Best Of

Thank you to @JosiasJessop for an excellent, if depressing, thread header.

Iran's targeting of non combatant Gulf states in retaliation further emphasises the precarious nature of global stability.
What was the world's policeman since WWII has now become a semi-rogue state itself.

It's not at all clear how we climb back out of this.

The same way we climbed out of it last time - bouncing off the bottom of the abyss. If we fight hard for our principles and have a large slice of luck.

Otherwise, we don't. Human society regresses to a more violent and less free state.

It's not looking great, is it? It seems like after every major conflict promises are made to abide by jointly drafted rules and to eschew violence as far as possible. To never forget. Then after a few decades, we do indeed forget and it all kicks off again.

I'm not looking forward to China being the global hegemon, and enforcing its style of global order, because the US destroyed itself by following the deranged whims of a corrupt narcissist.

The major risk is that the forces of democracy and law don't win the next major conflict. And then we're really in the shit.

ChatGPT obviousy still doesn't care about factual information, because they're still giving inaccurate election results. Just tried it again with a question about votes cast at the 1983 election in a random constituency.

I was at the Current Archaeoogy Conference in London a couple of weeks ago. Loads of really interesting talks. One was on notable women in Archaeology, of which there have been plenty over the last century.

To start the talk the two presenters said that they had asked Chat GPT to tell them the 3 most notable men in British Archaeology since 1900 and give a resume of their careers. With a few minor errors the answers provided were broadly correct although of course there is always room for debate in these things.

They then asked Chat GPT to do the same for the 3 most notable women over the same period. The answers duly came back along with resumes, notable achievements, positions held and papers published.

The problem was that none of these women actually existed. These were detailed answers including citations for 3 women who are no more real than Miss Marple.

Of course the lack of immediate women for Chat GPT to seek out and use is part of the problem (though there are dozens of very notable female archaeologists down the years) but it also served to illustrate the problem with LLMs. They are trained to provide answers and when they can't they just make shit up to fulfill their task.

Talking about uncessary wasteful pay outs...The Hundred...Sunrisers Leeds

Zak Crawley, £180,000

Some of us enjoy the hundred.

For example £240,000 for Joe Root to Welsh Fire is better value than £75000 for Mandelson to Epstein Fury.

Most traditional cricket fans don't like the hundred, but then it was never for them. I'd find it more palatable if they added 20 balls per innings and dropped the stupid PC game graphics...

I used to spend many a summer's day at New Road.

120 balls? It wouldn't be the hundred then.

I don't see what benefits the Hundred offers that the T20 Blast - which is a wonderful competition - does not.

(Apart from the added prominence it has given to women's cricket).

Does anyone know how to preserve an old website? I had one twenty years ago that still exists, although not by typing the address anymore, and I’d like to keep the content for posterity

QUESTION FROM ME TO PERPLEXITY. AI

• Can you give me a list of sites that enable you to archive your site, like "archive.is"
  

ANSWER FROM PERPLEXITY.AI, THEN CHECKED BY ME
Here are well-known services that let you save or view archived copies of web pages, similar to archive.is:

• Archive.is (archive.is and others) - Quick and dirty method of archiving a single page
  
• Wayback Machine (archive.org or https://web.archive.org/) – The largest public web archive; you can both browse historical snapshots and use “Save Page Now” to archive a URL on demand.
  
• ​Google Web Cache (obsolete from 2024) – Enabled you to view a recent cached copy of a page via Google’s search results or cache: operator. Removed in 2024. The "https://cachedview.com/" site *may* replicate the functionality
  
• ​CachedView (cachedview.nl) – Front-end that lets you quickly see Google, Bing, and other cached versions of a page.
  
• Conifer (obsolete from June 2026) (conifer.rhizome.org) – Free (with limits) archiving for interactive and JavaScript-heavy pages, useful for preserving dynamic sites and web apps.
  
• ​ArchiveBox (archivebox.io) – Self‑hosted tool that lets you run your own personal “archive.is”-style service and store snapshots you control.
  
• Perma.cc (https://perma.cc/) – Creates permanent, citation-style snapshots, aimed at academics, lawyers, and institutions; limited free tier, stronger focus on long‑term reference.
  
• Memento Time Travel (obsolete from 2025) (was timetravel.mementoweb.org, now https://mementoweb.org/) – Was a Meta-service that queried multiple archives (like the Wayback Machine and others) for past versions of a URL. Shut down 2025.
  
• Pagefreezer (https://www.pagefreezer.com/) – Commercial archiving services aimed at compliance, legal, and enterprise use; they automate ongoing captures of specified sites.
  
• Stillio (https://www.stillio.com/) - As pagefreezer
  
• MirrorWeb (https://www.mirrorweb.com/) - As pagefreezer
  

ADDITION BY ME

• I'd go with archive.org if you want to preserve your website a page at a time. Preserving its functionality is a bit more difficult. How many pages are we talking about?