[go: up one dir, main page]

I used to be @ambitiousslab@lemmy.ml. I also have the backup account @ambitiousslab@reddthat.com.

  • 19 Posts
  • 66 Comments
Joined 3 months ago
cake
Cake day: January 11th, 2026

help-circle










  • My partner, dad’s partner, and so many colleagues at my job, wasted so many weeks cramming for this stupid, irrelevant test. If you add up all the people who have to take this, how many person hours have we wasted as a society, all to be forgotten anyway, because it’s useless information.

    We really need to get rid of this test, or at a minimum strip it down and make it about how to vote and access public services. But even then, if someone wants to learn that, they will of their own accord and in their own time anyway.




  • I’m quite crazy:

    • Marginalia with the no JavaScript filter, then with that filter turned off
      • Good for finding technical stuff, and sometimes recipes! I often find cool blogs this way.
    • SearxNG via farside.link
      • Unfortunately (and understandably) many of these sites use Anubis now, so I have to turn on JavaScript, and thanks to Google’s ratelimits the results are either fantastic or not helpful at all
      • But, the public instances can work, so I try with 3 instances before moving on

    Depending on the thing I’m searching for, I have search shortcuts set up. These shortcuts are really handy. It seems much easier to get good results on dedicated search engines for each task, than finding another general purpose search engine that’s as good:

    • Wikipedia, Wiktionary and Wikidata
    • Some other wikis
    • Lemmy (of course!)
    • Peertube and podcast indexes
    • Websites of grocery shops near me

    Finally, if all else has failed, I use Google (which still unfortunately happens at least a couple of times per day 🙁). Although, reading the posts now, I should switch this stage to DuckDuckGo instead.

    I’d quite like to set up my own instance of SearxNG + YaCy at some point. It’d be nice to configure SearxNG to basically do all of these steps at once that I’m doing manually, prioritise my YaCy index, but use other engines to fill in the gaps, and then gradually fill in the gaps in my YaCy index.



  • I used Language Transfer and Michel Thomas’ courses when starting to learn Italian and found them really helpful in getting a foothold into the language.

    The Michel Thomas course was longer and went in more depth, but I preferred the vibe of language transfer. The Michel Thomas course seemed to be aimed at people looking to cheat on their wife on a business trip, because a lot of the conversation was about inviting women to get a drink :( Despite that, it was still useful.

    Unlike the language apps, these courses did a good job of getting me to think in real-time. Despite only being able to express and understand basic things, they gave me confidence to try and say things. Even without much vocab, I was able to express myself in a simple way: “I like that red thing over there”, and I was able to pick up new words with “what does this part mean?” or “can you repeat?” etc. So far, it’s the best method I’ve found to bootstrap enough of the language to start talking and picking up the rest by osmosis.



  • He added that the government would provide an extra £63m to the 21 areas affected.

    This issue has trundled on for so long, the concern being cost on local government, and then it turns out it can be sorted out overnight with just £63m.

    If you ran on a pledge to clean up politics, and you’re up against Reform saying “they’re all the same and all as bad as each other”, then it should be obvious that delaying elections because you’re scared about the results is not a good look.







  • I like the idea and also want to support independent journalism, but in the UK context, I don’t think a separate community makes sense. I had a look at !uk_politics@feddit.uk and !unitedkingdom@feddit.uk and I think most (70%+) of the posts from my unscientific sample would count as independent (in the sense of “free from government and corporate interests, and not controlled by a major media conglomerate”).

    I wonder if it would make sense to set up a bot to automatically crosspost articles from allowlisted domains from these general news communities? And if unknown links were found, there could be a mechanism to add them to the allowlist?

    These were the sources I found:

    • The Guardian
    • Al Jazeera
    • Novara Media
    • The Canary
    • Socialist Worker
    • Morning Star
    • Democracy for Sale
    • BBC
    • Politics Joe
    • ITV
    • New York Post
    • Metro
    • The National
    • Associated Press
    • Big Issue
    • London on the Inside

    A lot depends on the definition of independent, and I’m focused on the text rather than perhaps the intent of the definition. If that was stronger, a lot of these could be excluded and a separate community might make more sense.



  • I agree with you, and I think there’s a tension between the technical solution (meeting users where they are) and political solution (persuading the users to come to our way of thinking).

    The technical solution is an unequal fight. We have to provide a familiar and equally good experience - integrating everything into these easy-to-use everything apps, on a shoestring budget compared to the proprietary apps. And, without the “education”, users will converge on particular instances because that’s what’s most convenient, giving a lot of power to particular players in the network.

    If we can persuade people to prioritise freedom over convenience, then we end up with a much more resilient userbase who will go help with the existing networks.

    I don’t know how we can make people care, though. The free software movement has been trying for 40 years to make regular users care, but the message only really lands with developers. There’s certainly more interest in taking down big tech nowadays, but convenience still seems to come first.


  • Searching for a single Discord alternative may be asking the wrong question however. Discord itself is an extensive bundle of functions smashed together: real-time chat, persistent forums and documentation, voice chats, events and even games. Rather than replicating that bundle in a single app, the open social web may be converging on a different model entirely, where specialised services handle specific functions while sharing identity and social connections across protocol boundaries. These individual services themselves do not have to share the same protocol underneath, and may actually work better if they don’t, with each protocol handling the part it is best designed for.

    This is the most interesting part to me. Can users be persuaded to have different expectations from the proprietary apps they’re used to?

    Whenever these sudden migrations happen, the alternatives that win seem to be the ones that look and behave as similarly to the proprietary app as possible, as the people switching don’t care about decentralisation, and are much more sensitive to any changes in experience.

    I think we need to create separate experiences, backed by the same protocol, for people who care about decentralisation and freedom (and discover the fediverse naturally, outside of these big migrations), and those that show up during the big migrations.

    For the first group, we want software that’s easy to self-host, customisable, spreads users between instances, ultimately empowers them to have the exact experience they want. For the second group, we should just copy the exact experience of the proprietary networks as much as the protocol allows.

    Of course, the risk is that we get even larger influxes of people who never had to learn the community norms. Is that worth it? - I’m not sure.