Thankfully, it's almost ideal conditions for a project, the road is already there, just not sealed, it's flat, very flat. That means no tunnelling, landscaping, viaducts, moving existing infrastructure.
There's no one to complain about anything, no compensation to give for heavy works, no urban traffic management, etc. The hardest part is logistics and actually paying someone to basically go out there for absurd spans of time slowing moving up the road until completion.
A questionable Roblox shirt and an six-year-old Discord moderation comment are not in the same evidentiary universe as criminal sex crimes, in like, any reality, and it truly demonstrates how desperate you are to try and find wrongdoing with this person.
Also, just to drill the nails into the coffin that is your argument, Coolyoshiluigi99 existed as far back as 2020. So it wasn't even a recent change in name, he's had that name since at least 2020, and likely prior. He cannot hide previous usernames.
Yes, the shirt that he was wearing was seemingly vore coded. Take it up with Roblox content moderation.
Any image uploaded to Roblox has be viewed by a moderator to confirm it follows platform rules, apparently, this did.
These are also frankly things that are years old? His avatar is no longer this, it changed at least after 2022. This is on a timeframe of years, and seemingly no one was bothered by it, until you came along and decided to dig up dirty laundry and create drama.
If you go through almost anyone, you'll find slip-ups. Considering that you've excavated back to 2020, either you're terrible at searching or there is practically nothing.
Weird: Yes
Dangerous to Kids: No
Are You Helping Anyone: Fuck No
@Btrfly0blivion "No hate, but can everyone please look at this person, infer the worst possible motive, ignore context, and help me convert vague discomfort into a public accusation?"
You're rightfully getting fucked over a barrel because everyone can see through your manipulative bullshit.
Bluetooth is trying to make sure that you still can hear decent enough audio when stuck in a metal tube flying underground with 300 other devices screaming on 2.4GHz near you, with tiny devices, with small batteries with strict sync, over possibly two or more individual devices? Very different to what Moonlight is trying to do.
@rickc42069@SimiStern Yes, precisely.
That's one of the biggest things that you gain with self-hosting over managed-hosting.
The entire value proposition is that they handle all the hard bits for you, which is great until that catches fire, you have customers howling at you and you cannot do anything.
"Expected metal" doesnโt mean "every critical dependency must be on that same metal." In fact, putting all recovery/control-plane state on your own platform can be how you create an outage like the Facebook 2021 outage. The real issue is whether the GCP dependency had too much blast radius, not whether any GCP dependency existed at all.
"What if our own outage means we canโt access internal systems or recover the platform?" is going to come up in engineering meetings way more often than "what if Google just kicks us out of the account?"
@AArdvarkErick@neetcode1 Railway owns the architecture and the blast radius. Google owns the account/control-plane action. Customers are allowed to be angry at both, but pretending this has nothing to do with GCP because it was "just" the datastore is not serious.
@shu_building@SimiStern Managed pipelines are great until something goes wrong in that pipeline. Automatically managed pipelines are based on a set of very specific instructions going perfectly every single time, the law of the universe says added automatic complexity means more points of failure.
@shu_building@SimiStern From what I've seen, yeah, you're right. But I'm considering this specific outage in a vacuum based on the circumstances.
I don't like these services since you have limited control when something goes wrong and the only yelling you can do is at your account manager over email.
@draver461573@SimiStern@googlecloud https://t.co/lXuXQfMyyU
This seems to be what happened, Railway was worried about circular dependencies, database migrations are hard. Google Cloud unexpectedly threw a brick at them, single point of failure that on paper should've never been there, but reality is more complex.
Yeah, I did look into it a bit more, and they do indeed advertise that. Thatโs quite a bit more awkward then. They shouldโve at least had some credible fallback path on their own premises. That being said, hosting things yourself on your own platform can also create weird circular dependencies.
I donโt even like Railway, or the "managed hosting" ecosystem in general. But the actual problem here is the continued bullshit @googlecloud keeps pulling: "oopsies, our finger slipped :3" account/control-plane failures that can take real infrastructure offline, while everyone keeps pretending this is acceptable behaviour from a serious cloud provider.
@SimiStern I mean, yeah, sure. But Google Cloud suddenly blocking Railwayโs account is the kind of black-swan-ish platform dependency failure thatโs very hard to fully design away. What are you supposed to do? Replicate everything on two different cloud providers, self-host everything?
Labor should absolutely have expected the backlash and had a much clearer positive case ready. But I don't think a startup carve-out is some pathetic capitulation. Passive capital gains from property speculation and genuine founder/employee equity risk are not the same thing economically, and treating them identically is how you turn a good reform into a dumb one. The carve-out should be tightly designed so it doesnโt become a loophole for wealth to cosplay as innovation.