Follow-up: after some tests I discovered that the ISP is blocking the TCP layer, but UDP is allowed (e.g. dig and nslookup work). So my WireGuard connection can easily pass through.
Now I'm really curious to know how many of you are going to don't pay the next invoice to test this possibility ๐
View quoted note โ
daniele
_@dtonon.com
npub10000...vwqk
Working on https://fevela.me, https://nstart.me, https://njump.me, https://oracolo.me and other inspiring nostr projects. I love to build helpful things that people are pleased to use, mixing tech, design, usability and accessibility.
I just discovered a fascinating thing.
I've been struggling to solve an internet problem at my parents home for several days, a router and some other stuff cannot connect to internet; the strange part is that my PC and my phone work. My first thought was related to the fact that I use a VPN, but this doesn't make much sense, since VPNs usually are the problem, as they add one more hop to the connection chain.
I did a lot of tests and at the end, exhausted, I decided to order a new cheap router to see if this was the culprit. Today I installed it, same exact problem, damn.
So I noticed that the modem has a diagnostic area with the usual ping and traceroute tools, and I tried them: "unknown host".
This is really crazy, all the WAN/LAN are down, the modem itself cannot see the internet, but with my PC I can connect around and shitpost on Nostr?
So I contacted my ISP and discovered that they suddenly, and without any alert (bastards), disabled the line because there was an unpaid invoice from months ago for an unknown reason, and they are not smart enough to use the credit card they already have to automatically pay it.
Here comes the most interesting part: how does a VPN bypass the administrative suspension on a line and give me free internet?
Did I exploit a huge hole in my ISP (and who knows how many others) infrastructure?
Can any networking guru explain this?
Could #GrapheneOS ship a sufficiently-secure-and-degoogled fork for non Pixels phones?
Pixels are cool phones, but they are niche, expensive, and accessible only to a small portion of the population. If we want to promote a culture of privacy and security through free & open source software, we need a much broader and more flexible hardware base.
I know they are considering producing their own phone, but I fear this is not a solution for the average user. This is also because creating, distributing, and maintaining large-scale hardware support is an extremely challenging and risky operation.
Obviously, this hypothetical fork will not be able to guarantee the same security as a phone with selected hardware, but it will have a setup that will ensure good basic protection and, above all, *educate* users in the use of free and open software through a dedicated store. If it were possible to install it on $100 phone, we would have the opportunity to educate entire new generations of young people.
View quoted note โ
It's apparently a not so important issue, probably already discussed, but: what do you think if your client switched from the "Follows" label to the "Contacts" label?
"Contacts" seems more immediate, also for not English speakers, it removes the follows/following possible confusing overlap, and especially it opens to a more interoperable scenarios (DMs, "other stuff" use cases, etc). I don't see any particular drawbacks.
What do you think?
#asknostr
Imperfections as beauty.
View quoted note โ
Nostr prompt hackers: it is time to persuade Grok to make its own npub.
View quoted note โ
Small but significant bug fix for performance in Chronicle v0.4.1
View quoted note โ
GitHub
Release v0.4.1 ยท dtonon/chronicle
Avoid duplicate processing of the event when fetching the full conversation