For me even just the config format is such an upgrade that it was worth switching. Just a simple text file I can backup anywhere and actually read without clicking in a tabbed GUI where I never seem to remember what goes where. iTerm2 still has that weird thing you have to export right?
It’s subjective. The config format is a downgrade for me. The only settings I change is the color theme anyways. The terminal default should work well. It takes four seconds to use the mouse to open the theme config and pick a theme. It takes at least five minute to find the config file location, the syntax for the config file, understand whether it uses fs events to auto reload, and then decide to include or exclude it in your backups. I don’t intensely customize my terminal emulator so it’s not worthwhile. On the other hand my Emacs config is more than 500 lines long and that’s worthwhile.
I think you hit upon what bugs me about file config. It's not the file, it is having to find it, find the docs, find the syntax, find the keys, find the values, etc.
If it was always in a standard format, it wouldn't be a problem. Something like dconf, without the poor design.
OK, that may be true (and yes, iTerm still uses weird plist xD) but... On the day to day basis I don't care about config file. I mostly use one OS, I don't do reinstalls (had MBP from 2013, then after switching to MBP M1 moved the files and configs _once_) so having nice and readable config doesn't matter at all.
Having to search for some obscure configuration names and options is simply getting in the way.
For that reason I stayed away from Zed for ages - it is a fine editor but tweaking it options without even rudementary settings GUI was just super-meh.
I do hope that Ghostty will go the same route and they will add something that would help manage the simple config options (backed by the config file) without having to deal with purely-textual form)
This... It's weird how companies and people think adding something is the same as adding value.
Feature creep was an issue even before AI, now we see one worse feature being added after the next (does anyone even know all the touch controls of iOS or even macOS, as an example), with some expectation of that there will be money coming in because of it.
I benefit greatly from AI myself, but I use it to augment my capabilities, not as a slot machine for building things nobody will ever use, which seems to be the norm based on the products I use.
I much preferred the fake photos—as in not photos of food served at that restaurant—than the AI generated images of food that doesn't exist at all on menus, as an example.
I also hate that Etsy is now mostly an AI-image gallery.
A one-shotted POC can also put you straight into the very wrong direction, with a very high degree of certainty.
I also prefer doing a POC, but it's not without risk, especially if your stakeholders are just a bit too eager. Another problem I've seen is that a POC can hide complexity, so a lot of weight is on you as the author to do the due diligence before and bring that along.
This is an often underrated problem. I used to work at a company where we would hold demos every other friday, with drinks and nibbles. We would put a lot of thoughts into what we were going to present, and were really careful about the messaging, because execs and sales were watching.
The last thing we wanted was for an "Isn't it neat?!" hacked-together demo to be mistaken for something shippable, if workable at all.
Apple has a lot of actual money, Anthropic does not. Apple sells a lot of products, Anthropic doesn't. Apple have a platform multiple players want a piece of, Anthropic fight to stand alone.
There's very little comparison to be had here. Also, Anthropic likely have investors with way too much leverage for them to stop.
...exactly this were my thoughts when reading the response:
Maybe Apple is just waiting on the line until someone can be sucked up at a penny if the house is on fire - and pretty sure, this will happen. Not in 2027, but I expect MA activites & consoliation before 2030
> A truly skilled developer who makes a genuine attempt should be expected to solve 99% of their tasks within the same business day as of October 2026.
What? You know there's only 24h in a day right? I work 8 of those and most tasks takes my agent far longer. And thats not including me correcting it, adjusting the spec so it does the right thing, etc etc. And I don't have anyone doing a PR waiting, I go straight into production.
It sounds like either your “tasks” are too large or you have some drag somewhere like test cases being too slow to execute.
I’d look into where the model is spending its time and see if you can save some time. Eight hours is more then enough time to have an entire new microservice up and running and to adjust existing microservices to utilize it.
I hope he validates the output, the number of times colleagues have done exactly that and gotten the wrong output is.. well, not an insignificant amount.
Not saying it can't be done, but my recommendation is still "ask your LLM to make a tool that process the data, do not process it via the LLM itself".
Weird how many people on HN think this is about the count. Also it caused animal displacement from what I could gather, Google of course did this willingly and knowingly, because they know nothing will happen.
reply