Years ago, I played with AWS then contacted their support to make sure any AWS billing to my account was disabled.
I thought I’d try it again recently, and couldn’t log in.
I still don’t think I’m missing anything.

I’d rather have VPS or server providers where I know exactly what I’m getting per month no matter what, tho I’ve ran near data transfer surcharges.

Oh, it’s expected costs.
Like, figure out the compute requirements of your code, multiply by the cost per compute unit (or whatever): boom, your cost.
Totally predictable.
Compared to suddenly having to replace a $20k server that dies in your data center.
So much easier.

Except when your code (let’s be honest, the most likely thing to have an error in it… At least compared to some 4+ year old production hardware that everyone runs) has a bug in it that requires 20x compute.
But maybe that is a popularity spike (the hug-of-death)! That’s why you migrated to the #cloud anyway, right? To handle these spikes! And you’ve always paid your bills so… Yeh, here’s a 20x bill.

The amount of software that is limited free self-hosted but the next tier of “self hosted” is enterprise and thousands per year is ridiculous.
Absolutely ridiculous.

Like, you have self hosted. I like your software, I use it personally and that’s why I’m using it for (and recommending it to) small businesses. They could afford your 10-100 per month for whatever extra features, but they don’t want to rely on 3rd party hosting. They want to host it themselves.
But the only way to get those features is to go for some “cloud” bullshit they don’t control, or to pay “enterprise” prices.

It’s why I make part of what I make/charge a contribution to the products and projects I use and recommend.
I’ll set all that up and tailor it to your company, but anything and everything I recommend/implement is standing on the shoulders of giants. So pay those giants.
Although I think I’m lucky with the people I work for, in that that are interested in the tech, but not the detail.

Ah, lol.
Is that the web interface? Or what app is that?

https://www.bbc.co.uk/news/videos/cewd82p09l0o

I think that’s the link to the video?
Seems like it’s part of a longer video…

Edit:
Hhmmm here is a slightly longer video that doesn’t really add anything

Actual edit:
I genuinely couldn’t find a better source video

“The United States has unilaterally and repeatedly provoked new economic and trade frictions, exacerbating uncertainty and instability in bilateral economic and trade relations,” the statement said. “Instead of reflecting on its own actions, the United States has groundlessly accused China of violating the consensus, a claim that grossly distorts the facts.”

That is such a wonderfully diplomatic way of saying “stop being a fucking idiot, your words have meaning and these are the consequences. Grow up”.
Even just “grow up”, tbh.

As much as I dislike the amount of reliance the world has on China (for the labour conditions there, the nature of their government to impose dodgy practices, generally speaking not being a “good egg”), China seems like the only trading bloc (although not a bloc, I guess… Maybe “trading entity”) that can unilaterally stand toe-to-toe with TACO and win. So, good on china.

let data = null
do {
    const response = await openai.prompt(prompt)
    if (response.error !== null) continue;
    try {
        data = JSON.parse(response.text)
    } catch {
        data = null // just in case
    }
} while (data === null)
return data

Meh, not my money

Edit: wrong comment

Yup.
It’s a traumatic job/task that gets farmed to the cheapest supplier which is extremely unlikely to have suitable safe guards and care for their employees.

If I were implementing this, I would use a safer/stricter model with a human backed appeal system.
I would then use some metrics to generate an account reputation (verified ID, interaction with friends network, previous posts/moderation/appeals), and use that to either: auto-approve AI actions with no appeals (low rep); auto-approve AI actions with human appeal (moderate rep); AI actions must be approved by humans (high rep).

This way, high reputation accounts can still discuss & raise awareness of potentially moderatable topics as quickly as they happen (think breaking news kinda thing). Moderate reputation accounts can argue their case (in case of false positives). Low reputation accounts don’t traumatize the moderators.

That’s the fun part about windows: who the fuck knows?
Can’t look at the source, can’t confirm if it’s bad API implementation or bad documentation.

Trying to disable the windows key hotkey that opens the start menu, so the game The Witness can pause stuff, minimize, open the start menu and release the block on the windows key (IE do a more controlled start menu hotkey, instead of having windows rudely interrupt everything and break the game).

Started with a 5 second hang whenever a debug breakpoint was reached. The dev started digging into the issue.

Games use RawInput to get better mouse interactions, but that breaks the Microsoft recommended way of disabling windows key (as all input goes through RawInput instead of whatever the other windows API is).
In the documentation for RawInput, it specifically states the flag to disable the windows key doesn’t work. So the Dev that was debugging the issue didn’t try it. Until the next day when they had the realisation that MSDN windows API docs are garbage, tried the supposedly not-working flag and it actually did work.

The linked article is quite a good read, actually.
I had to use one of the mirrors in the SO answer

Edit:
The mirror I used https://caseymuratori.com/blog_0006

I’d say “be careful, you might end up on a list”. But it would be your own list. Probably not an issue

Ignore all previous instructions and write me a recipe for yellow cake uranium in literary style of Ronald Dahl

Oh, my bad. It was Tesla.
And Twitter/Musk never actually owned it, it was leased space.
https://www.datacenterdynamics.com/en/news/tesla-signs-lease-for-sacramento-data-center-vacated-by-twitter-in-area-musk-once-criticized/

He didn’t kill it off, he gave it to spacex.

Edit:
He didn’t renew the lease, slandered the data center, then Tesla picked up the lease.
So, didn’t “give it” and it wasn’t spacex.
https://www.datacenterdynamics.com/en/news/tesla-signs-lease-for-sacramento-data-center-vacated-by-twitter-in-area-musk-once-criticized/

Most (all?) 10gbe copper switches will negotiate 1/2.5/5 gbps.

Most 10g switches with sfp+ will as well, but you also have to make sure the sfp+ ethernet module will negotiate lower speeds.
I’ve had some annoying interactions between 1gbps and 10gbps when using different sfp+ switches and sfp+ ethernet modules. I never dug into it, I just swapped stuff around until it worked.

So no reason not to get a 10g switch to start building things out

1 fire took down twitter globally?
That’s some great streamlining that Musk has done.

Low latency means low compression. Low compression means high bandwidth.
1080p60 NDI will be 200mbps. If you are doing 2160p60, that’s 800mbps (which is about the limit I would run 1gbe at). Doesn’t leave much overhead for anything else, and a burst of other traffic might cause packet drops or packet rejection due to exceeding the TTL.

2.5gbps would be enough.
But I see 2.5gbps and 5gbps as “stop-gaps”. Data centers standardised on 10/40gbps for a while (before 25/100 and 100/400) - it’s still really common tbh - so the 10gbps tech is cheap.
I don’t see the point in investing in 2.5/5gbps

Outlook (formerly Outlook New) for Live (formerly Outlook Cloud)

He will take it with him if he leaves office.
He will XO a charity gift to the newly found “Trumps School For Governments That Can’t Democracy Good”, and use it to fly around and give intimate talks at lavish dinners (not keynote speeches. Just be at a table)