Fair point but it is already protected by the box

At least remove the screen’s protective cover :)

looks like an open source project got fucked over by a particular overpriced hardware company

Nice.

You send your binary to Apple and they do some checks before signing it. That’s what “Apple checked this application and found no malware” prompt is.

Also, those signatures can be revoked or even blacklisted if required.

This is pretty misleading, as it shows nodes, not capacity, and excludes Tor.

There’s also the fact that at least for a lot of the smaller share ones, they may be used just for proxying for a public IP.

Not that this means anything as if it is shut down, new nodes can come up to replace them in other networks

that and Solana, the other networks at least work sanely

Solana falls apart because they somehow managed to fuck up even PoS beyond repair that also means the network can go down due to one bad node

these deserve a new category

an LLM is just a big data aggregator, anyway.

if it can be solved easily then it probably should be, and that time spent on actual development work instead of centering a div

after some point you run into code quality problems, but as long as you don’t trust but verify, it’ll work

Avoid any with ATECC chips. Trezor Safe checks most of my boxes, including being able to use easily with another app.

Or you can use a SeedSigner depending on what you are doing and you would like to include stateless signing devices and not just HWWs.

What HWW would you recommend?

This issue has been resolved nostr:note16pgwjrgdxh8dvkpwn6676x0lr47tlfxw9sv67u05n9z0asc7xuss0kj8s3

Some nostr.land services are temporarily unavailable due to a networking issue with the upstream.

The issue is currently being resolved.

a raspberry pi and a camera

you will find that you can use standard protocols like RTSP

It never can until users stop having to manage nsecs in a very error prone way.

I guess the best way to redesign would be to have a set of keys (with their bunker nsecs), a set of users, secrets for pairing, and already authorized connections (possibly with descriptive names)

I see several minor issues:

- why encode nsecs as bech32 and not hex?

- bunker npubs should be limited as much as possible (1 key = 1 npub or 1 key per bunker) as it adds additional filtering costs to relays, and therefore be independent from users

- secrets only need to be kept for pairing and should most likely be separate from users

and some other issues

It’s a shitty processing pipeline

PDF > PostScript (rasterized PDF in PS) > PS again > PPM > some color processing > PCL XL

And the PDF to PS converter + Ghostscript is really slow. And it’s performed sequentially.

Some things:

rastertopclx falls apart on sRGB raster

pdftoraster is slow as hell, use mupdftoraster or pdf2printable from ppm2pwg instead

Use the CUPS PPD compiler instead of writing your own

Ghostscript is slow as fuck

A lot of printers have weird PJL commands they want to just work nostr:note1t3n8dma7yhjvkqevacc2dsxtda94u2fue6557eagrm0ange5lthqzj8ct5

My printer’s drivers took 10 seconds to rasterize a single page, so I made my own CUPS filter chain to output acceptable PCL for the printer

how tf do you get 24h?