#Sockets

2025-07-10

I'm in a self-imposed std::os::net::UnixListener, std::sync::mpsc, mio::Poll hell today.

For... reasons... I was trying to do something "simple" and just use a single thread for accepting connetions and a single thread for handling the client. But turns out that's far more complicated than just depending on tokio and doing everything with async.

KMJ πŸ‡¦πŸ‡Ήkmj@mastodon.ctseuro.com
2025-07-06

Decided to trash this 4 #matter #nous #a8m #power #sockets because of the need for full Google Play AND be logged in to Google,

Install the full app on a trash phone with no Google play and no Google Login failed too.

This and this DCL is unacceptable for privacy.

So will stay away from Matter. Hopefully others with privacy interest does not loose money too.

#homeassisant doc says it fully local, so there should be a correction to the website to make peopel aware!

community.home-assistant.io/t/

KMJ πŸ‡¦πŸ‡Ήkmj@mastodon.ctseuro.com
2025-07-06

Looks like i need to trash this box of 4 # matter #nous #power #sockets because it looks like to pair them with #homeassistant one must install the #homeAssistantCompanioApp from goole play, whoch is impossible on a degoogled #GrapheneOS mobile. All this for sending the Wifi Data to the device. This is sick IMHO and if this is the only was matter will stay in the trash for me forever.

2025-07-05

Last 2 days #coding #antibrowser #gui #apps in agnostic shell wrapping for any language (22 so far including #ada #cobol and #sed (yes sed)) shown scenario 74 apps on a single machine, slower than spreading out the tuples server,client,test driver on 3 different machines. Still effin fast because no #browsers, no windows manager no fekking bloated libraries, nothing but #C and #sockets no fonts either! don't need em

grayrattusgrayrattus
2025-06-24

Ok hear me out

which connect between each other seem to be an antipatern.

Let's think outside the box for a moment.

Throw them out replace with single layer between client and server. Every microservice which you had either needs to be converted into a library or use domain to communicate.

The only exception is communication between data-stores. For that you implement separated service.

Diego CΓ³rdoba πŸ‡¦πŸ‡·d1cor@mstdn.io
2025-06-04

Enorme artículo acabo de encontrar sobre packet flow en #netfilter, #iptables y #nftables! 😲

πŸ‘‡

Nftables - Packet flow and Netfilter hooks in detail

thermalcircle.de/doku.php?id=b

Si les gusta #linux, #kernel, #netfilter, #routing, #sockets, #vpn y temas afines, el blog completo de Andrej Stender tiene varios posts muy interesantes!

Felix Palmen :freebsd: :c64:zirias@bsd.cafe
2025-04-16

On a #coding mission to improve my #poser lib 😎.

In the current implementation of #swad, I don't really like that I need an extra thread, just to control a child #process. A first piece to add to poser is generic "child process support", which I'm testing right now. I realized I could reuse my #Connection class, which was built for #sockets, but works just as well with #pipes πŸ™ƒ

TODO now is mostly testing. See screenshots for some mimimal testing code and its output ... would you like this kind of interface? πŸ€”

#C #programming

Testing code for the new PSC_Process classTerminal output of the testing code
Dendrobatus AzureusDendrobatus_Azureus@bsd.cafe
2025-02-05

@DianeBruce @stefano

At least we are lucky enough to have experienced that time

πŸ–‹οΈ #bash #sh #zsh #ksh #csh #sockets #Linux #POSIX #NIC #DNS #relay #Usenet #networking #programming #freeBSD  #Fediverse

Dendrobatus AzureusDendrobatus_Azureus@bsd.cafe
2025-02-05

One thing I like about that period in which the System Operator could literally see what you typed; the Age of UUCP and Usenet, way before www, is that there was something on those federated networks that we shall not easily get nor easily gain again;

**TRUST**

Inherited Trust of the period was so beautiful. All the Networks are open and there were no trolls that could do any real damage.

Many decades have gone by, but I've now finally had contact with an actual Operator of a Usenet node. Thank you for your reaction

@DianeBruce
@stefano

πŸ–‹οΈ #bash #sh #zsh #ksh #csh #sockets #Linux #POSIX #NIC #DNS #relay #Usenet #networking #programming #freeBSD  #Fediverse

Dendrobatus AzureusDendrobatus_Azureus@bsd.cafe
2025-02-04

Do understand that all federated networks need relays, many, so that reach is increased, otherwise you'll end up having to wait hours to days before posts reach wider audiences

Excerpt:
{
One of the biggest advantages of relays is that they help smaller instances thrive. When a new or niche server joins the Fediverse, it often starts as a quiet place with limited interaction. Relays can dramatically change this dynamic. By participating in a relay, a small server’s content reaches a larger audience. This doesn’t just give more visibility to the server – it helps its users feel more integrated into the broader Fediverse community. People are more likely to stay active on their instance when they see others engaging with their posts.
}

Low reachability means you will have to wait long times, just like I did in the UUCP days. We could only afford one dialup session of 60 minutes per week to upr2.clu.net.
So our reach / ping time was 7 days!!!

πŸ–‹οΈ #bash #sh #zsh #ksh #csh #sockets #Linux #POSIX #NIC #DNS #relay #networking #programming #freeBSD #Fediverse

Dendrobatus AzureusDendrobatus_Azureus@bsd.cafe
2025-02-04

In order to better understand federated networks you should keep in min the following exerpt from the article

Excerpt: An ActivityPub relay acts as a middleman, **bridging multiple servers** to facilitate broader content sharing.

**Without a relay, communication across instances relies entirely on direct relationships between servers**

This was also the case with classic news servers and SMTP mail servers.

In my country I was one of the three people who brought the internet in fruition. Explicit detail is on my github {message me for details}
When we dailed from 454074 to upr2.clu.net we spoke UUCP with that big server in Puerto Rico. Our mail could only reach the rest of the world when upr2.clu.net pushed them to the upstream servers.

Upstream pushing could take days.

Relays simply tell the network that your server can send the traffic and they relay them to others, working like a classic network hub

BTW upr2.clu.net still runs, incarnated in a modern form (screencap included)

Go read about how it works on this Federation and learn

@altbot

gyptazy.com/hosting-a-new-acti

πŸ–‹οΈ #bash #sh #zsh #ksh #csh #sockets #Linux #POSIX #NIC #DNS #relay #networking #programming #freeBSD #Fediverse de

dark-themed webpage shows a profile for "gyptazy," identified as a "Deve." A cartoon penguin-like avatar is featured. A sidebar contains links for "About," "Blog," "Talks," "Projects," "Skills," and "Contact." The main content displays an image of a server rack labeled "Fediverse Relay," accompanied by text explaining the Fediverse and ActivityPub protocol's role in decentralized communication. A smaller screenshot shows what appears to be a web browser window. Social media icons (GitHub, LinkedIn, Twitter) are also present under the profile name. There are also buttons for "GPG" and "Blog" at the bottom.screenshot shows a terminal window displaying the results of a ping test to upr2.clu.net, with 21 packets transmitted and received, and 0% packet loss. The ping statistics show individual response times in milliseconds. To the right, a portion of a web browser window is visible, displaying a Mastodon instance's homepage and some hashtags related to networking and programming.
Dendrobatus AzureusDendrobatus_Azureus@bsd.cafe
2025-01-28

@altbot @micr0 @stefano @gyptazy

The image you see in the first post is a efficient status page as generated by the boxyBSD Network

boxybsd.com/status

πŸ–‹οΈ #bash #boxyBSD #sh #zsh #ksh #csh #sockets #freeBSD #netBSD #openBSD #POSIX #programming #ProxMox

Dendrobatus AzureusDendrobatus_Azureus@bsd.cafe
2025-01-22

@mboelen

Thank you for that valuable link

I've book πŸ“š marked
linux-audit.com/system-adminis

πŸ–‹οΈ #bash #sh #zsh #ksh #csh #sockets #Linux #POSIX #commandline #proc

Dendrobatus AzureusDendrobatus_Azureus@bsd.cafe
2025-01-17
Dendrobatus AzureusDendrobatus_Azureus@bsd.cafe
2025-01-15

Another wonderful view of the boxyBSD hypervisor status. Make sure you get your hours of studying to build a *BSD config from scratch by working on it, in your VM

boxybsd.com/status

πŸ–‹οΈ #bash #boxyBSD #sh #zsh #ksh #csh #sockets #freeBSD #netBSD #openBSD #ProxMox

A dark-themed webpage displays server status information for boxybsd.com.  The top shows the stylized text "BoxyBSD" in a blocky ASCII art font. Below this, a command line interface shows the command cat status.md being executed.  The results show the status of six hypervisors, each with a location (France, Germany, Canada, and Netherlands), miscellaneous information about the website and services (all online), and statistics including 500+ boxes provisioned, 7 OS images, and 99.9% uptime. 
Radio AzureusRadioAzureus
2025-01-10

@drunkenmadman

A nice find. Thank you for sharing
With plain
>>
$ traceroute bad.horse
traceroute to bad.horse (162.252.205.157), 30 hops max, 60 byte packets

I get this output. For me its closer

πŸ–‹οΈ

The image you sent appears to be the output of a network diagnostic tool, likely a ping command or a traceroute.

Here's what we can interpret from the information:

1. Hostnames and IP Addresses:

    The output lists a series of hostnames and their corresponding IP addresses (e.g., "63.245.90.162").

2. Ping Times (Latency):

    For each hostname or IP address, multiple ping times are listed, usually in milliseconds (ms).
    These times represent how long it took for a network packet to travel from the source to the destination and back.

3. Network Hops (Traceroute):

    It's possible that this is a traceroute output, showing the path that network packets take to reach a destination.
    Each line might represent a "hop" or a device (router, switch, etc.) along the path.

4. Network Performance Issues:

    Some of the ping times are relatively high (e.g., over 100ms), which could indicate potential network congestion or issues with the network path.
    There might be some packet loss, indicated by asterisks (*) in some lines.

Without more context, it's difficult to say definitively what this output means. However, it provides some basic information about network connectivity and performance.
Radio AzureusRadioAzureus
2025-01-08

Continuing with the handbook while playing in the sandBoxy

docs.freebsd.org/en/books/hand

πŸ–‹οΈ

A screenshot of a FreeBSD documentation webpage section titled "3.3. Users and Basic Account Management".  The left side shows a table of contents with chapter headings like "Synopsis," "Virtual Consoles and Terminals," and others related to FreeBSD system administration. The main section explains that FreeBSD allows multiple users, detailing how to manage user accounts, including adding, removing, and modifying them, as well as setting resource limits and creating groups.  A sub-section, "3.3.1. Account Types," describes system accounts, user accounts, and the superuser account, with examples of system accounts like daemon, operator, bind, news, and www.
Radio AzureusRadioAzureus
2025-01-08

installed zsh & bash in boxyBSD for a nice variance in shells.

sh is the default shell

building up freeBSD from *sh is fun and brings memories from doing that with Linux in the pre-Alpha days where you had to do EVERYTHING including build your own boot device image and using rawwrite to dump it on that 3.5" DSHD disk

πŸ–‹οΈ

Client Info

Server: https://mastodon.social
Version: 2025.04
Repository: https://github.com/cyevgeniy/lmst