Pals, what's the least egregious TV I can buy today, in the UK?
I want as little "smart" internet-connected nonsense as possible. Not bothered about 4k or massive size, but it should sound and look good.
@christianp we bought our most recent TV in 2011 IIRC. It had better work for another decade because buying a new one will turn me into even more of a grumpy old man. 50 inches?? Why not just live in a cinema??
Something I don't get about #eduroam is that I'm typing my own password into a wireless access point controlled by another university.
Is it just inertia that has prevented it from moving to a web-based SSO process where I log in to my own university's website, and they pass a signed token on to the institution I'm connecting to eduroam from?
A couple of weeks ago, someone crashed a car on the big road next to the pedestrian and cycle path that all the local kids use to get to school.
About a week later, someone came and towed the car away, but they left a lot of the debris scattered on the path. My daughter's bike got a puncture from rolling over it.
Another week has passed, and nobody's cleared it up. At some point the council mowers have been down and rolled over it all, so there were loads of shards of sharp metal, plastic and glass scattered over about 20m, on the pavement and hiding in the mulched grass.
So on my lunch break I decided to pick it up myself. Got some lower back pain for my trouble! #WhitleyBay
Going to join the #KidicalMass this afternoon, from #WhitleyBay to #Tynemouth. Apparently enough of the new cycleway is now open that we can use it. It's been a long time coming!
Who would know how to make a guess at the answer to this question: is internet traffic pretty much equal numbers of 1s and 0s, or is there noticeably more of one than the other?
Just remembered that last night I dreamt that Liz Truss had managed to get back in as PM, and she was tooling around in a car with "GOOD LUCK 2TH TIME" painted on the side
This morning I'm looking at #typst.
The first thing in the tutorial (https://typst.app/docs/tutorial/writing-in-typst/) is how to write a header, and it annoys me that it makes the same mistake HTML, and everything following HTML, made: you specify the level of the heading absolutely, and it's not scoped to a section of the document.
So when you want to have a heading one level lower, you have to know what level the previous heading was. And you can't tell how much of the document the heading applies to, only inferring it as going until the next header of the same or higher level.
I've always wondered why #TeXLaTeX has
\section{name}
instead of
\begin{section}{name} ... \end{section}
@christianp that's a really interesting thought. It seems like you want a markup kind of language with scopes which can be composed, and the ultimate meaning of any bit of syntax depends on how the whole thing is composed together.
It actually reminds me a bit of Haskell, except Haskell code defines a program not a document to display. Probably could be nearly any other programming language though.
I can't think of anything off hand (XML doesn't count because it doesn't have any canonical display semantics), but it seems like someone must have played with something like that at some point!
Elsevier, everyone's favorite copyright maximalist closed-access publisher, argues that their high costs are necessary because they're the arbiter of quality.
The arbiter of quality keeps publishing LLM-written papers. Thanks for making my argument for me, Elsevier! They didn't even read it.
"In summary, the management of bilateral iatrogenic I'm very sorry, but I don't have access to real-time information or patient-specific data, as I am an AI language model."
@cliffle I'm not sure if it was that paper but I saw one this week where the very first sentence started something like "Certainly, an argument for ... could start..." in that LLM chatbot style.
It's just so brazen to not even double check the opening sentence of something you're trying to pass off as your own. Were they rushing that much to submit??
I kind of feel that not enough is made of how the far-right has weaponised the internet and is totally undermining democracy. Like everyone knows it but is there much being done to oppose it? At this point I think there's every possibility that reform uk will win loads of seats in the next election here, and win the election after that, just by manipulating public opinion on otherwise uncontroversial issues like active travel and human rights.
@yaxu I think you may be overestimating Reform's chances. Even though they're polling at roughly 10% now, that is spread around geographically and likely won't get them any seats at all. See https://www.electoralcalculus.co.uk/homepage.html
They may get some council seats in the local elections, which certainly isn't good, but probably isn't that strong an indicator for a general election.
In my opinion, the reason the Tories are lurching to the right is that they have lost a lot of broader support and are now fighting it out amongst themselves to slice up the much smaller pie they're left with. They're appealing to Tory party members, people who've voluntarily paid into the Tory party, rather than Tory voters, and there are roughly 2 orders of magnitude fewer members than voters.
All that said, complacency isn't the answer and there's clearly a growing far right tendency globally, it's just the UK voting structure means it's quite hard for narrow issue parties to get much electoral traction.
Academic pals: do any of you work in an institution with a functional system for requesting help from your IT people?
We have a terrible ticketing system which, apart from its inaccessible web interface, is really hard to use: if you want to request help from someone, you have to browse through a list of dozens of ready-made forms for "common tasks", invariably none of which fit what you want. So there's a single "report something is broken" form with a single free-entry text field, and somebody on the service desk triages it. So all you have to do is resist the urge to report the ticketing system is broken, and hope that you can describe who you want to reach well enough for the service desk to deliver it correctly.
@christianp if you can find a database vulnerability, maybe you can create a super secret ultra high priority ticket (reserved for the chancellor or something) and actually get a useful response.
Important update from the "is this prime" game: 87 is now by far the most incorrectly tested number, ending just under 1 in 7 games.
After that, 51 and 57 are still almost neck and neck, ending 4.6% of games each.
An annoying aspect of having mathematics training: if you want to learn about a new topic, X, you'd like to read "X made easy", but most people's idea of "X made easy" is to present X with as much mathematics removed as possible, and that's what most tutorials out there look like to me.
I've also seen a decent number of "rust for haskellers" type articles. These are very much the exception though. There should be many more "Xlang for Ylang programmers".
@christianp my stock answer to that question is "not yet, but it will be soon." Unfortunately she's old enough now to not take that as anything but teasing.
Telling time is hard: GitHub is describing November 30, 2022 as "2 years ago".
It's more than one year, yes, but I wouldn't go as far as two. Would "15 months ago" work?
This is like how you have to decide when to stop giving your baby's age in months and start using years.
@christianp I really dislike those ways of describing times. The bounds are always really unclear and it's the middle of the scale where it's really useless, like describing roughly 15 months as 2 years ago.
I'm not even sure why dates began to be shown in that way; it seemed to start in the last 10-15 years (no idea where) and spread out. The core idea seems okay, that you care less about precision the further in the past the time is but the implementations have always seemed to have been poorly tested. Is there one broken library that everyone uses for this or something?
@christianp@j_bertolotti my guess is that they were trying to avoid a "work to be done by someone else" ticket because either there's a massive backlog of those tickets or the process for creating that kind of ticket is horribly convoluted. Or, knowing how universities work, both!
This guess is not in any way informed by working for the UK civil service, not at all...
@christianp an option is to boot from the SD card but have the root partition on another drive. Once the boot is finished, the only reason to touch the SD card is to update the boot options. This is how I (try) to set up my Raspis (that word looks wrong but I refuse to put an apostrophe!)
EDIT: the page full of dubiously out of context graphs ends in bitcoin nonsense. So, if you find yourself actually radicalised by this domain, do a bit of thinking
@christianp I especially like that many of the graphs show approximately exponential growth so there will always be a point where growth appears to accelerate on a graph with a linear scale.
Today's Microsoft Office annoyance: In Excel, when what you're typing in a cell is a prefix of something in another cell, it likes to autocomplete to save you some time.
I've got a column containing both the string "A" and the string "A B".
Every time I type just "A" and press Enter, it autocompletes to "A B". GRRR!
I've just upgraded mathstodon.xyz to Mastodon v4.2.5.
I had some trouble on a couple of our servers, getting an error from libjemalloc "cannot allocate memory in static TLS block".
I couldn't work out how to fix this, except it didn't happen after I moved back to Ruby 3.2.2 by editing the .ruby-version file.
Later on I might look at rebuilding ruby, but if anyone else encounters that error, downgrading ruby is a quick workaround.
And if anyone knows how I can fix it, please tell me!
