Funwayguy

I have similar setup on my home server except my Crowdsec & Anubis live in a dockerised VPS proxy that tunnels over WireGuard to the real server behind OPNSense. Saves me a ton not having bandwidth obliterating AI crawlers hit my internet bill. As an added bonus, I can move or shutdown the public VPS entry if things get too ‘spicy’ without any downtime on connections via the VPN.

I dumped Portainer on my home server when they did the ‘Fortnite’ graphical rebranding. Now using Arcane to manage my servers and edge nodes. Very much enjoying the CVE scan tools and in-place image updating.

Delta Goodrem is another artist like this who locals in AU would consider ‘famous’ but otherwise don’t make it on the international charts let alone bring anything unique that anyone cares to remember. It’s even worse watching them pad their gigs with covers of bigger artists’ songs. If not for constant TV appearances I suspect these artists would struggle to stay relevant even locally.

I too wish it was further ahead but when I think of the decades of duct tape MS has built around the NT kernel with horrors such as 95, Vista and 8 drivers, it makes sense that reverse engineering all that in a clean room is no easy feat. Their work certainly isn’t wasted either.

Not that it matters anymore. GN summarised it best stating that we’re already long past the point where the extortionate prices matter to even enthusiast consumers who wouldn’t buy this shit at 3x MSRP anyway. Even prebuilts are selling empty case ‘kits’ to keep inventory moving.

This approach only works in daylight with cameras that rely on colour only. Infrared spectrum cameras such as what Australia uses on mobile phone traffic cameras are much harder to deal with. Patterns don’t cut it and it’s a huge part of military research into hiding vehicles from infrared spotters.

Disabling the cameras entirely is much more effective.

Less sites are doing that to intentionally avoid tipping off the AIs about thresholds and trigger behaviours. The two main solutions are browser crypto challenges like this, or silently diverting flagged connections to a tar-pit of fake pages. The former is easier to implement but the latter actively sabotages the scraped data.

That would imply preserving the original. No, they are having their AIs revise history and its literature to whatever they deem ‘acceptable’ in their new world order, hence the rampant book banning spree they’ve been on as of late.

But they did such a great job of feeding the world’s most problematic corporations government IDs and photo identification of minors and storing it in databases ripe for hackers.

Brought to you by the government whose national health network failed its security audit last year, and had 13 million records breached through a 3rd party the year before that.

This is the part that the “walled garden” arguments I see conveniently ignores so often. Humble games and key reselling, good and bad, exist because of this rule. That 30% otherwise covers Steam dealing with sales, key distribution, payment providers, and all the legal liabilities that comes with that for you. Unless you can securely and continuously run your own shopfront below that 30% margin, there’s not a whole lot of incentive to do so… But Steam isn’t stopping anyone, not even Epic in fact. The so-called wall is like a foot high.

The DRM isn’t even that deep either and has known tools to remove it if you want. It exists as a bare minimum requirement for copyright law and Steam friends but not much else, hence why publishers often use things like Denuvo still.

We don’t ‘defend’ Valve’s monopoly so much as they really aren’t doing anything special to maintain it besides making Steam libraries accessible on more hardware. They compete by merely existing in the same space.

I think anyone who’s paid any attention to targeted advertising has know this long before LMMs were a thing. Removing the ‘easy answer’ is just a legal formality at this point and does little to nothing for the privacy it claims to protect. Advertisers like Google like it that way and will fight tooth and nail against any ‘reconsiderations’

From a techical perspective I’m still annoyed that the amorphous blobs were not the agreed default. They are the simplest to store in SVG, have no gradients to render, and being non-human makes them not require skin/gender modifiers. All expression, nothing more.

I have my reolink cameras setup on an internal network without direct internet access, but have a server running Frigate and a VPN that I can remote into from my phone. Gives me full control of where the recordings are backed up and remote access controls. This setup works for their doorbells too which is neat.

I got the same treatment recently. All tech departments were issued M4 Mac Book Pros because that was more cost effective than than dealing with the non-compliant fuckery of W11. Unfortunately non-tech departments got the old inventory and are suffering the abhorrent instability of W11. It somehow refuses to play nice with just about everything in our corporate ecosystem.

See that’s the fun part. Google is the ad company so it’s all 1st party data. Google can package the Trojan horse however they please, which why it’s such a fine line for the blockers to walk.

It wouldn’t matter whether it was intentional or not. Put simply, Google can continue indirectly punishing creators for tolerating adblockers then redirect blame, even though they could have easily separated the metrics from the advertising and telemetry endpoints that blockers filtered. This way they get their money either from unblocked ads or from creator’s reduced view counts, win-win for Google.

As an added bonus for Google, by ensuring view metrics get fucked up, it double punishes creators featuring sponsored content that rely on those metrics to determine how much the sponsor should pay them. Meanwhile Google could, in theory, sell ad placements attached to their own internal metrics that differ from the affected ones publicly visible.

Running that much power next to a data line sounds like a terrible idea for signal integrity, especially if something shorts to said data lines. It just sounds sketchy or filled with so many asterisks that it’s functional impossible to reach their claimed throughput.

It really isn’t. At best it can generate short snippet solutions but it absolutely botches projects that require contextual knowledge of the program stack.

My job actively avoids because of the havoc it tends to cause not to mention legal issues.

I’d like to think Australia has a nice middle ground design to their sockets/plugs without the foot destroying bulk. Still get the shutter variants for bathrooms too.

It exists to abuse and profit off search engine optimisation. More than half of most image search results are polluted with their results which forces you sign up to do anything useful with them.