andyg_blog
Joined 7 years ago
Karma
317
Submissions
64
Member since
May 10
Recent submissions
Comment · 25 days ago
10x really isn't hyperbole at this point
Comment · 1 month ago
I don't think anyone would mind a society without work so much if there wasn't the threat of starvation and destitution there currently is.
Comment · 2 months ago
It doesn't, and FWIW the article says that. A meta note: ABI stability is a holy war inside the C++ community. Those against it argue that it's holding back real language evolution. Herb Sutter is even working on his own C++ offshoot that shows what could be.
Comment · 2 months ago
I think it's a pretty good direction to go down. I wonder if we can add a little more accounting to it. Cheap shoes solve an immediate problem but you pay more long term versus expensive well built ones. How much $ (in dev time + tokens) was spent in incidents and/or fixing code later. Credits and debits. Weigh money earned (somehow) against money spent.
Comment · 5 months ago
This reminds me a bit of "Command-line Tools can be 235x Faster than your Hadoop Cluster" from 2014 https://adamdrake.com/command-line-tools-can-be-235x-faster-...
The constraint I, and I bet many here, have is just how much data there is. 3GB like in the 2014 article is one .pdf
Enterprise level data store is measured in hundreds of GB for a single customer, and you'll get murdered on data egress costs if you try to search an entire corpus, if you can even get through it all before the request times out or the customer decides after 5 minutes that enough is enough.
You'd need a true distributed filesystem to even start attempting what the authors suggest at any scale outside of your local machine.
Comment · 7 months ago
Completely agree on alignment. Without it, it's a shortcut to rejection. I actually wrote a lot about this in a blog post I called "Minimum Reviewable Unit" https://gieseanw.wordpress.com/2025/03/21/minimum-reviewable...
Comment · 7 months ago
I don't think it contradicts the OP. OP says the system is unreliable. Memory leaks that lead to out of memory failures for example. Smart pointers would stabilize things. (Also note that OP says their smart pointers PR was rejected).
Comment · 7 months ago
This is a small April Fools' project I created while exploring git commit hooks. I wondered if you could fail a commit based on the commit hash itself. Turns out you can't; you have to wait until the post-commit hook fires, and then do a reset on failure. Eventually I decided to see just how far I could take it, and well, now there's an official website and python sdk distributed through pip.
Comment · 7 months ago
>the whole article assumes the only language in the world is Python.
This was my take as well.
My company recently started using Dspy, but you know what? We had to stand up an entire new repo in Python for it, because the vast majority of our code is not Python.