gcr

18 ms·

361.
by gcr 1y ago
To me, Ring Racers feels like some bizarro version of a "what if we made a kart racer that plays like what normies think a fighting game plays like" future (complimentary, not derogatory). It's fascinating to get a glimpse. I
362.
by gcr 1y ago
Kart Krew accepts patches from outsiders. I've found the community to be quite collaborative. I'm an outsider and I've gotten a few patches accepted to Ring Racers, like this small screenshot bugfix[1], or this more complicat
363.
by gcr 1y ago
I play on dvorak. All keybinds are adjustable in the menu, even in this web version. If you care to compile from source, I have a patch that makes SDL input by scancode which lets you type in dvorak in the chat box etc, lmk if you'd li
364.
by gcr 1y ago
Working on SRB2 mods kickstarted my computer science career about twenty years ago. It's a really good game and, incidentally enough, an amazing learning tool for springboarding level into level design and practical C/C++ knowledg
365.
by gcr 1y ago
Blender doesn't use the nice boundary representations that you might be familiar with from Fusion 360 or Solidworks. Everything in Blender is "bag of triangles/quads," so it's lower level than what you may be used t
366.
by gcr 1y ago
Anyone who likes this might also like Stefan Miller’s paper, “a simple category theoretical understanding of category theory diagrams“, appearing in SIGBOVIK 2014. See https://sigbovik.org/2014/proceedings.pdf (starts
367.
by gcr 1y ago
why did they choose to air this live? For an internal team sure absolutely, but for public-facing work, prerecorded is the way to go
368.
by gcr 1y ago
What's the relationship between the search appliance and Google Desktop? I'm not seeing any connection.
369.
by gcr 1y ago
Which bank was this? Please name them so I can avoid doing business
370.
by gcr 1y ago
If you want "several tables and database-like semantics in one file," then what you want is DuckDB. If you want modern parquet, then you want the Lance format (or LanceDB for DB-like CRUD semantics).
371.
by gcr 1y ago
Does notion have a proper offline sync mode these days? I was under the impression that it doesn’t either
372.
by gcr 1y ago
To my best understanding, my old paper documents were just deleted. I can still log into the old paper.Dropbox.com but see an empty list. I assume this means my account was never migrated.
373.
by gcr 1y ago
In 2020, Dropbox launched “new paper,” which turned the flat list of Paper documents on paper.Dropbox.com into literal files that live in your Dropbox folder. This was a staged rollout. Mine wasn’t migrated until five years later. For a whi
374.
by gcr 1y ago
Giving an author 90 minutes of lead time before public negative press doesn’t count as responsible disclosure. Especially when that press doesn’t mention the specific security vulnerabilities you’re reporting to them. Here is a link to the
375.
by gcr 1y ago
Your changes are almost certainly in the op log, even as just bare snapshot operations. Have you been able to search that and recover them? (i've found jj undo quite robust, i'd be surprised if you ever actually lost work tbh)
376.
by gcr 1y ago
My gosh you're right! The paper in question is https://arxiv.org/pdf/2204.05149 , "The Carbon Footprint of Machine Learning Training Will Plateau, Then Shrink"
377.
by gcr 1y ago
Google was not using deep learning for search in 2011. Deep learning as a concept didn't really take off until AlexNet in 2012 anyway. Various ML "learn-to-rank" tooling was in use at Google for a while, but incorporating doc
378.
by gcr 1y ago
That's just a press release. Here's the actual PDF of the actual tech report: https://services.google.com/fh/files/misc/measuring_the_envi...
379.
by gcr 1y ago
I'd still love to see a report that accurately captures training cost. Today's report[1] notably excludes training cost. Not just "one training run," but the cost of a thousand AI engineers starting failing runs to get t
380.
by gcr 1y ago
In 2011, Google claimed that each search query takes about 0.3Wh [1]. Earlier this year, Sam Altman also claimed about 0.3Wh avg use per query for OpenAI. I'm honestly surprised that they're so similar. I've thought of LLM qu
381.
by gcr 1y ago
characterizing netflix as a "read-only" website is incredibly shortsighted. you have: - a constantly changing library across constantly changing licensing regions available in constantly changing languages - collaborative filterin
382.
by gcr 1y ago
The whole point of GBNF is to serve as part of the API that lets downstream applications control token sampling in a high-level way without having to drop to raw logit distributions or pull model-specific tricks. Ollama has a hardcoded GBNF
383.
by gcr 1y ago
The entire HP series is about one million words.
384.
by gcr 1y ago
Fastest I can manage is 0.5s, can anyone die quicker?
385.
by gcr 1y ago
Nope! “Top K” means you’re on the high score list of size K. Only the best players are in the “top 1%” for example. Being in the top 50 is more prestigious than the top 300,000.
386.
by gcr 1y ago
bad news — top 80% means most players survived longer than you, sorry :-) perhaps you’re thinking of top 19%? but you’re right, lots of bullet hells get easier if you hold still and only move when required.
387.
by gcr 1y ago
ah, perfectly captures the meditative feeling of touhou, just needs the music :-) games like this have a lot to teach you about patience and reflexes
388.
by gcr 1y ago
The PR you linked to says “thanks to the amazing work done by ggml-org” and doesn’t remove GGML code, it instead updates the vendored version and seems to throw away ollama’s custom changes. That’s the opposite of disentangling. Here’s the
389.
by gcr 1y ago
This is misinformation. Ollama’s supported structured outputs that conform to a given JSON-schema for months. Here’s a post about this from last year: https://ollama.com/blog/structured-outputs This is absolutely possi
390.
by gcr 1y ago
Goodbye MPEG group, and to be frank, good riddance I think. I'm glad that open codecs are now taking over on the frontier of SOTA encoding. Maybe these sorts of handshake agreements and industry collaboration were necessary to get thin

More ›