They probably used AI to help write it.

Komodo

My experience with IPFS over the years has been abysmal, and I think people have said the protocol design cannot sustain any more growth, which is not even that big yet at all.

You also cannot realistically search for files reliably by its hash, because of how files are divided into smaller pieces, whereby the method of dividing can change between clients, making the hashes incomparable. BitTorrent v2 solves this to my understanding, but almost nobody uses it for some reason.

Often times you need to wait several minutes for IPFS to find a file, assuming it ever finds it, which sometimes fails even on two boxes next to each other.

I wondered how many hot takes there would be… was not disappointed.

buildbot.net

All you need in order to do this is for the client to encrypt their password before sending it to the server. Often services that advertise “zero knowledge” platforms that use end-to-end encryption will authenticate their users in this way. If this were a website for example, there could be a javascript/wasm library used within the client page that encrypts their password before a login request is sent to the server.

See Freenet/Hyphanet

Please don’t, because it is literally the largest place online that openly trades CSAM. Law enforcement even run their own nodes there to try to catch people.

How long until this gets overrun with 🍕 and nobody wants to use it…

Not sure how moderation would even be possible with this model.

which also removes the merits of you questioning if what he is claiming is fiction or not

yet you claim marcan is lina with zero proof?

If that does happen, I just hope there will be enough developers by then that can/will want to use it (as in, write rust code). Especially developers that can put up with the kernel process and its people.

Distros are still free to make their own RPM packages, they can’t go around the GPL there.

But having official flatpak release makes it very easy to update to the latest versions regardless of your distro.

rustdesk

last time I checked, blind users could not even install any mainstream distro anymore, because they all switched to wayland, and that broke screen readers in the installer.

Waste of time, they don’t even fix the massive bugs the app still has for years :/

Details like the fact that people can have differing opinions and perspectives, and should not speak in absolutes?

Very misleading writing style IMO. I would say most of their bullet points ARE actually true in most cases… they just keep bringing up somewhat rare/exotic exceptions as a way to call it a “falsehood”.

Care to explain?

I disagree, there are many resources for making and distributing android reproducible builds, including third-party F-Droid repos like IzzyOnDroid mentioned in my previous link.

And to my knowledge there is no technical requirement that F-Droid actually needs to build OR sign packages on behalf of anyone… I haven’t seen any actual official rationale listed for it, but I assume one of the main reasons is convenience for the developers so they don’t have to provide their own builds and deal with signing/losing keys.

I understand that the risk of problems can be somewhat mitigated in F-Droid by using reproducible builds, but I don’t consider that sufficient for the most privacy-conscious users because:

• reproducible builds are not required by F-Droid

• it is not made clear to the user that a particular package even supports reproducible builds

• the verification of reproducible builds is not made plainly visible somewhere publicly if at all

• a user can still easily be misled by a one-off rogue package that is NOT reproducible, due to the previous point

• independent verifications of those builds reliably made by others are not common

Conversely I stay clear of F-Droid as they build and sign packages on behalf of the original developers, adding yet another point of injection for malicious code or supply chain attacks.

In my experience… not really. I would say SDL makes the task of writing controller support code within your own applications easier and higher-level, but in reality it still has not much to do with “drivers” (I assume you mean kernel modules), which the kernel and OS stack already provide multiple unified interfaces for with things like jsdev/evdev/udev/hidapi, regardless of how you access those subsystems (via SDL or otherwise).