qudat

8 ms·

121.
by qudat 8mo ago
They still need someone to review and hopefully QA every PR. I doubt it’s saving much time except maybe the initial debug pass of building human context of the problem. The real benefit here is the ability for the human swe to quickly conte
122.
by qudat 8mo ago
The trajectory hasn’t changed: they scaled generating code, a great feat, but someone has to apply higher level abstract thinking to make the tool useful. Running agents in a cron or having non SWEs use it will not last longer than a protot
123.
by qudat 8mo ago
It’s a tool like a linter. It’s a fancy tool, but calling it anything more than a tool is hype
124.
by qudat 8mo ago
> It can be a co-worker and it can also replace you and your whole team. You drank the koolaide m8. It fundamentally cannot replace a single SWE and never will without fundamental changes to the model construction. If there is displaceme
125.
by qudat 8mo ago
Agreed. The bottleneck is QA/Code review and that is never going away from most corps. I've never worked at a job in tech that didn't require code review and no, asking a code agent to review a PR is never going to be "
126.
by qudat 8mo ago
Creative destruction is a fundamental component of economic growth and has been happening since economies were part of humanity. You are thinking too linearly. When the price of goods and services go down because the cost to produce those g
127.
by qudat 8mo ago
At this point I don’t trust any company that offers a core free tool with an upsell. Trials or limited access is one thing, but a free forever product that needs active maintaining, be skeptical. It’s been tough for us at https:/&#x2F
128.
by qudat 8mo ago
Have you tried qwen3-coder-next? Model moats are going bye bye
129.
by qudat 8mo ago
> In 2025, 1.1 million layoffs were announced. Only the sixth time that threshold has been breached since 1993. Over 55,000 explicitly cited AI. Believing what companies say is the reason for a layoff instead of figuring out the actual r
130.
by qudat 8mo ago
Agreed. Companies don’t have the capacity to be moral entities. They are driven purely based on monetary incentives. They are mechanical machinery. People are anthropomorphizing values onto companies or being duped by marketing speak.
131.
by qudat 8mo ago
I think you would be very surprised at a) how useful it would be and b) how lax prod can be depending on the company culture and stakes.
132.
by qudat 8mo ago
If the terminal is in alt-screen the history will only print the current screen. Happy to brainstorm on the zmx repo if you are interested
133.
by qudat 8mo ago
I have thought about writing a separate tool that resembles dvtm but I’m not exactly sure how I would build it. I don’t want to maintain a monster project like terminal multiplexing. Zmx is basically a single file with 1500 LoC and is “prod
134.
by qudat 8mo ago
What do you mean by snapshots? There’s a “zmx history” command which will print whatever is stored in libghostty as plain text, or with ansi escape codes, or even html
135.
by qudat 8mo ago
Nice! Big fan of abduco. I wrote a similar tool but use libghostty for rehydrating the terminal session: https://zmx.sh Works pretty well if you don’t need a window manager in your terminal
136.
by qudat 9mo ago
Nice job! I need to learn how to use tcpdump apparently
137.
by qudat 9mo ago
SSH suffers from tcp-in-tcp issues which means it’ll always take a performance hit over other protocols
138.
by qudat 9mo ago
There are other tunneling solutions that support both and https, websockets using ssh tunnels for the communication. For example I use https://tuns.sh which is a managed sish instance
139.
by qudat 9mo ago
> I don't think it's accurate to describe LLMs as pattern matching I’m talking about the inference step, which uses tensor geometry arithmetic to find patterns in text. We don’t understand what those patterns are but it’s clear
140.
by qudat 9mo ago
Anatomically good candidate, the thalamal-cortical loop: https://en.wikipedia.org/wiki/Cortico-basal_ganglia-thalamo-...
141.
by qudat 9mo ago
The only time I preferred light mode was as when I was on vacation in Florida and worked out of a sun room. I find myself normally working in low light environments which makes dark mode better overall. But it doesn’t have to be black backg
142.
by qudat 9mo ago
My take is that a huge part of human intelligence is pattern matching. We just didn’t understand how much multidimensional geometry influenced our matches
143.
by qudat 9mo ago
> Then you look around and see "startup X gets to $1M ARR a month after launch" and shit like that and I'm feeling terrible about how we're barely growing. Comparison is the thief of joy. I fall into this trap almost
144.
by qudat 9mo ago
This is a very cool idea and I like the simplicity of the business model! SSH has a ton of great features and its ergonomics are excellent for terminal enthusiasts. Most of us want to ssh into our cloud compute anyway. As a founder of an s
145.
by qudat 9mo ago
just install fish then
146.
by qudat 9mo ago
Betteridge's law of headlines is an adage that states: "Any headline that ends in a question mark can be answered by the word no."
147.
by qudat 9mo ago
That's true for internet searching. How many times have you gone to SO, seen a confident answer, tried it, and it failed to do what you needed?
148.
by qudat 9mo ago
This is the part that I like the most, which is why I created https://pico.sh Further, when building ssh "apps" you can build out tooling for client clis that already exist (e.g. rsync, sftp, scp, sshfs). This provide
149.
by qudat 9mo ago
Wrong. I will spend 30 minutes having the LLM explain every line of code and why it's important, with context-specific follow-up questions. An LLM is one of the best ways to learn ...
150.
by qudat 9mo ago
I know this isn’t quite the spirit of self hosting, but for people that aren’t ready to self they host can use https://prose.sh which interops with Hugo. It’s a safe stepping stone into self hosting a blog for anyone who wants t

More ›