Today, I wanted to check if our external mail gateway (which usesa TLS certificate from Let's Encrypt) was seeing a drop in incomingtraffic, since there are various problems cropping up as a resultof a Let's Encrypt related TLS root certificate expiry ( also ). Weextract metrics from our Exim logs using mtail , so the raw information wasavailable in metrics, but we get a sufficiently modest amount ofexternal email that our natural external mail arrival rate isrelatively variable on a minute to minute

When I started writing Wandering Thoughts , this blog, Itended to link to external URLs using at best a few words aboutwhat they were, and sometimes even less obvious link text andsurrounding text than that. This is a style that's still quite infavour on the web, in blogging and other usage, and you can findit used by many people (sometimes even me, still). However, overtime I switched to more often quoting the full title of the page Iwas linking to,

Today I noticed that one of our servers was regularly logging aburst of kernel messages about problems with ACPI data. These messages looked like:

ACPI Error: No handler for Region [POWR] (000000000c9d7b92) [IPMI] (20190816/evregion-129)ACPI Error: Region IPMI (ID=7) has no handler (20190816/exfldio-2

Somewhat recently I read Ariadne Conill's understanding threadstack sizes and how alpine is different ( via ),which is in part about how Alpine Linux has a very low default thread stack size, unlike other things, andthis can cause program crashes. As part of this, Conill says:

In general, it is my opinion that if your program is crashing onAlpine, it is because your program is dependent on behavior that isnot guaranteed to actually exist, which means your program is notactually portable. When

Back in 2016, I wrote an entry about how we'd never be able tohave everyone use two factor authentication .Recently, the university (and thus necessarily my department ) is moving to make two factorauthentication required. Given this, you might wonder what happenedbetween 2016 and now, five years later. The simple answer ischanging attitudes on smartphones, or if you prefer, a view thattwo factor authentication is sufficiently important that theadministration is willing to force people to go

One of the frequently noted things about OpenBSD is that its releaseshave a short support period. OpenBSD generally releases securityand other updates for the current release and the previous one, andonly updates third party packages in the current release ( cf ). Since OpenBSDreleases more or less every six months, this gives you at most ayear of support for what you install. One of the practical resultsof this policy is that people wind up running unsupported releases , but let's suppose that you want tokeep with

Ars Technica's Examining btrfs, Linux’s perpetually half-finishedfilesystem ( via ) isnot very positive, as you might expect from the title. I found ita useful current summary of the practical state of btrfs, which isby all accounts still not really ready for use even in its redundancymodes that are considered "ready for production". There's probablynothing new for people who are actively keeping track of btrfs, butnow I have something to point to if people ask why we'

Earlier this month I wrote a blog entry about Firefox having jankwith WebRender turned on , inwhich I mentioned that the problem had appeared when I stopped usingan environment variable $MOZ_X11_EGL that forced Firefox to use"EGL". The blog post led to someone filing a quite productiveFirefox bug ,where I learned in passing that Firefox is switching to EGL bydefault in the future. This made me realize that I didn't actually knowwhat EGL was, where it fit into the