out_of_protocol

Depends on the game obviously. e.g. Assassin's Creed N+1 - you've seen the formula like 20+ times already, all the previous games + all of Far Cry from 3 upwards. There's nothing new. And there's Cyberpunk2077 and BG3 which are still huge but unique and carefully designed

KiB means 1,024 bytes, KB is still 1000, if you want to be pedantic.

On other note, who the hell decided kibi, mebi, gibi are good prefix names? that's the reason no one uses them (me too)

Too late, people already talking like this too

Not even close to being as popular as sports or a gym stuff. There's basically common knowledge - to be a good <think profession> you need to hit a gym regularly. And it's going to be gym+something for a brain soon. At least i hope so

I really hope there will be social change - flexing mind - like flexing body in a gym/sport - becomes a thing.

20x was never about total usage, only 5h one. For 2x price you get 2x usage

count .png-s in the web. There's a lot. And .jxl's lossless is much better than png's. Basically any image with sharp lines (screenshots, image of a button etc) or gradients should be lossless.

also, don't forget - actual video format is better for animated stuff. Current default approach is .webm which is standardized .mkv container + vp9 (or av1) video + opus audio (usually no audio to as a feature parity to gif). IMO animated picture formats like .avifs are very backwards and don't add much

zstd with better compression level would be nice - these numbers are not really comparable since both time and compression level are too different

Apple dropped 32 bit support in specific OS version with no way to run it anymore sans downgrading the OS - so it's not hardware restriction

Tell that Apple and Android. I have 32 bit native software which can't run on modern devices at all.

> early hours of the morning

which TZ?

> There are no off-peak hours.

Probably US west coast hours, but there's no public stats

Dunno. Crease almost invisible when screen is on and not pitch black ( using Honor V3 while actual generation is V6).

Does that also mean owner must re-run very model ever released by them to detect anything?

1) That's a lot more entropy to gather, means sample size needs to be huge.

2) this also means check must be run against every user's ID in their database, 10^9 more expensive

Do i understand correctly - to check watermark you need full model weights, of all org models. So running the check is basically the same as running every model once? That's really expensive

Well yes, SV is abbreviation for sievert which measures health risk of radiation

*for a brief moment it actually runs.

from the point of the dev - work done on the port will rot very soon, return of investment is really low. Also, the studio may not exist at all by time changes will be necessary - as fast as few days after public release nowadays :(

Why bother with macos version at all then? Soon current version becomes unusable as well, and needs windows fallback to run since macos version is useless

> It trades something fungible and incredibly cheap (CPU)

it could be RAM-bound, which is very much NOT cheap nowadays :)

There's a lot of places without fiber, e.g. all the ships/jets etc. there's a lot of low-density areas, there's islands with no internet or VERY expensive internet

> 6. You can make money without doing evil.

You can but well, it's more profitable the other way around....

On Android, opera browser has text-wrap which fits text into screen on any zooom level. Dunno how people live without this feature

I blessed to live in a country that uses kilowatts. That works wonderfully (and advertised as such) for AC units

Did anybody compared these directly using exactly same prompts and harness? I assume V4 Pro could be real frontier model, and if it's true, it'd be better to use it in automation or routine steps instead of simple models (e.g. haiku or even sonnet if V4pro is better)

1) that's not an ORM

2) after ~8 years of using it, i find it ergonomic, light on congitive load and good for long term support

Elixir is very ergonomic when it comes to concurrency, e.g. parallel-map example:

    1..10
    |> Task.async_stream(fn x -> x*2 end, max_concurrency: 2, timeout: 7000, on_timeout: :kill_task)
    |> Enum.to_list()

Equivalent Go code would be very long and very ugly.

Golang has its positives but you also lose a lot. Whole ecosystem is not comparable, like "debug live production cluster" is one-command away for Elixir vs "fuck you" for Golang

Django is not very different than Rails and imo much less ergonomic. Many issues are still there like hidden state/setup needed to call something

> ActiveRecord is more pleasant to work with than the ORM of Phoenix IMHO, but not everyone shares the same feeling.

Well, depends on what you do. Ecto is closely follows SQL logic and allows to translate weird sql queries into code directly. All queries are explicit, e.g. you either do preload(...) or can't access nested records at all, no chance of N+1 by design.

Changesets are also different and are just functions you can define as needed.

    defmodule MyApp.User do
    ...
    def changeset(user, attrs) do
        user
        |> cast(attrs, [:email])
        |> validate_required([:email])
        # This matches the error from the DB uniq index to the :email field
        |> unique_constraint(:email)
    end

I'd say Rails is faster than Phoenix (as in development speed) only for the first day or so. After that you'll stumble upon impicit logic, method-missing and this kind of stuff, which will require more time to figure out how it works. Elixi/Phoenix is more exlicit in that regard, making long-term support (as in anything past first week) a breeze. No hidden state, no figuring out where ModuleName.method(params) is coming from, no need to setup stuff to launch said method (just pass right arguments). The only downside i see is smaller library of ready to use packages