ayaya

I don’t think Microsoft is capable of not fumbling everything related to the Halo franchise.

One of those rare games with an actually good native Linux version

You just reminded me there actually was a browser called Torch that could download torrents like a normal download. It was basically just Chrome with a built-in torrent client.

I remember trying it out when it first came out in 2012. It never caught on and looks like the last release was in 2020.

I guess by “cybercrime” they mean piracy, because that’s the main thing I’ve seen .su used for.

PIA is the best for torrents. It is $79 for 39 months which is $2.03/mo and they have port forwarding. That’s less than half of pretty much every other provider.

I have had 3 clients (one for a specific tracker, one for everything else, and an extra seedbox) going 24/7 for years with no problems. No complaints about the speeds either. I frequently saturate full gigabit on both downloads and uploads.

deleted by creator

deleted by creator

If you want to do less math you can just drop some zeroes and say it’s the same as making $70k while losing $2.50

+1 for MXroute. I have unlimited domains with 25GB of storage for $30 every 3 years. So less than a dollar per month. Looks like they are still offering it. It’s more than enough for email especially considering the Gmail account I used for 15 years was under 5GB.

I switched to them at the beginning of the year so about 9 months ago and have not had any issues.

Even on Linux where their drivers are supposed to be better, my 7900XTX has been crashing randomly for at least a month and it was only fixed in the latest 6.10.9 kernel release yesterday.

It might be their third most popular country but that doesn’t necessarily mean it’s profitable. Brazil isn’t exactly known for spending a lot on things like Twitter Blue or ads, especially per capita.

I have a theory they looked at the numbers and realized that shedding the load would save more money than they would gain by staying. But then again that might be giving Elon too much credit.

deleted by creator

This breaks any site that uses CloudFlare’s Turnstile for me. It will loop forever and never let me through if my user agent is set to Chrome.

Even if they fully render them into the video with absolutely no way for an extension to tell where it is something like Sponsorblock where people manually enter time codes could still get around it.

Plasma actually has a UI for smart TVs if you weren’t aware, although I have never used it myself so I’m not sure how good it is. https://plasma-bigscreen.org

It should all be opt in

Then you introduce self-selection bias and the data is worthless.

Aggregate data can be used to personally identify

You can’t identify someone based on how they interact with a service. If you spend 5 minutes on one page and 2 minutes on another that could be anyone. Even if you for some reason personally knew someone’s browsing habits it would be nearly impossible to pick them out in a sea of millions of data points.

I see you linked privacyguides.org in the thread as “alternatives”, one of the services it recommends is Proton (Mail, Drive, etc.). Look at their privacy policy:

2.1 Visiting proton.me or protonvpn.com website: We employ a local installation of self-developed analytics tools. Analytics are anonymized whenever possible and stored locally (and not on the cloud). IP addresses are not retained and stored for such analytics.

When you use our native applications, we (or the mobile app platform providers) may collect certain information. We may use mobile analytics software (e.g. fabric.io) app statistics and crash reporting, Play Store app statistics, App Store app statistics, or self-hosted Sentry crash reporting to send crash information to our developers in order to rapidly fix bugs.

Or how about addy.io that privacyguides recommends for email forwarding? From their privacy policy:

We use a self-hosted instance of Umami, an open-source, privacy-focused and lightweight option for website analytics. All the site measurement is carried out absolutely anonymously.

ALL online services collect this kind of data. Even the privacy-focused ones. There is nothing nefarious about it.

Like the comment I replied to already explained, this information is necessary to make informed development decisions. If you don’t know who is using what feature you might be wasting resources on something barely anyone uses while neglecting something everyone needs.

You also need some of that data for security purposes. You can’t implement rate limiting or prevent abuse if you can’t log and track how your services are being interacted with.

And this is aggregate data. I can promise you not a single person cares about what any individual user is doing (assuming it’s not illegal)

Yeah as someone who has worked in web development for over 20 years everything in here is completely standard. Almost every major website in existence collects this kind of analytical data.

Again, even an exact copy is not stealing. It’s copyright infringement. Theft is a different crime.

But paraphrasing is not copyright infringement either. It’s no different than Wikipedia having a synopsis for every single episode of a TV series. Telling someone about what a work contains for informational purposes is perfectly fine.

If the model isn’t overfitted it’s also not even copying. By their nature LLMs are transformative which is the whole point of fair use.