Recently I wound up using a FreeBSD machine, where I promptlyinstalled vim for my traditional reason . When I startedmodifying some files, I had contents to paste in from another xtermwindow, so I tapped my middle mouse button while in insert mode(ie, I did the standard xterm 'paste text' thing). You may imaginethe 'this is my face' meme when what vim inserted was the last thingI'd deleted in vim on that FreeBSD machine, instead of my X textselection.

Recently I read 32 MiB Working Sets on a 64 GiB machine ( via ), whichrecounts how under some situations, Windows could limit the workingset ('resident set') of programs to 32 MiB, resulting in a lot ofCPU time being spent on soft (or 'minor') page faults . On Linux, you cando similar things to limit memory usage of a program or an entirecgroup, for example through systemd ,and it occurred to me to wonder if you can get

A year or so ago I mentioned that I'd set up WireGuard on anAndroid and an iOS device in a straightforward VPN configuration. What I didn't mention inthat entry is that the other end of the VPN was not on a Linuxmachine, but on one of our OpenBSD VPN servers. At the time it wasrunning whatever was the then-current OpenBSD version, and todayit's running OpenBSD 7.6, which is the current version at the moment.Over that time (and

I mentioned recently a case when power cycling an (x86) serverwasn't enough to recover it , althoughperhaps I should have put quotes around "power cycling". The reasonfor the scare quotas is that I was doing this through the server'sBMC , which means that what was actuallyhappening was not clear because there are a variety of ways theBMC could be doing power control and theBMC may have done something different for what it described as a'power cycle'. In fact

In a recent entry on short lived TLS certificates and gracefulcertificate rollover in web servers , I mentioned that oneissue with software automatically reloading TLS certificates wasthat TLS certificates are almost always stored in multiple files.Typically this is either two files (the TLS certificate's key anda 'fullchain' file with the TLS certificate and intermediatecertificates together) or three files (the key, the signed certificate,and a third file with the intermediate chain). The core problemthis creates is the same one you have

Ever since its Solaris origins, ZFS has supported automatic NFS andCIFS sharing of ZFS filesystems through their ' sharenfs ' and' sharesmb ' properties. Part of the idea of this is that you couldautomatically have NFS (and SMB) shares created and removed as youdid things like import and export pools, rather than have to maintaina separate set of export information and keep it in sync with whatZFS filesystems were available. On Linux, OpenZFS still supports this, workingthrough standard Linux NFS export permissions ( which don

I recently wrote a little Python program where I ended up addingtype hints, an experience that I eventually concluded was worthit overall even if it was sometimesfrustrating. I recently fixed a small bug in the program; like manyof my bugs, it was a subtle logic bug that wasn't caught by typing(and I don't think it would have been caught by any reasonabletyping).

One view you could take of type hints is that they often don't catchany actual bugs, and

We have various sorts of servers here ,and generally they run without problems unless they experienceobvious hardware failures. Rarely, we experience Linux kernel hangson them, and when this happens, we power cycle the machines, as onedoes, and the server comes back. Well, almost always. We have twoservers (of the same model), where something different has happenedonce.

Each of the servers either crashed in the kernel and started toreboot or hung in the kernel and was power cycled (both were

Over on the Fediverse, I said something :

Current status: respinning an Ubuntu package build ( ... painfully ) because I forgotthe golden rule that when I add a hack to something, I should alwaysmake it log when my hack was triggered. Even if I can observe the sideeffects in testing, we'll want to know it happened in production.

(Okay, this isn't applicable to all hacks, but.)

Every so often we change oraugment some standard piece of software or