbage there is straightforward unless you checked it in)
noosphr 14 hours ago [-]
Just have an emphemal VM which you nume daily.
Its the only way I trust to install anything from these companies.
judge2020 19 hours ago [-]
Installed "clor.com" and it got into a loop of downloading its docker image from ghcr them throwing it away immediately because it was supposedly out of date (iirc). 65mbps continuous transfers for a week until I looked at my network traffic and saw 500gb of rx per day.
s3p 16 hours ago [-]
Sisyphus' updater
klodolph 18 hours ago [-]
Over eager use of /tmp is the worst. That’s RAM, my boy! Precious gold.
arcanemachiner 17 hours ago [-]
I don't think that's always true.
Although I just looked into it, and even Ubuntu 26.04 appear to default to tmpfs for /tmp storage, so I guess it's fairly universal now.
Not sure if I agree with that idea in principle... What's wrong with just wiping it intermittently??
trollbridge 16 hours ago [-]
It’s RAM unless you have swap enabled.
matltc 12 hours ago [-]
Did not know this. Was looking at htop yesterday scratching my head as to why there was over a GB in swap with ram <0.25 capacity. Had been teeing logs there pretty heavily
clickety_clack 4 hours ago [-]
Why wouldn’t you just rebuild the VM every day? Then you don’t have to run around trying to catch and delete things, it’s set up with exactly what you want.
InsideOutSanta 1 hours ago [-]
It's funny how differently people use LLMs. I have agents that run for weeks, building something from a detailed plan with clear goals that gate progress. I don't like to constantly baby LLMs and tell them what to do. It would be inconvenient to rebuild the VM every day, then restore it and start everything again.
rurban 12 hours ago [-]
Went down the same hole, seeing all the massive logs, in plain text and as json, but ever eventually found the real culprit: snap caches of insane sizes.
sroussey 22 hours ago [-]
I just erased 20GB of Cursor’s cruft.
TeMPOraL 22 hours ago [-]
Is that cruft or past conversation history / subagent history and observability logs?
InsideOutSanta 21 hours ago [-]
Just garbage. Temp files, worktrees that weren't removed after subagents finished, build files... all kinds of things.
thdr 21 hours ago [-]
But that's agent behavior, and not harness responsibility...?
klodolph 18 hours ago [-]
Harnesses should be responsible IMO for managing lifecycle stuff, including worktrees and temp files.
It’s simple, deterministic stuff. At least most of the time.
agentdev001 22 hours ago [-]
Harness problem, or model problem?
corndoge 22 hours ago [-]
Let's be real, what do you think wrote the harness?
NooneAtAll3 21 hours ago [-]
> I now have an LLM run once a day to just look through the disk and delete everything that looks like it's leftover garbage
...are you sure it wouldn't consider your personal files garbage?
InsideOutSanta 21 hours ago [-]
"In my dev VM". Everything of value in there is in a git repo.
umvi 22 hours ago [-]
Can't wait until we get really well designed harnesses written in a compiled language that doesn't bloat my computer. Will probably have to be written by a company other than OpenAI because codex is like the exact opposite of everything I just listed
Yeah, using it a bunch with local LLMs. It's fast (small system prompt), focused on the basics, made around simple choices, and possibility to call it in scripts.
So far I have very positive experiences, the results are generally better (and not just faster) than in other harnesses with the same models. I've done the same (user) prompts for tasks in a range of them, and compare the outcome to eval between them. Simple (harness) beats complicated one many times. I do wish it could use MCP though, but not as often as I think, and CLI tool usage often compensates on this front with little extra effort)
That is now my usuall go-to harness for things usually. Still figuring out the edges though (when we get off the happy path, like models reaching their max context, or such...)
It's very light, recommed you try it out, and see if it fits your toolbox.
freakynit 15 hours ago [-]
Just tried. And wow!! I opened my mac's activity monitor to see how much load it was putting on my system, and i had to scroll all the way down ... unlike those current harnesses that sit right on top (sorted by used cpu, descending).
And it worked. I'm switching to this unless i find a reason not to.
serf 21 hours ago [-]
it's okay.
it's minimal and somewhat opinionated, not integrated well with memory subsystems (although that's kind of the point)
my only real criticism is that last I played with it it deals with slow provider responses very poorly , so lots of locally hosted llms require a lot of tuning to get it to handle streaming/non-streaming/timers well.
way less minimal (but still pretty minimal) I prefer Pi; but different markets I think.
hax fit my openbsd laptop well, espescially since Pi and related derivatives are kind of un-spokenly incompatible with the BSDs.
OleksandrC 21 hours ago [-]
(author passing by) Could you please report the issues you've experienced on the repo? I did quite a bit of testing myself on local models (Qwen, Ornith, Gemma, Muse Glimmer) with llama.cpp, and they all worked fine. Details would be helpful.
Also one brief note for Codex users specifically. hax can work out-of-the-box as codex-cli drop-in replacement (since it can use the already authenticated codex token), so it's super easy to try it - basically it will be a zero-config experience.
seanhunter 3 hours ago [-]
Firstly thanks for writing it and making it available. The problem I had when I tried it is it would consistently blow out the small context window of my local qwen and although there was seemingly a config knob I could use to prevent that, seemingly that didn’t do anything.
imrehg 16 hours ago [-]
Thanks for making it :)
crossroadsguy 15 hours ago [-]
That landing page is hard to look at, let alone read. Have LLMs really not yet learned to write webpages that read the sytem theme and look that way? Must be hard.
dummydummy1234 22 hours ago [-]
Codex is written in rust, fyi... Not that I disagree.
umvi 22 hours ago [-]
Oh, I haven't been keeping up. I assumed it was still TS since it prompts me to run npm all the time.
Twirrim 21 hours ago [-]
They just use npm as a binary delivery mechanism, for whatever reason (I guess they expect all their users to have npm?)
judge2020 19 hours ago [-]
It's pretty darn ubiquitous, yeah. Plus enterprises tend to run npm mirrors, so delivery integrates into existing IT compliance frameworks.
lynx97 22 hours ago [-]
Which generates unnecessarily huge binaries.
rfgplk 22 hours ago [-]
It's because they statically linked it, and probably rely on a trillion pointless crates (as is standard in Rust development). A good statically linked harness can easily be ~5MB MAX
does it achieve token efficiency? vs something like maki? It's not at all clear what the method of that "efficiency" is.
my main issues with maki are that the line editor (what you use to enter your prompts) is very barebones, copy-paste is broken (no middle-click paste, or copy to PRIMARY; implicit copying to CLIPBOARD that can only be used with a mouse), and scrollback is not native. So it's not a real cli, but some sort of weblike tui.
I tried 3code; it does not mess with selection, but code printed in the conversation has | markers in every line, and is thus uncopyable in practice. Don't know how to use that.
koe123 22 hours ago [-]
Omp
rfgplk 22 hours ago [-]
Just write your own harness. Ironically Codex/Claude are both abysmal.
theowaway213456 22 hours ago [-]
How do you even write a harness that can use a Claude subscription? I thought it was not allowed by Anthropic's ToS and can get you banned?
_flux 13 hours ago [-]
I don't use Claude, but it appears it is permitted to use Claude as a stdin/stdout -communicated tool using JSON, or use Claude SDK, with the monthly plans. Monthly Max/Team plans also include credits that could be used directly via the API: https://support.claude.com/en/articles/17154008-monthly-api-... .
E.g. OpenCode can be used with all three methods. Looks like it needs a plugin for the two Claude methods, though.
colinmarc 20 hours ago [-]
It's trivial to bypass this by tweaking the system prompt in other harnesses. They use string matching and a blacklist.
You can’t just use your Claude sub on a third party harness without getting blocked even if you use the Agent SDK. Apparently, Anthropic gives “credits” depending on your sub and you can use those _credits_ when you use third party harnesses or Claude -p mode. That usage is not the same Claude sub one would assume it is… It is all quite deceiving. Don’t get a Claude sub unless you’re okay with being locked into Claude Code…
sroussey 22 hours ago [-]
That’s a different discussion.
fasterik 22 hours ago [-]
I suspect a lot of people wouldn't use AI if they had to pay per token at current rates. Yes, it's technically a different discussion, but "just write your own harness" isn't really an answer for the vast majority of us riding on the investor-subsidized subscription plans.
rfgplk 21 hours ago [-]
If you're using the sub, only OpenAI (via the OAuth token) is viable. Claude is API_KEY only
agentdev001 22 hours ago [-]
It's allowed, but you get bumped to per token pricing.
arcanemachiner 17 hours ago [-]
Not true; They rolled back some of the changes you're thinking about IIRC.
bhelkey 21 hours ago [-]
That defeats the purpose of using the Claude subscription.
sroussey 22 hours ago [-]
I did, and the weird things you have to do to make work at least decent. I’ve been doing evals and I got some things better than opencode but not as good as pi. Though my goal is to optimize for a browser and browser based llm so it can create stuff in a web container or at least generate ui by requesting json schema (a2ui style).
joquarky 19 hours ago [-]
They're not made of fairy dust; you can build your own.
nicce 22 hours ago [-]
> People moan about how badly made Claude code is but Codex seems much worse on every metric
Also, it feels that chatgpt.com and aistudio.google.com are the buggiest websites I have seen for long time. Why it is always something related to LLMs?
stravant 7 hours ago [-]
Because LLMs are changing so fast that nobody wants to spend time carefully engineering an efficient setup for them which might just get thrown away in 6 months when the modality changes again.
steve1977 2 hours ago [-]
But are we not told that LLMs are so powerful that nobody would need to spend time to engineer that?
sdcfgy 22 hours ago [-]
All the LLM turds we've curled out at work are the same.
bakugo 22 hours ago [-]
Because they're the most likely to be 100% vibe coded. And as much as some people still deny it, AI code = buggy code.
demibabs 21 hours ago [-]
It’s actually insane how buggy the Gemini web version is (or at least was a few months ago, since I stopped using it).
novafunc 17 hours ago [-]
It's improved. The scrolling used to be terribly glitchy and would make it difficult to read as text was generated.
But in the same fix, they also removed the reasoning-sample, which kinda sucked.
platevoltage 21 hours ago [-]
ChatGPT.com is dreadful. The whole thing chokes when the conversation gets moderately long. It's amazing that it's been allowed to stay this way.
dannyw 16 hours ago [-]
I suspect it’s intentional.
Chat sessions are not metered by token usage but rather a fairly high limit of messages per day or something.
The longer chats you have, the more context and token input. So I think that discourages longer chats, and encourages users to start new chats.
slopinthebag 20 hours ago [-]
[dead]
jonplackett 22 hours ago [-]
It’s extremely ironic that the software that’s meant to end software is itself pretty terrible software.
for some reason I had 11 separate codex and codex-code-mode-host binaries in my ~/.bun/install/cache/@openai . each weighing about ~300 MB. might be worth a look to anyone running low on disk space.
rfgplk 22 hours ago [-]
Yep, .codex/packages/standalone/releases keeps all the old versions. Also each codex bin is somehow ~300MB.
22 hours ago [-]
JodieBenitez 22 hours ago [-]
thanks, I just removed all the old stuff.
wodenokoto 12 hours ago [-]
I mean, they are the premier showcase of vibe coding - can we really expect anything but the finest soup of slop?
I haven't worked with the update subsystem before but that seems like an oversight to me. I'm gonna fix it.
orliesaurus 21 hours ago [-]
what's the lazy way to check? can you share where/how you cleaned up?
icantevenhold 17 hours ago [-]
I just asked Claude where all my disk space went and it gave me a really useful breakdown; turns out i had some unused local models in random folders ~200gb worth.
Plus aforementioned harness bloat plus 100gb of unused npm stuff etc
verdverm 17 hours ago [-]
Don't get me started on my custom NES harness, it generates ~10MB per turn. At least the harness itself is pretty tiny
Onavo 22 hours ago [-]
Even if you installed codex with brew?
icantevenhold 17 hours ago [-]
Not sure I installed it via some random install.sh which they promoted
bayindirh 22 hours ago [-]
Remember: VSCode runs a complete headless Eclipse instance as its Java LSP, so I don't expect quality code from any of the big companies anymore.
Yes, my DEV VM tended to fill up with VSCode backups as well. I finally expanded the disk so it's less of an issue today, but for a while ENOSPC was resolved by cleaning up ~/.vscode-server/cli/servers.
dllu 22 hours ago [-]
The daemon has many bugs. For example, when auto-updating, it would randomly ruin the permissions and block the goal as .git became read-only. Generally, it is extremely annoying how many bugs Codex has. I basically can't trust it to execute on any goal since it would often get blocked every few hours due to random bugs.
Also, in Codex, one randomly can't click "New Chat" or "Scheduled Tasks". You know, the absolute two basic functions.
kristjansson 21 hours ago [-]
The CLI has been pretty good on this score, way better than the App. Has been…
dllu 21 hours ago [-]
My complaint applies specifically to the CLI. I have never used the gui app. Here are the various goal-breaking bugs I've run into:
- Goals get blocked due to .git and other stuff becoming randomly read-only when daemon auto-updates in the background (the linked issue in my comment above)
- Goals get blocked due to transient "Selected model is at capacity"
- Goals get blocked due to bwrap randomly failing with `bwrap: loopback: Failed RTM_NEWADDR: Operation not permitted`. I traced this one to `apparmor` on Ubuntu 24.04 interfering with bwrap and found a local workaround.
calamari7063 17 hours ago [-]
[dead]
reacharavindh 22 hours ago [-]
I’ve been using GPT models via hax[1], a bare bones harness with minimal tools on my system and have the models rely on bash for the most part the last few months. It’s been a breath of fresh air amidst all these overengineered harnesses.
If I need special actions, I just make them CLI tools on my Mac and tell the AI model that such a tool exists(via a single line in its system prompt) and that’s it. It’s been very nice.
I swapped to hax after the atuhor talked about it here on Hackernews, it really is very much a huge recommendation. Most importantly for me, it has been the only harness who respects the XDG specification and does not ask me to install by curling into sh. From my previous to harnesses, pi and maki, neither did both.
On top of that, it also feels very much like a normal CLI tool, which is a big plus for me.
It does not support plugins or MCP, which could both be a big negative for many people, but for now I was able to find a good terminal alternative for everything I needed.
Just a huge recommendation from my side as well, not affiliated in any way myself.
agentdev001 22 hours ago [-]
You seem to care, so I wanna point out a semantic nit; "such a tool exists" could be interpreted as a harness provided tool, rather than a command line tool. (Edit, on a re-read its pretty obvious you mean cli tool, but keeping this statment here as im sure it'll open someone's eyes to the difference)
I've gotta give hax a go, as a pi user who came on-board due to the no frills direction pi takes. I say takes rather than took, because regardless of the (heated) discussion surrounding mcp and code mode support, I cannot fight the numbers- augmenting the harness provided tools is a notable boost to execution round trip performance.
reacharavindh 21 hours ago [-]
I may need to take a second look at Pi again if it has any niceties like that notable boost in round trip performance compared to hax.
I mostly stuck around with Hax because of a few design choices. The sessions are merely plain text files indexed by a dirpath slug. This means I can just start a hax session in that directory and expect sessions history to be available in any model.. (I occasionally dabble in GLM for security checks, Gemini for writing stuff, Deepseek for tightly scoped tests etc).
asa123 15 hours ago [-]
In the end what you said about letting the model know what tools exist is more or less what pi does regarding tools/extensions/plugins.
you can also switch the model at any point too after resuming.
But honestly i dont think there is that big a difference between these “minimal harnesses” to begin with so long as they have the basic functionality…
mglvsky 5 hours ago [-]
can I use this harness with subcription plans without being afraid to banned from violating some ToS?
denverllc 2 hours ago [-]
ChatGPT: likely yes (Tibo tweeted that it was approved)
Claude: probably not (officially against the rules after the openclaw fiasco)
timdorr 22 hours ago [-]
This is the same operational model as Claude Code.
ls ~/.local/share/claude/versions/
ps auxww | grep "claude bg"
I see a few prior versions of the Claude Code binary and several running background tasks, without any copy of Claude Code in the foreground. I regularly get a "Update available" message in my Claude Code UI too.
whimblepop 22 hours ago [-]
This kind of stuff often breaks for applications packaged with Nix, so it's often removed from Nix packages for compatibility reasons. Both behaviors you mention are helpfully patched out of the Claude Code package I use: https://github.com/numtide/llm-agents.nix/blob/dc285bfd94ec7...
As a result, I never knew that Claude Code does a nasty thing like this, either. :-\
Fwiw, packages in that repo get version bumped each day, so if that's a fast enough pace for you it may be a good alternative.
farlight 20 hours ago [-]
If running systemd, wrap that crap into a cgroup and force kill the whole process tree on exit. I've been doing it for years with badly behaving applications. On phone, from memory, may not be exactly what is needed:
This sort of persistent auto closing is just so infuriating. Shows 0 respect for the user and not wasting users' time.
Sha1rholder 22 hours ago [-]
Off-topic: I think the front-end quality of ChatGPT/Codex Desktop and Android app is simply dreadful—hard to imagine how such a large number of frequent failures made it into production. Do they not use human testing at all?
ncr100 20 hours ago [-]
Is there evidence to suggest that open ai business leadership values money over inspiring general public customer delight?
slopinthebag 20 hours ago [-]
probably not much, they're all using workflows where the agents pull issues and automatically merge code when the tests pass. so lots of stuff slips through and the overall experience is kinda shite.
verdverm 17 hours ago [-]
Eyes free coding is a badge of honor for some folks.
I think it's a crazy notion given what I see these things do and generate, and hoped they would too after recent incidents.
slopinthebag 14 hours ago [-]
yeah and it's not even about reading code in particular. i keep a close eye on architecture, project structure, data structures & data flow, etc. and then i do manual testing to make sure things feel right. i cannot imagine developing anything users interact with without doing this.
verdverm 10 hours ago [-]
I have recently hit the point where I am struggling in a vibe coding experiment with the state management and control flow sequencing. It's a very complex scenario (detecting game mode change for Final Fantasy 1 in a NES harness where the game loop and agent loop need to run independently, yet interleave). I'm insanely impressed with how far the open models have brought the project, I would never have on my own, life is too short. It remains unclear if I can work with the agents through this challenge. Probably time I took a deeper look at the code I suppose.
silverbluep 21 hours ago [-]
Codex is very badly structured. I was an early user with codex. But it mixes state and configuration data; which is why I can't really declaratively manage it in my nix-darwin environment; so abandoned it a while ago. This issue being raised, and the way that the codex development responds back to it makes me not trust their harness ever again. I only use codex in opencode; which seems to be doing OK. At least openai doesn't gate their product to their harness like anthropic.
Not trying to sing praises; i think opencode is doing a bunch of weird stuff too but it's not codex level.
dizhn 17 hours ago [-]
Antigravity's agy is the same. I suspect ACP and GUI too. It has no env var or parameter to set its config dir. You can set home but that also set's agent's home. How big of a piece of crap it is is literally hard to fathom.
It doesn't even have a place to check your remaining time or limits except for an in app /quota command that the CLI does not expose.
dingaling911 22 hours ago [-]
This stuff is moving fast, should update automatically. This is kind of standard now. Browsers do it. No reason for harnesses not to do it as well.
isityettime 22 hours ago [-]
Ordinary package managers are totally capable of updating thousands of programs without manual effort, including automatically and in the background if that's a thing you want. But they do it without violating users' sovereignty over their own computers or turning every app into a special snowflake that gets to install software (namely itself) in its own bespoke way.
icantevenhold 22 hours ago [-]
My browser doesnt update automatically, i have to approve it.
Maybe this kind of stuff is standard for some people but I absolutely loathe auto updating software for many reasons
dv35z 22 hours ago [-]
One practical reason - system stability for important events.
I travel for work, and often have a tech demo, solution, tool, or something to show - or am running a workshop.
I usually do NOT update things right before my trip, as I want to minimize the potential for disruption.
vorticalbox 22 hours ago [-]
Isn’t this what docker is for?
Jokes aside I do the same I tend not to upgrade my Mac, my phone unless I really have to.
The “update” button on the top of my brave install has now turned from green to orange and I’ll likely wait until it goes red
TeMPOraL 21 hours ago [-]
> Isn’t this what docker is for?
What? Pinning a browser?
First thing it'll want to do after starting is update itself to newest version anyway.
nicce 22 hours ago [-]
For example, Firefox updates by default and forces you to restart it, disabling using websites or seeing any site content before doing so.
Zambyte 22 hours ago [-]
On your setup.
Firefox does not only not prompt me for updates (it only updates if I manually run an update), even if I do update with Firefox running, it does not stop me from continuing to use Firefox without restarting it. All I did was install Firefox from flathub (packaged by Mozilla (a default configuration provided by Mozilla)).
nicce 22 hours ago [-]
I am running Firefox Developer Edition on macOS, where it is the default option. Have to set "Check for updates, but choose when to install" to get the similar behavior.
RobotToaster 22 hours ago [-]
I've never seen that behaviour, it downloads updates in the background and then gives me a little notification that I should restart it to update, but it's dismiss-able.
nicce 22 hours ago [-]
I think this is the old default option. What options you see in the update options?
djsavvy 22 hours ago [-]
I've been using Firefox for years and don't think I've ever seen this. Are you running some custom configuration?
TeMPOraL 21 hours ago [-]
Standard on Linux, I believe. One of many reasons I got back to Windows as my main OS (with Linux needs fulfilled by WSL).
Iirc, this is a pattern of the flatpack version of Firefox. I do not believe the .deb does this.
rootnod3 14 hours ago [-]
Unless I rebuild my ports, Firefox does fuck all to my system. If it ever did, I would nuke it immediately.
six_seven 22 hours ago [-]
Are you sure this isn't your organization's update policy kicking in here?
icantevenhold 17 hours ago [-]
Not on my machine
judge2020 19 hours ago [-]
> My browser doesnt update automatically, i have to approve it.
Have you ever actually blocked it due to something in the new version, or is it mostly to control the timing of when it updates?
icantevenhold 17 hours ago [-]
Only to control the timing
jstummbillig 22 hours ago [-]
Being able to turn auto-updates off is okay, but that being the default would certainly not be okay in Codex, giving the reach, what it can do and how immature it is.
bayindirh 22 hours ago [-]
You can do this without a daemon, and without storing n previous versions as noted by another user.
jmbwell 22 hours ago [-]
Yeah this seems like a decision yielded by a model. One of countless examples of the model’s outputs careening into individually defensible but holistically nonsensical results
A human seems less likely to embark upon a path that most humans know leads to something a human is unlikely to want
LLMs produce amazing things, but only one token at a time. Anything beyond that is retrospectively emergent at best
yoyohello13 21 hours ago [-]
I guarantee the prompt “add auto update feature” was the first and last human input applied to that feature.
22 hours ago [-]
wtallis 22 hours ago [-]
Serious tools shouldn't do this. Sure, you can offer an opt-in rapid release beta update option, but if you can't go more than a month without forcing me into a "stable" update that changes the UI and behavior I'm getting used to, then I'm not going to consider your software stable and reliable enough to be worth developing expertise in using.
dingaling911 21 hours ago [-]
The entire class of software (harnesses) you're talking about was literally invented like 2 years ago...
wtallis 15 hours ago [-]
I don't see how that excuses anything. It certainly doesn't give harness developers an excuse to be disrespectful to their users. It shouldn't give harness developers any reason to not even try to produce a stable release with a well-designed UI instead of a janky mess that gains new bugs every time you try to use it.
There are contexts where "move fast and break things" can make sense, but usually not when it's other people's workflows that you're breaking.
Just how old do you think a category of software product needs to be before it's reasonable for users to start expecting professionalism from the developers?
dingaling911 14 hours ago [-]
The majority of harness users are the equivalent of boomers using browsers. They don't really care.
22 hours ago [-]
wren6991 22 hours ago [-]
If software engineering is solved they should be able to get it right the first time
Yizahi 21 hours ago [-]
Software Engineering clearly is solved, that's why Google has a tiny animated banner on the AI Studio sidebar (advertising subscription upgrade), which shows a simple animated border around it and utilizing exactly 100% of the GPU to render just that banner. And there is a browser extension removing that styling element (without banner GPU load is obviously under 0.1% as it should be) released way back in May. So the almighty AI didn't manage to fix a major UI blocker for a flagship AI product in half a year now. :)
wren6991 21 hours ago [-]
When I had a Claude account (cancelled ~May time) I vividly remember the web chat having broken editing for backtick-quoted monospace text. If you typed <space>`foobar`<space> the last space would move the cursor backward into the closing quote, and the space would be emitted inside the monospaced region. It was like this for several months (on Firefox). I always found this baffling for a product marketed to programmers.
whythismatters 20 hours ago [-]
Same with their desktop app by the way, absolutely infuriating.
rurban 11 hours ago [-]
You don't understand Google business. It's an ad company, not a tech company. Nobody cares about that banner.
phoghed 22 hours ago [-]
Coding is solved, software engineering still TBD
22 hours ago [-]
dingaling911 22 hours ago [-]
CODEX WRITES ITSELF! DONT YOU PEOPLE UNDERSTAND!!!!!
bayindirh 22 hours ago [-]
So you imply that Codex has a hoarding issue?
Rapzid 22 hours ago [-]
It loves hording issues. Over 15 THOUSAND on GH.
jonplackett 22 hours ago [-]
It gets very attached to itself
TeMPOraL 21 hours ago [-]
Nah, you're confusing it with Autolith.
swatcoder 22 hours ago [-]
Products moving fast doesn't mean that they're moving in the direction I want, or that I want them to move underneath me while I'm in the middle of a project.
Auto-update systems have value but are not a suitable default for professional tools.
crossroadsguy 13 hours ago [-]
Only if OSes have some easy and sanenly robust way of limiting app access in such a granular way that these apps might be as well forced to start behaving.
Like a simple:
- Dropbox:
- Asked Permissions: File Access [{~/Dropbox, rw}]; App Install/Exec [/Applications/Dropbox.app]; Internet Access [*.dropbox.com]
- They can have additional system defined accessible path inside ~/.cache, ~/.config etc and accessible only via their app package names (if they change their signature - not notarised or shit, user gets a notification and approves it) -> this is the tricky part. I'd rather have an app get a ~/.local/appdata/<app/package name> sort of blanket access for anciliary things and do whatever the shit you want do there.
- optional permissions: []
and so on…
(also: the example given above is not fair game for nitpicking :D)
The thing is: the Internet access part one can control with Lulu or other similar apps. But anything else is a user nightmare - next to impossible even for a cli/tweaking/etc comfortable user. And mac's TCC is a royal mess - even their UI part for this is utterly confusing. I think only the Linux world can achieve something sane+easy like this (if it doesn't have that already; though they have not been known for being "easy").
I have achieved a lot of these for harnesses using sandbox-exec (by basically doing: "deny *; then selectively allow") on mac but it was a pain and I can't even think of trying something like that for "normal" apps.
RobotToaster 22 hours ago [-]
Then use the package manager, or windows store on windows.
johnnyApplePRNG 22 hours ago [-]
[flagged]
dingaling911 22 hours ago [-]
you run it on the bare metal?!! Balls. I run it in a container
piker 22 hours ago [-]
It's not necessarily hostility. We considered different updating approaches with our legal IDE where security, stability, etc. are paramount, and having a buggy application out in the wild is user-hostile. Sure, in a dream world we'd say "don't ship until it's perfect", but that's just not realistic.
We considered the daemon thing, but ended up with a hard-update check on every boot. It's thus a user choice: if you don't like the automatic updating, you can just not run the application anymore.
And even with all that, it's hard to stay on the right side of AV algorithms, etc.
> It's thus a user choice: if you don't like the automatic updating, you can just not run the application anymore.
That's not a choice. It s just what most apps do these days, but it's the first time I've seen anyone stating this so directly, or trying to frame it as a choice.
(And behaviorally, this makes your product just a SaaS website with extra steps.)
piker 21 hours ago [-]
I guess I mean it's a choice in the sense that it's a transparent part of the product that one chooses how and when to use (or not). That's opposed to something running in the background on its own schedule, with its own agenda whether the user is running the application or not.
[Edit: yes, that is kind of the idea. Native version has better features because it can access them.]
thebeardisred 22 hours ago [-]
It's right about now that having a operating system distro model is great - you get folks who will package the software & test it out, make generally sensible defaults, and generally keep shenanigans like this from going on
Cyph0n 22 hours ago [-]
Nix/NixOS takes it a step further and makes you have to go out of the way to enable self-updating behavior when packaging something :)
dv35z 22 hours ago [-]
What are the top ones you recommend? ( I am running Mint Debian Linux )
esjeon 22 hours ago [-]
I think the quality of Codex CLI (and Claude Code) is an honest reflection of the status quo of LLM-driven development. The future doesn’t look really promising.
qoez 22 hours ago [-]
To be fair we've gone from barely capable at writing code to what we have today in barely 3 years. I wouldn't put money on the counter bet that they won't get beyond this hump.
afavour 21 hours ago [-]
I wouldn’t assume that progress is linear. Maybe I’m wrong and it will be! But I’m reminded of early developments with self driving cars, it felt like manual driving would be a thing of the past within a decade. But progress has been much slower than that.
I_am_tiberius 21 hours ago [-]
> But I’m reminded of early developments with self driving cars, it felt like manual driving would be a thing of the past within a decade. But progress has been much slower than that.
I think that's exactly the standpoint of Yann LeCun.
programjames 20 hours ago [-]
[flagged]
benrutter 20 hours ago [-]
No need to be rude! Besides OP just said:
> I’m reminded of
They were drawing a comparison, not saying "x will definitely happen because y"
tenuousemphasis 20 hours ago [-]
No need to be such a dick, it reflects poorly on you.
jknoepfler 20 hours ago [-]
You committed your own fallacy of applying a sloppy heuristic to understanding the person you were responding to without bothering to check if it applies. Spectacular. The comical levels of arrogance that infect over-hyped tech marketing boondoggles (one area among many in which GenAI is clearly similar to self-driving cars) is one of the reasons nobody trusts the hype anymore.
The breathless astroturf articles, the brain-dead talking head conversations, the endless flood of obvious misinformation about industry uptake, the nauseating LinkedIn puff-pieces about PREPARING YOUR ENTERPRISE OR ELSE... it's a hostile information economy at best, and its being driven by the same people who brought you bullshit like self-driving, crypto and the Metaverse... for the exact same reasons (to get rich).
There's no epistemic credibility left in the tech industry in general. Maybe the Dario's or whoevers of the world believed what they were talking about at one point, but I don't believe a single word I read from Anthropic or OpenAI for the same reason I don't believe anything I read from Russian state media. I know it's being produced for reasons that have nothing to do with the truth.
As someone who interacts with the technology on a daily basis at a non-trivial level... I'm not seeing anything that would cure my skepticism about GenAI. The information economy surrounding it is broken. The fundamentals don't line up. The hype about "nearby capabilities" is identical to the hype of generations after generations of hype in the AI space (dating at least back to Turing). It smells like a bad bet.
apsurd 21 hours ago [-]
The product and UX of the tool is distinct from the raw ability of the tool.
ctxc 21 hours ago [-]
Claude code has great UX.
gwking 21 hours ago [-]
It is hands down the most user hostile UX I put up with. The fact that I put up with it speaks volumes about the value proposition that the underlying models bring to the table.
I’m in the process of dumping codex for more open harnesses that don’t add problematic new features every week. Because Anthropic insists that I have to use their harness with the subscription, eventually I’m either going to wrap `claude -p` to get away from the crappy TUI or else I’ll just unsubscribe and use other models. Their bad UX has alienated me as a paying customer.
ctxc 21 hours ago [-]
Can you give me examples of things you don't like? Because I unironically admire how they improve UX. It's subtle, minimal, yet insanely useful.
Listed a few in my other comment
arcanemachiner 20 hours ago [-]
Try using the "rewind" feature to go back to a previous message.
Went too far? Oops, it's not accessible from the Claude Code UI anymore! You can get the agent to dredge up the info from the session logs, but you'd better hope they put the pieces together correctly.
Also, you can't rewind to a steering message. Want to make sure you can rewind to it? Wait for the agent to settle, or stop them, enter your message, and hope you don't throw them off their current train of thought.
To see a proper rewind UI, see how Pi does it: Every message is available, forked conversation history is preserved.
It has its own issues, but is miles ahead of Claude Code in this regard.
apsurd 20 hours ago [-]
I'm not who you're asking but for me I am right now going through the insanity of providing plugin/mcp/tool access across environments. This is in codex which as I mentioned is intentionally trying to provide mass-market ergonomics. But with that comes incomprehensible configuration that's also a moving target. chat/work/codex/remote/cloud are all different things. And each client exposes different abilities and hides others. And looks like only the web-ui is the most up to date because other targets need releases. As a developer I understand why it's in its current state, but that in no way excuses it as acceptable ux.
Like of course codex is the most powerful and "just" run `npx blah blah` but no, I'm not running randomly piped code on my personal computer and even if I didn't care about privacy it means I can only run this locally with my laptop on. "leave your laptop on and access from your phone" is not a good idea. It only adds to the surface area of confusion. Ux no good.
toonvanvr 21 hours ago [-]
No, the tooling I make has better UX very quickly. It is built for their financial goals, not for the end user.
carterschonwald 21 hours ago [-]
this 1000x. a lot of details in context compaction, editing, reverting ux, etc are all really centered on keeping the load and cost of subscription users as small as possible. its no always that nor is it necessarily correct, but its a huge driver of how those subscriptions and first party apps are designed
ctxc 21 hours ago [-]
I know this is an unpopular opinion, but it's actually good.
- pasting lots of content became [35 lines] and they added a tiny notification to show it as-is if you paste again
- interactive questions layout
- paste images
- the recap
- auto-updating session titles
And so many more. I tried codex and it was lacking in many UX aspects. Remember we got from a text-paragraph-only terminal to this at breakneck speed.
wtallis 14 hours ago [-]
Half of what you're listing as features people should be impressed are stuff that causes as much frustration as it solves. The image pasting behavior is a super-annoying footgun, because sometimes I just want to tell the LLM the path to the image file without it trying to burn a ton of tokens on computer vision—and dragging and dropping a file into the terminal pastes the path in any normal context. The auto-updating session titles, recap, and notifications are token-burning distractions that are glitchy; I had to disable them when I found that having the agent launch an app would mean I'd get a random beep about a minute later when the agent had been waiting for further input in the background.
And they can't seem to go two whole months without scrolling breaking in some new way.
asgraham 21 hours ago [-]
I’m not deeply pro-codex, but it does have all those features. Who knows since when.
Claude Code is better than Codex UX. But as per the sibling response it's more a developer tool made by developers for developers. It does have a lot of niceties. Codex/openai on the other hand is more clearly trying to reach mass-market and it's a UX shit show because of it. Consumer UX is its own thing. So the original point is that raw ability of the tool doesn't magically solve ergonomics.
ctxc 21 hours ago [-]
Won't disagree with you because I haven't used codex/OAI a lot - but Claude's non CLI UX is great too.
Eg - I asked for a recipe and it says leave x for y time - there's actually deeplinks that link to your clock timer with the time preset! And to change quantities, there's input boxes in the response that update all quantities accordingly.
Many such cases of use-case specific UX
aschobel 21 hours ago [-]
agreed. the thing i reach for most is its ability to coordinate my agent graph, e.g. seeing what sub-agents are doing, messaging agents, adding new capacity. claude code really is a delight to use here
rhdunn 20 hours ago [-]
The current approach of the big companies is to keep increasing the number of parameters (there was a report that recent Claude models are on the 4T parameter scale). Even recent SOTA open weight models are in the 1T-2T range.
This is partly due to the multiplicative scaling factor -- the number of tokens; the number of transformer layers; the number of features; etc.
I suspect that for models to have a step upgrade either/both of 1) a new/improved architecture model; and 2) better training of smaller parameter models are needed. I say the latter as I think that with the larger weights there is a lot of unused/redundant parameters.
wonnage 21 hours ago [-]
If we’re learning anything from the past 15 years it’s probably not to extrapolate AI progress in a straight line.
Progress is spiky and not guaranteed. The capital costs have also ballooned to the point where the rate has to level out
dylan604 21 hours ago [-]
> The capital costs have also ballooned to the point where the rate has to level out
That's a pretty bold use of "has to". None of the financing behind AI seems sane, yet it continues to defy rational expectations. Reminds me of the comment "the market can remain irrational longer than you can stay solvent". It's not the exact meaning but has the same feeling. The circular financing going on does not help anything either
bhaak 21 hours ago [-]
Time doesn't solve bad designs. We still have lots of buffer overflows in C.
We don't see any indication that the LLM companies are slowing down and go for better code.
Instead it just looks like a cash grab until the AI bubble bursts.
lifeisloving 21 hours ago [-]
I dont think the quality has gotten that much better. Take claude 4.5, the probability of getting usable outputs was less likely than with opus 5.5, and the usuable outputs weren't great.
I think newer models are more likely to produce the usuable output, almost always, but its still not great. The frequency in which you get usable results leads one to believe that quality is improving.. No usability is improving but the result is still meh.
Also put an llm in an environment without an lsp or compiler to check against and let me know how capable you think they are on their own. The "loop" is doing a lot of heavy lifting here.
monster_truck 21 hours ago [-]
I say this about 30 times a day, to nobody. "How the FUCK is this all so bad? They have unlimited tokens"
Jimmc414 21 hours ago [-]
I think it’s a new spin on Brook’s Law. Adding tokens to a project makes it less efficient not more.
chuckhoupt 21 hours ago [-]
The "Mythical Man-Month" is now the "Mythical Token-Minute"?
duckmysick 19 hours ago [-]
As an aside, I often wonder how would classics like Mythical Man-Month or Project Phoenix would be (re)written today.
r_lee 21 hours ago [-]
that's what I'm also wondering about.
in the last 2 weeks or so:
- scrolling stopped working
- subagents no longer work (main agent doesn't get notified when they finish)
- subagents don't work (a subagent can't make its own agent even if instructed to do so)
what's going on? I thought coding was solved?
tasuki 21 hours ago [-]
> They have unlimited tokens
That is exactly why it is so bad. Tokens spent generate code. Large amounts of code is a liability.
abighamb 22 hours ago [-]
It has pretty rapidly reached the point where I do not think any acceptable LLM-harness is widely available. Every single one I've tried has broken in bizarre ways on even minor updates.
entropie 21 hours ago [-]
I used codex the first time yesterday and quickly decided I need to sandbox it and came up with a pretty easy and straightforward solution (I think).
I use nixos and use a little wrapper that wraps Codex CLI in Docker. Run codex-box /path/to/project and it starts or reuses a container with only that project bind-mounted, keeping login and history in a separate volume. It uses tmux for detaching and reattaching
I was delaying it but one day Codex+GLM went ahead and did things in my ~/.dotfiles folder which I had granted only READ access for. I asked and I was given a reason which I didn't understand and didn't really want to. Right then I used sandbox-exec/safehouse (macOS) and tightened every harness installed (and the one not even installed, lest I install it later and forget to add/track them in the sb zsh init script).
But I am not satisfied. I am trying to find a really lightweight container/VM for macOS that eats very less RAM on its own and expands as needed. Because I really don't want to run an agent outside a container on my mac.
entropie 2 hours ago [-]
The script is working like charm. I updated the packages so it has more access to tooling (ffmpeg, rmagick etc)
natpalmer1776 21 hours ago [-]
Build your own, using the broken harness
Maybe thats just the tool nerd in me that just wants the perfect tool for every job.
rnxrx 20 hours ago [-]
Very much this... The utility of Claude has grown exponentially with the amount of work I've put into memory/context management, tooling, skills development and CLAUDE.md (and friends). I freely admit I may just be fortunate, but I've yet to hit a bug in a new release that rose above the level of minor annoyance.
The thing that really validated this point was when I recently tried Codex again. Out of the box it was OK, but not great. I swung over the memory apparatus and had it import directives from my existing Claude install and it was almost immediately quite useful in its own right and, especially, as a counterpoint to Claude when dealing with some fairly complex operational issues.
mpalmer 21 hours ago [-]
Pi as well? Comparatively it's quite a small surface. I haven't seen issues when I upgrade.
NewJazz 21 hours ago [-]
Pi went searching across my entire home directory unprompted... Idk to do that without some kind of sandboxing builtin is rough and chews through tokens on things I don't really want polluting my context.
ghosty141 20 hours ago [-]
A coworker of mine has a pretty neat script that uses bubblewrap to wrap any application. I'm using that one too for pi and it's pretty great since you don't have to trust the harness and can quite easily customize the sandbox to your needs, e.g. disallow reading to certain directories etc.
Well yeah, it's YOLO-by-default by design. That's not related to upgrade churn.
crossroadsguy 12 hours ago [-]
Yes, Pi as well. All these harnesses have gone off kilter and are all for the LLM, by the LLM, and of the LLM at this point.
Maybe it's like online services/tools. Use them until they are VCed and gone pivoting and then move to another tool.
19 hours ago [-]
22 hours ago [-]
flir 21 hours ago [-]
Both CLIs are getting worse. I think it's because they keep bolting in unconsidered "features", not because of the code quality.
But maybe being able to code up a feature at the drop of a hat has turbo-charged their ability to make bad UX decisions.
bakugo 20 hours ago [-]
I believe it's all related. They keep adding new, badly thought out features because it's much easier than improving existing features built on low quality code.
The latter requires modifying existing, badly architectured and tightly coupled code that will almost definitely break something somewhere else if touched, so what do you do? Just keep adding new code on top and hope for the best.
This is not a new phenomenon by any means, but AI amplifies it by 100x.
whimsicalism 22 hours ago [-]
I find codex CLI to be a pretty high quality application. Is the existence of a backend daemon what you're referring to ?
michaelt 21 hours ago [-]
I've used the CLIs for Codex, Gemini and Claude and they all give off the feeling of being on an very rickety, very fast runaway train.
You install an update. Oops, they added a sandbox that doesn't work inside a docker container, it's stopped working.
You install an update. Oops, you can't log in inside a docker container any more, it's stopped working.
You install an update. Oops, they started catching the copy+paste shortcut and replaced it with one that doesn't work.
You revert a broken update. Oops, the update changed the format of your login details, the old version can't read them, gotta log in again.
You install an update. Oops, can't scroll with the mouse wheel any more.
Not showstopping bugs (if you're used to persuading broken oauth logins to work) but it certainly makes updating an "ugh, what have they broken this time?" experience.
contingencies 20 hours ago [-]
You got further than I did. I checked them out and went back to manual prompting. Just don't have a need for codegen that much, and when I do I'd prefer to be in control.
mattmanser 21 hours ago [-]
To be fair, how many products this old even have a revert update mechanism?
warpspin 22 hours ago [-]
It's buggy like hell once you do not run the default settings. Whole permissions system is completely broken, for example.
demibabs 22 hours ago [-]
It’s kinda vibe coding in a nutshell. Lots of cool and varied features, that often just don’t quite work right.
ultrarunner 20 hours ago [-]
It does some neat things that, ultimately, are super obnoxious. Background animations and "sparkle" effects that make my terminal indicate activity at all times. Just subtly broken behavior.
crossroadsguy 12 hours ago [-]
So is OpenCode. More so. I use Claude Code and OpenCode almost equally and Codex slightly less and Pi the least of them (not that it's worse, but because I use it for "non coding" purposes). They are all bad. Even this newer breed of "agent orchrestrating" tools.
One goes to Github and one notices, the dev is done by LLMs, issue triage is done by bots, and the commenting is done by the bots. (btw, the triage mostly means closing them or playing the "this is dupe of that" game while noone gives a flying fuck about any of the issues). Another good one is "not planned". You report a breaking/critical change and "not planned". What demorlised me is attitude of OpenCode devs/maintainers. I think the Open in it more of the "Open" in OpenAI than in FOSS. That's when I started reducing my dependency on this harness whihc was slowly becoming my pretty much full time harness.
yearolinuxdsktp 11 hours ago [-]
OpenCode is barely usable with local LLMs. Plan mode to build mode switch and lose all your prefix. Compaction throws away your prefix (just fixed in v2... I guess v1 gets to stuff it). Pi throws away your prefix during compaction too. I think all these tools' devs and users use it with hosted models. Pi doesn't come with basic permission capabilities--YOLO, write your own or hope the extension you find in the Pi extension store isn't an absolute dumpster fire.
Claude Code turned on the remote-control server by default with an update (no explanation or notification). BUT Claude Code didn't break the expectation that the hooks and the commands you run inherit your environment. So it was hard to notice.
Whereas when Codex App server became enabled by default, and as per OpenAI, it was never contractual that hooks run in the client's environment. Sure, let's break long-standing expectations of where code runs when you run a CLI--it was never contractual, you see. Now you have to figure out how to turn this crap off, just to UNBREAK your workflow!
This is why it's apples and oranges here, with Claude Code and Codex.
crossroadsguy 11 hours ago [-]
> hope the extension you find in the Pi extension store isn't an absolute dumpster fire
That's why I didn't mind it when Pi added native MCP capabilities. What dismayed me was they didn't other tools like subagents, and other stuff. I mean why do I have to scour the Interwebs to find something that doesn't possibly screw everything up for me.
I have a feeling harness usage will finally move to lots of harnesses with usage clusters - as in not X million harness for X million devs/users but a lot and people will use what suits them, and if not a more inclined coder will fork/make one just for them. In all their faults LLMs have made this a possibility.
sdcfgy 22 hours ago [-]
Totally.
The problem is hardly anyone gives a shit if they can pump out stuff that makes them look like they are good at stuff.
I think the (lack of) quality of these tools shows that software engineering is maturing as an engineering discipline.
We were overbuilding trivial parts of the software like the Romans overbuilt small bridges.
Modern engineering is about making the thing that barely works. If you disagree, learn a little bit about bit narrow the tolerances of modern cars and bridges are.
worewood 20 hours ago [-]
> learn a little bit about bit narrow the tolerances of modern cars and bridges are
Buildings typically use a safety factor of 1.5 to 2.0, and cars typically use a safety factor of 2.0 to 3.0. These things are being built twice as strong as they need to be. I wouldn't call that narrow.
And with software, we're at a point where it can't even properly do what it needs to. If we were overbuilding it, damn software wouldn't crash so much.
scheme271 20 hours ago [-]
I think all engineering has safety factors of 2 or more. Maybe with the exception of aerospace (mostly space) where materials might be at the limits of what's possible right now so the safety factors are much reduced.
fooker 19 hours ago [-]
> I think all engineering has safety factors of 2 or more.
Not even close.
For example, a BMW X5 weighs about 2200 Kgs without passengers, and it's GVWR is 2800 Kgs.
If you have five people and heavy bags, you can comfortably exceed the GVWR. You can absolutely break control arms and suspension components in that scenario if you go over a speed bump or hit a pothole (which are real things that exist in the real world, like people doing something unintended and crashing software).
If this needed a 2x safety factor, there would have been no way to build this car.
Even the toughest cars like land cruisers are limited in this way.
gumby271 20 hours ago [-]
I think there's a difference between something that's engineered and precise vs something that's rickety. Just because you built something that might fall apart at any moment doesn't mean it's been engineered with narrow tolerances.
fooker 20 hours ago [-]
I disagree.
> [...] vs something that's rickety
A 100 percent of modern bridges are rickety. There has been dozens of high profile incidents because of this.
> Just because you built something that might fall apart at any moment doesn't mean it's been engineered with narrow tolerances.
Here the software they are building falls apart in ..say.. two days of use, but can be updated in 5 seconds at close to zero marginal cost. The software is useful enough that people keep updating and using it. This iteration gives them a competitive advantage over the safer choice of rigorously testing every new thing for several weeks.
bb-connor 20 hours ago [-]
I don't disagree but it's clearly a problem that already existed for decades in software and is just being 10x'ed. how many different frontend frameworks or AWS regurgitated and enshittified cloud services did we have to endure over the years.
neya 21 hours ago [-]
There is actually even worse code quality written by humans, is that small sample also a reflection of human-driven development? And therefore, the future doesn't look promising that way either?
jdiff 21 hours ago [-]
The difference is the source of the sample. This is a flagship product by a frontier lab. It should be reflective of the best that is possible. It should be biased in the most favorable possible way. It was designed by people without the normal limitations and quotas of outsiders in building it, as well as in learning experiences in driving it. This is not a random cherry-pick of bad code.
neya 19 hours ago [-]
Again, the same parallel can be drawn by frontier tech companies. Case in point: Even before AI (and even till date), iOS has always been buggy, crappy. Apple was a trillion dollar company when they had some of the worst bugs in iOS launched to public.
The same story for Google's products - their products were always buggy even before AI. And so is the case for Microsoft.
I think the reasonable argument is that AI or not, software problems are hard problems to solve. And software is constantly evolving. AI isn't a magic wand. It also depends a lot on the humans using it, including prompting and context supplied to it.
Drawing a massive conclusion from a tiny, unrepresentative sample is just hasty generalization.
techgnosis 21 hours ago [-]
Seems like its just a bunch of young engineers making bad architecture decisions. Not sure I would blame the culture of vibe coding everything. It might just be bad taste.
JohnMakin 20 hours ago [-]
If you blindly trust frontier LLM architecture 'decisions,' they often do look a lot like what a young/junior engineer would do.
switchbak 20 hours ago [-]
The restraint, rigour and responsibility you showed prior to LLMs relates directly to how much of that you apply now that they’re here.
Given how rare it was previously, and how little it was cultivated by shortsighted management, I’m not surprised with all the slop we see today.
The truth is there’s nothing stopping someone from using these tools to remove dead code, simplify project structure, remove dependencies, etc. but I don’t see it happening very often. I have done it a bunch myself though so it’s absolutely possible, it’s all a matter of investment and prioritization.
thousand_nights 20 hours ago [-]
i'm personally quite perplexed how sloppy the harnesses are from the leading AI labs, you'd think these are their primary products and would have some human attention devoted to them
it's embarrassing when every claude code release has obvious bugs in it like telling you to upgrade when you're on the latest version or codex opening visible powershell windows for every command that pop up on your screen
like 5 minutes of human testing would catch this obvious shit but obviously they don't bother because vibes are the future and stuff? idk
cryptolobster 20 hours ago [-]
I'm inclined to agree
rsoto2 21 hours ago [-]
Solar power has become a real transformative technology and instead of trying to recreate a better world we tariffed them and the execs of this country spent all our coins on power hungry data-centers that write mid code.
Rapzid 22 hours ago [-]
I'm sure if you file a bug on the 15k backlog they'll fix it eventually /s
Legit bugs still open 6 months on.. It's a bit ironic
gruntled-worker 21 hours ago [-]
You can just copy the executable then run it with --no-daemon, or edit the config equivalently. No autoupdates and no background process.
The recent terminal UI changes were more fastidious, but even those are easily disabled with CODEX_TUI_DISABLE_KEYBOARD_ENHANCEMENT=1 and the /tui command.
Alifatisk 21 hours ago [-]
Is anyone using the vscode extension of Codex? Its insane how brittle it is.
Just a couple of days ago, it was completely broken for about a week before it got somewhat fixed again. When I say fixed, I mean basic functionality is working again but bugs and regressions are still left. There are multiple issues tracking the same problems. The community have even started to patch the extension themselves as a workaround until its adressed.
You know whats even crazier? There are like a new release every other day with lots of commits and changes, yet it still in a bad shape. I have turned off autoupdate for Codex extension for this reason. The only extension I’ve had to.
I mean, it doesn’t take that much work to have some tests in the pipeline to make sure the extension continue to work, it almost feels like they are purposely doing this.
I have now experienced the consequences of ”move fast and break things”. I realize how much software quality means to me. Its not fun being on the receiving end, especially when it affects my daily workflow.
If any of you, see anyone hype up LLMs again, please refer them to the state of Codex vscode extension.
pluc 21 hours ago [-]
It's as if there was a reason the whole industry was obsessed with quality up to like 3 years ago
clutter55561 21 hours ago [-]
My 2-cents: never install those things into your host. A locked down Docker container is the bare minimum. Pin your versions and never update them without checking the changelog.
pkulak 20 hours ago [-]
Well, OpenAI at least is fine with you using whatever harness you like. Having a crappy first-party one is bad, but an escape hatch puts a floor on it.
ncr100 20 hours ago [-]
Continuing to be glad I trust my instincts around bad management making problematic products!
grimgrin 22 hours ago [-]
while this points it out, i'll show you what I had in place:
~/.codex/config.toml
[features]
daemon_auto_start = false
hell, I also have this redundancy configured in bash:
codex() {
command codex --no-daemon "${@:-.}"
}
umutsevdi 21 hours ago [-]
I discovered this while benchmarking the tool I'm developing. I realized Codex was surprisingly fast after the initial run. Some other harnesses do similar things, so I decided to use bwrap around these tools for accurate benchmarking.
thoughtpeddler 16 hours ago [-]
This is why I use all OpenAI products through the web (except where there's strong sandboxing e.g. Apple iOS). Makes me sleep a lot better at night.
vardalab 20 hours ago [-]
Cool thing is that I can just ask the llm to modify the harness to what I need.
I discovered this daemon thing about two weeks ago and had codex modify it's own custom launcher that I have to bypass the daemon thing. If as a recreational code farmer can do it then most "pros" should be able to handle this and other inconveniences. Cc has been pushing that no scrolling feature that codex recently introduced for a half a year now but again I had them put a stop to that. Software has never been this easy to configure.
AndreasBackx 21 hours ago [-]
I feel like so many of the new AI tooling is reinventing the wheel. We've figured out how to update software via package managers. Do we need a daemon to continually check for updates and do we need shell scripts to circumvent package managers?
Have a look at https://chatgpt.com/codex/install.sh. It's managing the installation by itself not using existing package managers. There's a good point to making it easy to install, but it leads to managing an installation when you shouldn't need to. Time would be better spent elsewhere and better solutions exist.
Am I the only one thinking this?
isityettime 19 hours ago [-]
It's taking a cue from desktop apps on Mac and Windows. But I agree. I hate self-updating behavior in both contexts.
jbverschoor 21 hours ago [-]
Sandbox everything if the OS is not doing what it’s supposed to do
Interesting. I have a very similar setup for myself.
21 hours ago [-]
22 hours ago [-]
quotemstr 22 hours ago [-]
Refusal to adhere to the XDG specification is a big signal that the app is going to be a general bull in a china shop. It's not that hard to do OS- and distribution-appropriate package management.
whimblepop 22 hours ago [-]
A well-behaved application should not only use XDG on Linux, but it should use platform-native conventions on macOS as well: "~/Library/Application Support", etc.
quotemstr 14 hours ago [-]
Yes, and the Windows app-data equivalent. The point is that you have to be a good citizen of the platform on which you're running. curl | bash is kind of a selfish model.
infinite_spin 14 hours ago [-]
I expect this sort of thing when installing/running the client, but I can see why it would be alarming for a CLI installed via NPM to do this
rootnod3 14 hours ago [-]
Slop coded CLI(!) via NPM(!)...I think that answers itself.
ilitirit 21 hours ago [-]
I use it at work. On Windows it also launches a new console window every it executes (?) a command. You should definitely disabled this using --no-daemon.
I don't know how and why they ruined this CLI with needless "features". It was perfectly usable two weeks ago.
sigotirandolas 19 hours ago [-]
A bit off-topic, but on the rare occasions I help some acquaintance with Windows I always feel a bit upset that random split second console windows still pop up when installing everyday software, even that from Microsoft.
I remember them from way back on Windows 98 and no one seems to care to hide them (?).
kristjansson 21 hours ago [-]
not to mention the daemon mode breaks a bunch of assumptions one might have about about codex CLI e.g. that codex CLI inherits the env vars of the shell that invokes it.
—no-daemon is the way, but still feels like the noose is tightening …
elAhmo 20 hours ago [-]
So, similar as Google and Microsoft have been doing with all their tools?
ranger_danger 21 hours ago [-]
In some circles this is called a backdoor.
soletta 11 hours ago [-]
Deleting codex and unsubbing. Thank you, OP, for highlighting this important PSA.
warpspin 22 hours ago [-]
All I'd want is a CLI with a few tools AND a built-in, properly configurable and working sandbox. Seems that's too much too ask.
phoghed 22 hours ago [-]
Just be the 10th person to make one this week.
zombot 6 hours ago [-]
Talk about your box not being yours any more. First Apple and Microslop take it away from you, and then the service providers pile on.
sva_ 20 hours ago [-]
Reminder that OpenAI allows you to use the Pi Coding Agent, which is a lot more lightweight, and in my experience much better than any of the proprietary harnesses.
I've been using Pi nearly every day for 4-5 months now, and even though it is still very much in development with new updates arriving nearly every week, it has to my mind never regressed. I've never lamented that a Pi update ever broke anything I depend on or made anything worse.
I immediately adopted the new tuiMode "fullscreen" feature the moment it landed, and it's been exceptional. I love the way it now manages the viewport -- this was also right about the time I migrated to Ghostty so that also made a big difference. Showing inline images with the Kitty graphics protocol has been a huge improvement too, I love it when a model calls `read` on an image file and I get to see the results in my terminal. The OSC-8 hyperlink support is super cool too. I love the way it auto hides/shows the scrollbar now too, and recently they've even made it so concurrent tool calls show different "themes" when pending/running vs success vs error.
hackerbrother 21 hours ago [-]
Claude has been doing this already.
Garlef 21 hours ago [-]
ah... so this is why my agents all made a signal noise when i used chatgpt on my phone?
carlosjobim 21 hours ago [-]
Claude also now hacks your MacOS to put an "Ask Claude" item into every right-click menu. You can remove it through the system's keyboard preferences.
jongjong 13 hours ago [-]
100 years ago, if you had told people that everyone in the future would be summoning 'persistent daemons' inside black boxes using something called a 'Codex CLI', they'd have thought we went crazy and that the future is a horrifying place.
cromka 18 hours ago [-]
It's off topic, but this weekend I caught both Codex and Claude accidentally read some credentials in config files and not flag them. When confronted, they both went "oh yeah, that, I should have mentioned that".
Except in the past they both would automatically stop and fire an alarm almost. So something clearly changed and I think they just want us not to notice, to avoid damaging the trust in their tools.
I seriously feel at loss with this current AI situation: can't live without them anymore, hate their product still and feel taken advantage of. Local LLM still to expensive and hard to see any light in the tunnel.
20 hours ago [-]
HackerThemAll 22 hours ago [-]
Yeah, and for the contrary I spotted Claude app being ugly. When I turned off Claude, then put my computer to sleep and later woke it up, it made my fans spin loudly, and after I checked, there were like 2 dozen Claude processes and a lot of Python. What was it doing - I have no idea, but it was crazy. I'm setting up a VM with nothing but Claude and necessary tools. I feel Claude is not honest with everything it is doing.
whythismatters 20 hours ago [-]
Today I used Claude Desktop after not having used it for a while. It ran bash commands on my machine, like starting brew services etc. It did not have that capability before (I used to hook up some mcp to allow selective file editing), and nobody warned me that it currently has. Shady tactics, very invasive, freaked me out.
dv35z 22 hours ago [-]
Can you share some more detail of how you would implement this? Are you thinking something like VirtualBox (?), running a virtual machine, and you're using - for example, OpenCode terminal in the VM...?
If you want something not NixOS-related, smolmachines.com (I haven't tried this yet) gets mentioned here a lot.
TZubiri 16 hours ago [-]
A brief story of these agentic harnesses, from the perspective of a dev that barely used them.
I remember when opencode was out, I installed it as a separate user and tried it and it was cool.
Then Copy.fail came out, and I worried that a separate user really wasn't enough, but I felt I was overworrying at this point.
Then it worked fine and I felt silly for going through the trouble of a separate user even.
Then the opencode guys started self developing the tool itself and it became slop, this was like 2 months into the fad, and it was clear the quality merited uninstalling the thing, even if it did some cool magic ( https://github.com/anomalyco/opencode/issues?q=is%3Aissue+au... )
Then Codex came out, and uninstalling opencode became the optimal move, if only because it removes a third party. It's one of those things where people will continue to use it out of inertia (either legacy code, or people reading/watching old tutorials), like using Axios in 2026 after Fetch was made standard, or using Mono after Microsoft released multiplatform dotnet.
Everything was good for some time, but then it quickly became obvious that OpenAI wasn't developing this tool, but rather letting it build itself, which results in a lot of slop (High RAM usage, inefficient looping that only works out of sheer GPU power and tight token limits, resulting in Skinner-box based development)
Then I uninstalled Codex, but this Pi thing came out, from the authorship of the dude from Flask, but I'm tapping out here, then Jev came out and I'm just not playing this game, good luck to everyone who is using these things.
It feels like this is an attrition game where we are waiting for the bubble to pop, but the bubble is made very long to tempt people into taking the bait. I'm betting the house on the day of reckoning, and hope that those that didn't take the AI bait, (or took the AI pill and repented) will be one of the few players left with a reputation.
I remember going in interviews and people asking me if I could use these tools, they wanted to see what projects I made, but I wasn't comfortable even showing a Chess Engine thing, like I didn't make this, this is not what I'm going to be delivering professionally, I'd rather be unemployed for a couple of months or years if that's what it takes, I'm not vibecoding for a salary sorry.
wrinfo 1 hours ago [-]
[flagged]
bsoooh1 41 minutes ago [-]
[flagged]
zvictor 9 hours ago [-]
[flagged]
Wittie 8 hours ago [-]
[flagged]
popkoren 11 hours ago [-]
[flagged]
kydanet 20 hours ago [-]
[flagged]
favori995749721 14 hours ago [-]
[flagged]
mohamed01121986 12 hours ago [-]
[flagged]
pathly 9 hours ago [-]
[flagged]
3amar404 22 hours ago [-]
[flagged]
mococa 22 hours ago [-]
[dead]
uechan0525 14 hours ago [-]
[flagged]
talkarimli 19 hours ago [-]
[flagged]
mococa 22 hours ago [-]
[dead]
phoghed 22 hours ago [-]
[flagged]
cyanydeez 22 hours ago [-]
this is specifically about a "trillion" dollar company and the boot it needs you to lick to justify it's existence.
phoghed 21 hours ago [-]
We all know all the valuations are fake. Elon taught us this lesson well. Let’s not pretend they don’t have a gorillion users though, including every enterprise on earth.
cyanydeez 20 hours ago [-]
right, but when someone says the "market value" it should be distinguished between the public market that has a bunch of rules, insight, etc vs private equity which is a bunch of people who are directly liable for that value.
One of them has some kind of regulatory control, even if shitty, the other is basically just a bunch of greedy bastards with no oversight.
--
1: https://github.com/openai/codex/blob/0c329ad2c9d5a2592128a1d...
People moan about how badly made Claude code is but Codex seems much worse on every metric
It's like fighting fire by employing an arsonist.
But, I'm just wondering.
So you only snapshot the rest. (And since your source tree is under version control, finding stray garbage there is straightforward unless you checked it in)
Its the only way I trust to install anything from these companies.
Although I just looked into it, and even Ubuntu 26.04 appear to default to tmpfs for /tmp storage, so I guess it's fairly universal now.
Not sure if I agree with that idea in principle... What's wrong with just wiping it intermittently??
It’s simple, deterministic stuff. At least most of the time.
...are you sure it wouldn't consider your personal files garbage?
Anyone experiences with it?
So far I have very positive experiences, the results are generally better (and not just faster) than in other harnesses with the same models. I've done the same (user) prompts for tasks in a range of them, and compare the outcome to eval between them. Simple (harness) beats complicated one many times. I do wish it could use MCP though, but not as often as I think, and CLI tool usage often compensates on this front with little extra effort)
That is now my usuall go-to harness for things usually. Still figuring out the edges though (when we get off the happy path, like models reaching their max context, or such...)
It's very light, recommed you try it out, and see if it fits your toolbox.
And it worked. I'm switching to this unless i find a reason not to.
it's minimal and somewhat opinionated, not integrated well with memory subsystems (although that's kind of the point)
my only real criticism is that last I played with it it deals with slow provider responses very poorly , so lots of locally hosted llms require a lot of tuning to get it to handle streaming/non-streaming/timers well.
way less minimal (but still pretty minimal) I prefer Pi; but different markets I think.
hax fit my openbsd laptop well, espescially since Pi and related derivatives are kind of un-spokenly incompatible with the BSDs.
Also one brief note for Codex users specifically. hax can work out-of-the-box as codex-cli drop-in replacement (since it can use the already authenticated codex token), so it's super easy to try it - basically it will be a zero-config experience.
my main issues with maki are that the line editor (what you use to enter your prompts) is very barebones, copy-paste is broken (no middle-click paste, or copy to PRIMARY; implicit copying to CLIPBOARD that can only be used with a mouse), and scrollback is not native. So it's not a real cli, but some sort of weblike tui.
I tried 3code; it does not mess with selection, but code printed in the conversation has | markers in every line, and is thus uncopyable in practice. Don't know how to use that.
E.g. OpenCode can be used with all three methods. Looks like it needs a plugin for the two Claude methods, though.
https://x.com/FlorianKluge/status/2041855675295318039
(Whether that's against the ToS is not clear.)
You can’t just use your Claude sub on a third party harness without getting blocked even if you use the Agent SDK. Apparently, Anthropic gives “credits” depending on your sub and you can use those _credits_ when you use third party harnesses or Claude -p mode. That usage is not the same Claude sub one would assume it is… It is all quite deceiving. Don’t get a Claude sub unless you’re okay with being locked into Claude Code…
Also, it feels that chatgpt.com and aistudio.google.com are the buggiest websites I have seen for long time. Why it is always something related to LLMs?
But in the same fix, they also removed the reasoning-sample, which kinda sucked.
Chat sessions are not metered by token usage but rather a fairly high limit of messages per day or something.
The longer chats you have, the more context and token input. So I think that discourages longer chats, and encourages users to start new chats.
https://github.com/numtide/llm-agents.nix/
https://github.com/openai/codex/issues/52251#issuecomment-60...
- Goals get blocked due to .git and other stuff becoming randomly read-only when daemon auto-updates in the background (the linked issue in my comment above)
- Goals get blocked due to transient "Selected model is at capacity"
- Goals get blocked due to "Codex ran out of room in the model's context window". Somehow, both auto-compaction and manually running /compact don't work. https://github.com/openai/codex/issues/52567#issuecomment-61...
- Goals get blocked due to bwrap randomly failing with `bwrap: loopback: Failed RTM_NEWADDR: Operation not permitted`. I traced this one to `apparmor` on Ubuntu 24.04 interfering with bwrap and found a local workaround.
If I need special actions, I just make them CLI tools on my Mac and tell the AI model that such a tool exists(via a single line in its system prompt) and that’s it. It’s been very nice.
[1] - https://usehax.dev/
No affiliation whatsoever. Just a happy user.
On top of that, it also feels very much like a normal CLI tool, which is a big plus for me.
It does not support plugins or MCP, which could both be a big negative for many people, but for now I was able to find a good terminal alternative for everything I needed.
Just a huge recommendation from my side as well, not affiliated in any way myself.
I've gotta give hax a go, as a pi user who came on-board due to the no frills direction pi takes. I say takes rather than took, because regardless of the (heated) discussion surrounding mcp and code mode support, I cannot fight the numbers- augmenting the harness provided tools is a notable boost to execution round trip performance.
I mostly stuck around with Hax because of a few design choices. The sessions are merely plain text files indexed by a dirpath slug. This means I can just start a hax session in that directory and expect sessions history to be available in any model.. (I occasionally dabble in GLM for security checks, Gemini for writing stuff, Deepseek for tightly scoped tests etc).
you can also switch the model at any point too after resuming.
But honestly i dont think there is that big a difference between these “minimal harnesses” to begin with so long as they have the basic functionality…
Claude: probably not (officially against the rules after the openclaw fiasco)
As a result, I never knew that Claude Code does a nasty thing like this, either. :-\
Fwiw, packages in that repo get version bumped each day, so if that's a fast enough pace for you it may be a good alternative.
TLDR: They don't care.
I think it's a crazy notion given what I see these things do and generate, and hoped they would too after recent incidents.
Not trying to sing praises; i think opencode is doing a bunch of weird stuff too but it's not codex level.
It doesn't even have a place to check your remaining time or limits except for an in app /quota command that the CLI does not expose.
Maybe this kind of stuff is standard for some people but I absolutely loathe auto updating software for many reasons
I travel for work, and often have a tech demo, solution, tool, or something to show - or am running a workshop.
I usually do NOT update things right before my trip, as I want to minimize the potential for disruption.
Jokes aside I do the same I tend not to upgrade my Mac, my phone unless I really have to.
The “update” button on the top of my brave install has now turned from green to orange and I’ll likely wait until it goes red
What? Pinning a browser?
First thing it'll want to do after starting is update itself to newest version anyway.
Firefox does not only not prompt me for updates (it only updates if I manually run an update), even if I do update with Firefox running, it does not stop me from continuing to use Firefox without restarting it. All I did was install Firefox from flathub (packaged by Mozilla (a default configuration provided by Mozilla)).
https://www.firefox.com/en-US/firefox/141.0/releasenotes/#:~...
Have you ever actually blocked it due to something in the new version, or is it mostly to control the timing of when it updates?
A human seems less likely to embark upon a path that most humans know leads to something a human is unlikely to want
LLMs produce amazing things, but only one token at a time. Anything beyond that is retrospectively emergent at best
There are contexts where "move fast and break things" can make sense, but usually not when it's other people's workflows that you're breaking.
Just how old do you think a category of software product needs to be before it's reasonable for users to start expecting professionalism from the developers?
Auto-update systems have value but are not a suitable default for professional tools.
Like a simple:
- Dropbox:
(also: the example given above is not fair game for nitpicking :D)The thing is: the Internet access part one can control with Lulu or other similar apps. But anything else is a user nightmare - next to impossible even for a cli/tweaking/etc comfortable user. And mac's TCC is a royal mess - even their UI part for this is utterly confusing. I think only the Linux world can achieve something sane+easy like this (if it doesn't have that already; though they have not been known for being "easy").
I have achieved a lot of these for harnesses using sandbox-exec (by basically doing: "deny *; then selectively allow") on mac but it was a pain and I can't even think of trying something like that for "normal" apps.
We considered the daemon thing, but ended up with a hard-update check on every boot. It's thus a user choice: if you don't like the automatic updating, you can just not run the application anymore.
And even with all that, it's hard to stay on the right side of AV algorithms, etc.
We wrote about it here: https://tritium.legal/blog/update
That's not a choice. It s just what most apps do these days, but it's the first time I've seen anyone stating this so directly, or trying to frame it as a choice.
(And behaviorally, this makes your product just a SaaS website with extra steps.)
[Edit: yes, that is kind of the idea. Native version has better features because it can access them.]
I think that's exactly the standpoint of Yann LeCun.
> I’m reminded of
They were drawing a comparison, not saying "x will definitely happen because y"
The breathless astroturf articles, the brain-dead talking head conversations, the endless flood of obvious misinformation about industry uptake, the nauseating LinkedIn puff-pieces about PREPARING YOUR ENTERPRISE OR ELSE... it's a hostile information economy at best, and its being driven by the same people who brought you bullshit like self-driving, crypto and the Metaverse... for the exact same reasons (to get rich).
There's no epistemic credibility left in the tech industry in general. Maybe the Dario's or whoevers of the world believed what they were talking about at one point, but I don't believe a single word I read from Anthropic or OpenAI for the same reason I don't believe anything I read from Russian state media. I know it's being produced for reasons that have nothing to do with the truth.
As someone who interacts with the technology on a daily basis at a non-trivial level... I'm not seeing anything that would cure my skepticism about GenAI. The information economy surrounding it is broken. The fundamentals don't line up. The hype about "nearby capabilities" is identical to the hype of generations after generations of hype in the AI space (dating at least back to Turing). It smells like a bad bet.
I’m in the process of dumping codex for more open harnesses that don’t add problematic new features every week. Because Anthropic insists that I have to use their harness with the subscription, eventually I’m either going to wrap `claude -p` to get away from the crappy TUI or else I’ll just unsubscribe and use other models. Their bad UX has alienated me as a paying customer.
Listed a few in my other comment
Went too far? Oops, it's not accessible from the Claude Code UI anymore! You can get the agent to dredge up the info from the session logs, but you'd better hope they put the pieces together correctly.
Also, you can't rewind to a steering message. Want to make sure you can rewind to it? Wait for the agent to settle, or stop them, enter your message, and hope you don't throw them off their current train of thought.
To see a proper rewind UI, see how Pi does it: Every message is available, forked conversation history is preserved. It has its own issues, but is miles ahead of Claude Code in this regard.
Like of course codex is the most powerful and "just" run `npx blah blah` but no, I'm not running randomly piped code on my personal computer and even if I didn't care about privacy it means I can only run this locally with my laptop on. "leave your laptop on and access from your phone" is not a good idea. It only adds to the surface area of confusion. Ux no good.
And so many more. I tried codex and it was lacking in many UX aspects. Remember we got from a text-paragraph-only terminal to this at breakneck speed.
And they can't seem to go two whole months without scrolling breaking in some new way.
https://x.com/dvsj_in/status/2098620473550700771
Eg - I asked for a recipe and it says leave x for y time - there's actually deeplinks that link to your clock timer with the time preset! And to change quantities, there's input boxes in the response that update all quantities accordingly.
Many such cases of use-case specific UX
This is partly due to the multiplicative scaling factor -- the number of tokens; the number of transformer layers; the number of features; etc.
I suspect that for models to have a step upgrade either/both of 1) a new/improved architecture model; and 2) better training of smaller parameter models are needed. I say the latter as I think that with the larger weights there is a lot of unused/redundant parameters.
https://finance.yahoo.com/sectors/healthcare/articles/godfat...
Progress is spiky and not guaranteed. The capital costs have also ballooned to the point where the rate has to level out
That's a pretty bold use of "has to". None of the financing behind AI seems sane, yet it continues to defy rational expectations. Reminds me of the comment "the market can remain irrational longer than you can stay solvent". It's not the exact meaning but has the same feeling. The circular financing going on does not help anything either
We don't see any indication that the LLM companies are slowing down and go for better code.
Instead it just looks like a cash grab until the AI bubble bursts.
I think newer models are more likely to produce the usuable output, almost always, but its still not great. The frequency in which you get usable results leads one to believe that quality is improving.. No usability is improving but the result is still meh.
Also put an llm in an environment without an lsp or compiler to check against and let me know how capable you think they are on their own. The "loop" is doing a lot of heavy lifting here.
in the last 2 weeks or so: - scrolling stopped working - subagents no longer work (main agent doesn't get notified when they finish) - subagents don't work (a subagent can't make its own agent even if instructed to do so)
what's going on? I thought coding was solved?
That is exactly why it is so bad. Tokens spent generate code. Large amounts of code is a liability.
I use nixos and use a little wrapper that wraps Codex CLI in Docker. Run codex-box /path/to/project and it starts or reuses a container with only that project bind-mounted, keeping login and history in a separate volume. It uses tmux for detaching and reattaching
https://gist.github.com/entropie/8f30b9e33e9412fda3adf2f7507...
But I am not satisfied. I am trying to find a really lightweight container/VM for macOS that eats very less RAM on its own and expands as needed. Because I really don't want to run an agent outside a container on my mac.
Maybe thats just the tool nerd in me that just wants the perfect tool for every job.
The thing that really validated this point was when I recently tried Codex again. Out of the box it was OK, but not great. I swung over the memory apparatus and had it import directives from my existing Claude install and it was almost immediately quite useful in its own right and, especially, as a counterpoint to Claude when dealing with some fairly complex operational issues.
This blogpost describes a similar thing: https://bxt.rs/blog/easy-sandboxing-on-linux-with-bubblewrap...
Maybe it's like online services/tools. Use them until they are VCed and gone pivoting and then move to another tool.
But maybe being able to code up a feature at the drop of a hat has turbo-charged their ability to make bad UX decisions.
The latter requires modifying existing, badly architectured and tightly coupled code that will almost definitely break something somewhere else if touched, so what do you do? Just keep adding new code on top and hope for the best.
This is not a new phenomenon by any means, but AI amplifies it by 100x.
You install an update. Oops, they added a sandbox that doesn't work inside a docker container, it's stopped working.
You install an update. Oops, you can't log in inside a docker container any more, it's stopped working.
You install an update. Oops, they started catching the copy+paste shortcut and replaced it with one that doesn't work.
You revert a broken update. Oops, the update changed the format of your login details, the old version can't read them, gotta log in again.
You install an update. Oops, can't scroll with the mouse wheel any more.
Not showstopping bugs (if you're used to persuading broken oauth logins to work) but it certainly makes updating an "ugh, what have they broken this time?" experience.
One goes to Github and one notices, the dev is done by LLMs, issue triage is done by bots, and the commenting is done by the bots. (btw, the triage mostly means closing them or playing the "this is dupe of that" game while noone gives a flying fuck about any of the issues). Another good one is "not planned". You report a breaking/critical change and "not planned". What demorlised me is attitude of OpenCode devs/maintainers. I think the Open in it more of the "Open" in OpenAI than in FOSS. That's when I started reducing my dependency on this harness whihc was slowly becoming my pretty much full time harness.
Claude Code turned on the remote-control server by default with an update (no explanation or notification). BUT Claude Code didn't break the expectation that the hooks and the commands you run inherit your environment. So it was hard to notice.
Whereas when Codex App server became enabled by default, and as per OpenAI, it was never contractual that hooks run in the client's environment. Sure, let's break long-standing expectations of where code runs when you run a CLI--it was never contractual, you see. Now you have to figure out how to turn this crap off, just to UNBREAK your workflow!
This is why it's apples and oranges here, with Claude Code and Codex.
That's why I didn't mind it when Pi added native MCP capabilities. What dismayed me was they didn't other tools like subagents, and other stuff. I mean why do I have to scour the Interwebs to find something that doesn't possibly screw everything up for me.
I have a feeling harness usage will finally move to lots of harnesses with usage clusters - as in not X million harness for X million devs/users but a lot and people will use what suits them, and if not a more inclined coder will fork/make one just for them. In all their faults LLMs have made this a possibility.
The problem is hardly anyone gives a shit if they can pump out stuff that makes them look like they are good at stuff.
Vanity before engineering.
We were overbuilding trivial parts of the software like the Romans overbuilt small bridges.
Modern engineering is about making the thing that barely works. If you disagree, learn a little bit about bit narrow the tolerances of modern cars and bridges are.
Buildings typically use a safety factor of 1.5 to 2.0, and cars typically use a safety factor of 2.0 to 3.0. These things are being built twice as strong as they need to be. I wouldn't call that narrow.
And with software, we're at a point where it can't even properly do what it needs to. If we were overbuilding it, damn software wouldn't crash so much.
Not even close.
For example, a BMW X5 weighs about 2200 Kgs without passengers, and it's GVWR is 2800 Kgs.
If you have five people and heavy bags, you can comfortably exceed the GVWR. You can absolutely break control arms and suspension components in that scenario if you go over a speed bump or hit a pothole (which are real things that exist in the real world, like people doing something unintended and crashing software).
If this needed a 2x safety factor, there would have been no way to build this car.
Even the toughest cars like land cruisers are limited in this way.
> [...] vs something that's rickety
A 100 percent of modern bridges are rickety. There has been dozens of high profile incidents because of this.
> Just because you built something that might fall apart at any moment doesn't mean it's been engineered with narrow tolerances.
Here the software they are building falls apart in ..say.. two days of use, but can be updated in 5 seconds at close to zero marginal cost. The software is useful enough that people keep updating and using it. This iteration gives them a competitive advantage over the safer choice of rigorously testing every new thing for several weeks.
The same story for Google's products - their products were always buggy even before AI. And so is the case for Microsoft.
I think the reasonable argument is that AI or not, software problems are hard problems to solve. And software is constantly evolving. AI isn't a magic wand. It also depends a lot on the humans using it, including prompting and context supplied to it.
Drawing a massive conclusion from a tiny, unrepresentative sample is just hasty generalization.
Given how rare it was previously, and how little it was cultivated by shortsighted management, I’m not surprised with all the slop we see today.
The truth is there’s nothing stopping someone from using these tools to remove dead code, simplify project structure, remove dependencies, etc. but I don’t see it happening very often. I have done it a bunch myself though so it’s absolutely possible, it’s all a matter of investment and prioritization.
it's embarrassing when every claude code release has obvious bugs in it like telling you to upgrade when you're on the latest version or codex opening visible powershell windows for every command that pop up on your screen
like 5 minutes of human testing would catch this obvious shit but obviously they don't bother because vibes are the future and stuff? idk
Legit bugs still open 6 months on.. It's a bit ironic
The recent terminal UI changes were more fastidious, but even those are easily disabled with CODEX_TUI_DISABLE_KEYBOARD_ENHANCEMENT=1 and the /tui command.
Just a couple of days ago, it was completely broken for about a week before it got somewhat fixed again. When I say fixed, I mean basic functionality is working again but bugs and regressions are still left. There are multiple issues tracking the same problems. The community have even started to patch the extension themselves as a workaround until its adressed.
You know whats even crazier? There are like a new release every other day with lots of commits and changes, yet it still in a bad shape. I have turned off autoupdate for Codex extension for this reason. The only extension I’ve had to.
I mean, it doesn’t take that much work to have some tests in the pipeline to make sure the extension continue to work, it almost feels like they are purposely doing this.
I have now experienced the consequences of ”move fast and break things”. I realize how much software quality means to me. Its not fun being on the receiving end, especially when it affects my daily workflow.
If any of you, see anyone hype up LLMs again, please refer them to the state of Codex vscode extension.
~/.codex/config.toml
hell, I also have this redundancy configured in bash:Have a look at https://chatgpt.com/codex/install.sh. It's managing the installation by itself not using existing package managers. There's a good point to making it easy to install, but it leads to managing an installation when you shouldn't need to. Time would be better spent elsewhere and better solutions exist.
Am I the only one thinking this?
I don't know how and why they ruined this CLI with needless "features". It was perfectly usable two weeks ago.
I remember them from way back on Windows 98 and no one seems to care to hide them (?).
—no-daemon is the way, but still feels like the noose is tightening …
https://pi.dev/
I immediately adopted the new tuiMode "fullscreen" feature the moment it landed, and it's been exceptional. I love the way it now manages the viewport -- this was also right about the time I migrated to Ghostty so that also made a big difference. Showing inline images with the Kitty graphics protocol has been a huge improvement too, I love it when a model calls `read` on an image file and I get to see the results in my terminal. The OSC-8 hyperlink support is super cool too. I love the way it auto hides/shows the scrollbar now too, and recently they've even made it so concurrent tool calls show different "themes" when pending/running vs success vs error.
Except in the past they both would automatically stop and fire an alarm almost. So something clearly changed and I think they just want us not to notice, to avoid damaging the trust in their tools.
I seriously feel at loss with this current AI situation: can't live without them anymore, hate their product still and feel taken advantage of. Local LLM still to expensive and hard to see any light in the tunnel.
https://michael.stapelberg.ch/posts/2026-02-01-coding-agent-...
If you want something not NixOS-related, smolmachines.com (I haven't tried this yet) gets mentioned here a lot.
I remember when opencode was out, I installed it as a separate user and tried it and it was cool.
Then Copy.fail came out, and I worried that a separate user really wasn't enough, but I felt I was overworrying at this point.
Then it worked fine and I felt silly for going through the trouble of a separate user even.
Then the opencode guys started self developing the tool itself and it became slop, this was like 2 months into the fad, and it was clear the quality merited uninstalling the thing, even if it did some cool magic ( https://github.com/anomalyco/opencode/issues?q=is%3Aissue+au... )
Then Codex came out, and uninstalling opencode became the optimal move, if only because it removes a third party. It's one of those things where people will continue to use it out of inertia (either legacy code, or people reading/watching old tutorials), like using Axios in 2026 after Fetch was made standard, or using Mono after Microsoft released multiplatform dotnet.
Everything was good for some time, but then it quickly became obvious that OpenAI wasn't developing this tool, but rather letting it build itself, which results in a lot of slop (High RAM usage, inefficient looping that only works out of sheer GPU power and tight token limits, resulting in Skinner-box based development)
Then I uninstalled Codex, but this Pi thing came out, from the authorship of the dude from Flask, but I'm tapping out here, then Jev came out and I'm just not playing this game, good luck to everyone who is using these things.
It feels like this is an attrition game where we are waiting for the bubble to pop, but the bubble is made very long to tempt people into taking the bait. I'm betting the house on the day of reckoning, and hope that those that didn't take the AI bait, (or took the AI pill and repented) will be one of the few players left with a reputation.
I remember going in interviews and people asking me if I could use these tools, they wanted to see what projects I made, but I wasn't comfortable even showing a Chess Engine thing, like I didn't make this, this is not what I'm going to be delivering professionally, I'd rather be unemployed for a couple of months or years if that's what it takes, I'm not vibecoding for a salary sorry.
One of them has some kind of regulatory control, even if shitty, the other is basically just a bunch of greedy bastards with no oversight.