285 karma · joined June 4, 2023
Claude Code is actually far from the best harness for Claude, ironically...
JetBrains' AI Assistant with Claude Agent is a much better harness for Claude.
they tried to do something with remembering "how you left things" between sessions, and even when disabled things are still weird...
Also some power management related hooks are not working as well as before. Like if you put the computer to sleep at night, and wake it up in the morning, the automatic dark-to-light theme switch doesn't trigger. at least not always.
Still the best system to work with though!
the original function was full of mutable state (not required), full of special cases (not required), full of extra return statements (not required). Also had some private helper methods that were mocked in the tests (!!!).
All of this just for a "pure" function. Just immutable object in - immutable object out.
and yes, he was a junior.
Extremely impressive feat nonetheless!
I regularly review code that is way more complicated that it should.
The last few days I was going back and forth on reviews on a function that had originally cyclomatic complexity of 23. Eventually I got it down to 8, but I had to call him into a pair programming session and show him how the complexity could be reduced.
This sounds more serious than it has gotten attention for :(
People generally don't do firmware updates on their headsets, unless I live in a bubble...
Considering that the brain has hard-wired neurons (since birth) to automatically count/recognize up to 3 objects, we need some icons as anchor points. With emphasis on some.
You can have max of 3 icons in a row, and a max distance of 6 items without icons.
Then you can VERY quickly find any menu item, by doing "+/-3 from X anchor" math, which your brain does extremely efficiently.
Further, this means that an action is represented by a relative path (e. g. save as = 1 down from the save icon). This further helps the brain to store these associations, because it's also extremely efficient at storing relative paths. So, long time users automatically become more and more efficient at using your software.
The pay-per-crawl thing, is about them thinking ahead about post-AI business/revenue models.
The way AI happened, it removed a big chunk of revenue from news companies, blogs, etc. Because lots of people go to AI instead of reaching the actual 3rd party website.
AI currently gets the content for free from the 3rd party websites, but they have revenue from their users.
So Cloudflare is proposing that AI companies should be paying for their crawling. Cloudflare's solution would give the lost revenue back where it belongs, just through a different mechanism.
The ugly side of the story is that this was already an existing solution, and open source, called L402.org.
Cloudflare wants to be the first to take a piece of the pie, but also instead of using the open source version, they forked it internally and published it as their own service, which is cloudflare specific.
To be completely fair, the l402 requires you to solve the payment mechanism itself, which for Cloudflare is easy because they already deal with payments.
Anyway, for anyone actually looking for good cloud drive hosting, without any BS: rsync.net (you encrypt on your side before sending anything. I use Vorta with them).
Also the same server can be used by multiple (trursted) users, like family members etc.
I saw another comment calling it "webview app", which is also valid, but we call it "hybrid app".
Good luck to anyone wanting only ideologically compatible software. They'll end up with pretty much nothing left to use.
Same applies to companies that produce goods. It's a never ending hole.
It's even sadder to me that the author says this is not GPT. I believe them. Which means we have reached a point where the style of how ChatGPT writes has made its way into our sub-conscious...
And the answer to "why now" is quite simple - Because of the whole Rust in kernel debate, people started scrutinizing the situation.
If we jump down to the bolts and nuts, let's say on a json API, it's about including extra attributes/fields in your json response that contain links and information of how to continue. These attributes have to be blended with your other real attributes.
For example if you just created a resource with a POST endpoint, you can include a link to GET the freshly created resource ("_fetch"), a link to delete it ("_delete"), a link to list all resources of the same collection ("_list"), etc...
Then the client application is supposed to automatically discover the API's functionality. In case of a UI, it's supposed to automatically discover the API's functionality and build a presentation layer on the fly, which the user can see and use. From our example above, the UI codebase would never have a "delete" resource button, it would have a generic button which would be created and placed on the UI based on the _delete field coming back from the API
I had to spend 8 hours in Stansted airport, and I managed to setup the tunnel while in the time limit of the free WiFi (I think it was 30'). It felt good, haha.
General population doesn't understand that fixed pricing includes an extra cost which is the risk that the electricity provider has to account for. That risk has a calculable price, which is passed down to the consumers. But because it's baked in the flat rate, nobody complains.
Smart/dynamic pricing actually benefits the consumer.