Modern PC BIOSes, including server BIOSes, almost always have asetting for what the machine should do if the power is lost andthen comes back. Generally your three options are 'stay poweredoff', 'turn on', and 'stay in your last state'. Lately I've beenrealizing that none of them are ideal in our current 'work fromhome' environment ,and the general problem is probably unsolvable without internalremote power control .

In the normal course of events, what we want while

Back in 2018 I wrote about Firefox's media autoplay settings asof Firefox 63 . Unsurprisingly,things have changed in the Firefox world since then, and for oncethey've changed for the better, at least from my perspective. Myinterest is in basically never having video autoplay. And when Isay basically never, I mean it. In particular, I don't want Youtubeto autoplay the next video in a playlist or the like.

The basic control for media autoplaying (for

For a long time now, I've needed (or at least wanted) to simultaneouslyput a stress load on a test NFS fileserver and a stress CPU and memory load onanother machine. Back in the old days on Linux, I would have donethis by compiling the Linux kernel repeatedly, but in today's modernage the Linux kernel isn't big enough any more; it builds too fastand isn't big enough. I eventually settled on repeatedly buildingFirefox from source, partly because it

Plenty of people like backwards compatibility, especially strongbackwards compatibility. But it has what is a sometimes unpleasantconsequence, which is that a strong commitment to backwardscompatibility requires keeping your mistakes . Or at least many ofthem. To put it one way, you need to keep mistakes that work, and ofcourse you have to keep them giving the same result as they currentlydo. For example if you provide an API that people can use to expresspotentially conflicting things and you don't reject the attempt butinstead

Recently I wrote about my failure to arrange a graceful TLS rootcertificate rollover for our OpenVPN servers .This might leave you wondering what we're doing about this instead,and the answer is that we've opted to use a brute force solution,because we know it works.

Our brute force solution is to set up a new set of OpenVPN servers(we have two of them for redundancy), using a new official name andwith it a new TLS root certificate that is good for quite a while(

The surprising TLS news of the time interval is that Let's Encryptmade a systematic mistake in issuing all of their TLS certificatesfor years. Instead of being valid for exactly 90 days, Let'sEncrypt certificates were valid for 90 days plus one second . This isn'ta violation of the general requirements for Certificate Authoritieson how long TLS certificates can be, but it was a violation of Let'sEncrypt's own certificate policy.

TLS certificates have a 'not before' and '

When I started with LSP-based Python editing in GNU Emacs , the Python LSP server was pyls . However,pyls is apparently now unmaintained and the new replacement is python-lsp-server ,also known as 'pylsp'. I noticed this recently when I looked intotype hints a bit , and then when I was editingsome Python today, lsp-mode or some sub-component nagged me about it:

Warning (emacs): The palantir python-language-server (pyls)is unmaintained; a

Every since I started aggressively overriding websites so I could seewhat links I'd visited , it's become reallyclear to me that modern web design apparently hates making visitedlinks look different and goes out of its way to not do so. This ismost visible on the quite a lot of text focused websites that stillstick to more or less the original web colours of black text on awhite background with blue links. In the original basic web colourscheme, visited links would be a purple shade. But