- AerolfosWell, it would be nice to see examples (or weights to be completely open) for the baseline model, without any GPT-5 influence whatsoever. Basically let people see what the "raw" output from historical texts is like, and for that matter actively demonstrating why the extra tweaks and layers are needed to make a useful model. Show, don't tell, really.
- > but surely consumer SSDs wouldn't just completely ignore fsync and blatantly lie that the data has been persisted?
That doesn't even help if fsync() doesn't do what developers expect: https://danluu.com/fsyncgate/
I think this was the blog post that had a bunch more stuff that can go wrong too: https://danluu.com/deconstruct-files/
But basically fsync itself (sometimes) has dubious behaviour, then OS on top of kernel handles it dubiously, and then even on top of that most databases can ignore fsync erroring (and lie that the data was written properly)
So... yes.
- > https://github.com/DGoettlich/history-llms/blob/main/ranke-4...
Given the training notes, it seems like you can't get the performance they give examples of?
I'm not sure about the exact details but there is some kind of targetted distillation of GPT-5 involved to try and get more conversational text and better performance. Which seems a bit iffy to me.
- Ok so it was that. The responses given did sound off, while it has some period-appropriate mannerisms, and has entire sections basically rephrased from some popular historical texts, it seems off compared to reading an actual 1900s text. The overall vibe just isn't right, it seems too modern, somehow.
I also wonder that you'd get this kind of performance with actual, just pre-1900s text. LLMs work because they're fed terabytes of text, if you just give it gigabytes you get a 2019 word model. The fundamental technology is mostly the same, after all.
- > To make `Z` a column vector, we would need something like `Z = (Y @ X)[:,np.newaxis]`.
Doesn't just (Y @ X)[None] work? None adding an extra dimension works in practice but I don't know if you're "supposed" to do that
- > Why not use X.transpose()?
Or just X.T, the shorthand alias for that
- The article is AI-written. Obviously can't be relied on to accurately convey the original information
"But Arduino isn’t SaaS. It’s the foundation of the maker ecosystem." is a giveaway, but the whole set of paragraphs of setup before that is chatgpt's style
- Video editing? Adobe has set themselves up for failure there, everyone wants an alternative
Davinci Resolve is probably competitive with Premiere, but while free it's not actually open source. But either a viable competitor catching up or Davinci publishing the code could change that fast
- I'm sure people will come out of the woodwork to tell me just how wrong this is, but say various Linux projects or the kernel seem to have better (better, not perfect) governance structures.
Not being obsessed with rapacious growth, not chasing trends and features that look good on delivery metrics but instead building a stable product would go a long way.
And frankly, for descriptions on what the product should be, standards to implement (or not), and overall strategy for a project that tries to do its best for tech as a whole - Mozillas own writeups are spot-on! They just don't seem to act in accordance with the "vibe" and ideals that blog posts etc. talk about.
- >The Mozilla leadership seems to have a unfortunate tendency to emulate the behaviors of the tech companies that their core Firefox project is often seen as an alternative too.
But of course, they need "competitive" salaries so they can hire "great talent" from the tech sector so that the company doesn't fall behind or something.
- Actually more common than you might think.
Bethesda games have the same ecosystem - they do provide an official plugin system, but since modders aren't content with the restrictions of that system, they reverse engineered the game(s, this has been going on since Oblivion) and made a script extender that hacks the game in-memory to inject scripts (hence the name).
- Standard policy is I think mostly the same, but in Europe there's been arguments that those policies don't follow the actual consumer protection laws, which is a whole thing that I don't think really resolved one way or another.
It varies with country but I believe a number of protection laws specify normal use/testing a product is allowed, so you can open boxes and test functionality (norwegian law does this for sure). Excepting videogames from this is arbitrary, the argument from consumer protection agencies goes.
I believe in practice a number of games did get refunded when threatened with formal complaints along these lines, but that's far from a guaranteed thing.
Anyway, GOG decided to go with the generous interpretation (and the one all kinds of electronic goods except games and CDs/DVDs have), which is nicer for everyone, really
- Key activation is actually a good point.
I believe there were some pushes to get rid of opened box = no refund policies as being against standard 14-day returns in Norway, because the law explicitly says the consumer may (paraphrase) "reasonably test the use of a product" which allows you to open the box on other goods. But keys being consumable puts them in another category of goods (like food, which obviously can't be returned after "use"), so that doesn't apply.
- In some sense, GOGs entire existence is testing the hypothesis that it's impossible to run a consumer friendly digital store due to abuse.
If you took the common sense publisher view then no DRM = everything you make is instantly pirated and the whole store fails instantly. But GOG is a viable storefront, so that's demonstrably wrong.
The evidence is no better for Steam's refund policy than it was for DRM being necessary.
- > Refunds with no questions asked if you've played less than 2 hours of the game,
Weaker than standard physical store consumer protections (no playtime restriction on returns, obviously), and (much) weaker than GOGs refund: 1 month after purchase, no playtime restriction.
I believe they explicitly called out the equivalent for physical stores and european consumer protection in general when they announced the policy and lack of restrictions. Which is an indirect call out at Steam, which hasn't cared in the slightest and continues to have a worse policy.