this is useful for code that renders images (e.g. data-visualization tools). the image is the primary artifact of interest, but maybe it was generated from data represented in JSON format. by embedding the source data (invisibly) in the image, you can extract it later to modify and re-generate.
it's magic to see the repair process, but not magic to see how the mechanisms work? i'm unsure how you could draw enjoyment from the former without also appreciating the latter.
Kinda hoped that's what I wrote, but clearly not. For what it's worth, I'm already working out how I can give the same treatment to my Atari Lynx when it conks out.
ah, i believe i misunderstood. on re-reading, i see your statement about sacrilege was more about removing a functional watch from circulation before its time.
but did they choose the correct path for their org to get there?
> [Mozilla] hired someone from the mobile industry to run the company, this led to some culture changes : no more a flat org, but a pyramidal one with middle managers. Culture became way less engineering centric, and started being a bit more top -> down. Focus was now solely on B2G.
when i think about the occasional anti-user decisions Mozilla has been making in the past several years (most recently, AI training policies) i wonder now if it's all residual from this structural shift.
> when i think about the occasional anti-user decisions Mozilla has been making in the past several years (most recently, AI training policies) i wonder now if it's all residual from this structural shift.
I don't know enough about Mozilla to say, but the OP certainly makes it sound that way.
Eh, as someone who was there at the time I would say that these people largely left when B2G was shut down. I don't think there's much of a connection to whatever gripes you have about Mozilla today.
summary - Perifractic from RetroRecipes intends to purchase all 47 commodore trademarks to enable the community to easily and affordably license the brand name for new products (like the mega64, ultimate64, c64os, etc).
> NONE of them have even mentioned the privacy aspect
because the privacy aspect has nothing to do with LLMs and everything to do with relying on cloud providers. HN users have been vocal about that since long before LLMs existed.
LLMs at their core do produce reproducible results with a given seed. it's all the workflow stuff people do on top that tends to break reproducibility.
This is not the case for LLMs running on GPUs (which is most of them); GPUs are non-deterministic for this use-case due to the floating point math involved. there is no way to get perfectly deterministic output from OpenAI despite the presence of seed and temperature parameters.
i'm not sure i agree. maybe if you're "vibe-coding", but not if you're using AI as an assistant. a good abstraction makes it hard to write bugs, so telling AI to use a certain library (which i know to be high quality) is a good way to constrain the types of bugs i have to look for when reviewing the code.
reply