• 16 Posts
  • 90 Comments
Joined duela urte bat
cake
Cake day: uzt. 18, 2021

help-circle
rss

Thanks for the suggestion. Do you know where I could find the Linux kernel developer’s blog post you’re referring to?


Are you talking about having good search engines in things like private communities in Lemmy or chat rooms in Matrix and things like that?


Maybe changing your title’s “extract” to “deduce” or something like that will avoid this confusion, so that it’s more likely that you get the answer that you’re looking for :)


In general, I watch the very last couple of games because that’s usually when lots of people around me are excited about it and watching it.

With this year’s one, I’m so grossed out by its ecological and social insanity that I don’t really want to.


Thanks, Lemmy, for being a place for educated, empathetic, and principled people
I had an awful night. I was surrounded by incredibly racist, sexist, ignorant, and insensitive people. I was astounded by the lack of basic knowledge about the world that was possible. I knew conceptually that ignorant people with relative privilege existed, but seeing it in action is another thing. It just hit me in the gut to hear slurs flying around, and hearing unironical defenses for fascism, apartheid, and social Darwinism. This got me thinking. I have to be honest. I sometimes fear Lemmy could be a place for violence and dogma, but after tonight, after being surrounded by absolute pieces of shit, I have to recognize this place is much more special than I gave it credit for. Here in Lemmy, I consistently see posts that reflect a sound understanding of how the world works, be it ecology, politics and economics, programming, heck, even basic sciences. I also see humane concern for the wellbeing of others, regardless of who they are*. *Well, to be fair, maybe I do see a bit less concern for those who are most destroying the livelihoods of others. But even they are sometimes seen as victims of a system. The point is I am grateful for you. I love that you aim to grow your knowledge about the world. I love that you're empathetic and kind*. I love that this place is inclusive and fosters growth among us.
fedilink

Unless I’m missing something, I believe both WebP and AVIF have lossless compression as options. Of course, choosing that results on heavier files, but at least there’s the option.


Thanks for the recommendation. I didn’t know such little reduction in quality meant so much savings with JPGs and PNGs!

Yeah. Avif gave me some nasty artifacts on some pictures that I wanted to save long term. Not using it anytime soon… or at least until those issues are fixed.

Why do you dislike WebP? I can see that it isn’t widely used, so if longevity is my goal JPGs and PNGs are a better bet. But am I missing something about WebP?


The Snowden leaks show that the government programs rely on unencrypted data and data that companies are forced to give. Neither method relies on exploits of memory unsafety.

Memory safety refers to the software that is used internally.

Here’s an analogy. Imagine I sit in a park, and have a squad of friends who go out to listen to other people’s conversations. They may write down and pass me what they hear or they may simply tell me what they remember. That doesn’t matter. What matters is what I do afterwards.

Let’s say I want to write down what they said. I could write it down with a permanent marker on the palm of my hand. If I make a mistake, I’ll have to cross it and waste precious space.

Imagine I learn I could use pencil and paper, or a computer. Mistakes would stop costing me dearly. I would also have more space.

This change from ‘permanent marker on a hand’ to ‘paper and pencil’ or ‘computer’ is the equivalent to the NSA recommending memory safety. This doesn’t change the fact that my squad is out there spying on people. It just changes how I deal with the evidence they hand me over.




I’m sure it’s funny but I genuinely don’t get it(?)


Search for “loss meme”


Ironically, he may be a great source of knowledge for social scientists to determine how to fix a social problem.


Good point. I understand you’re saying it’s better to use Ubuntu than more locked-down and less libre alternatives. Fortunately, I daily-drive a GNU/Linux distro that I believe is less invasive than other alternatives.



I don’t get it.


Echo chambers are responsible for polarization if we define them broadly as groups that orient themselves towards similar values by listening to each other and not to other groups. This happens with friend groups over dinner, family reunions, and (for those who are religious) at religious events.

However, if we define it as necessarily something that happens in social media, I disagree with the notion that those echo chambers are responsible for radicalization. Notice I said “radicalization” rather than ‘polarization’.

Before, you’d simply gather around with your friends or family and discuss how the world is becoming more progressive. If your family happens to have regressive values, you’d complain about marginalized groups no longer being as marginalized. “Oh, the gays, they want to destroy the country by normalizing homosexuality”. The impulse to retain the values of a marginalized group (such as religious and racist zealots) motivates further cloistering and clustering.

Now, social media amplifies polemical voices that fiddle with your insecurities. If you’re a shy teenage girl, this might mean that you’re exposed to messages about how you’re ugly and poor. But if you’re not, it can be about anything. An effective technique is showing you your political enemy doing something threatening. “Democrats steal the election”, " Republicans burn a library because of LGBT books". This makes you want to engage with the content and watch more ads from the social media platform (Facebook, Twitter, YouTube). As a result, you spend more time reading about or watching videos about this threat (the insecurity-fiddling threat) and radicalize. These are the flat-earth YouTube rabbit holes. Or the toxic-masculinity red pill ideology rabbit holes. Or the fucked up terrorist rabbit holes. Or the hyper-religious and anti-democratic zealot rabbit holes. Or the racist rabbit holes. There are plenty. You can tell me which ones I’m missing. You probably know what I am talking about.

In other words, it’s not echo chambers that worry me. It is algorithmic manipulation to maximize engagement. That is a source of big trouble for the world. That is what we will have to deal with if we, as a world, pretend to live democratically.


If you’re not too anxious, reading Learned Optimism by Martin Seligman and applying it. If you’re too anxious, reading The Happiness Trap and applying it. Those are both books on proven ways to deal with many mental health issues.


Oh wow, the browser-OS comparison seems big! Do you have a favorite browser? Or are they all equally bad? I hope browser security and size can change for the better in the future!




I agree that Electron is a bottleneck. That’s something they’ll have to deal with. However, there’s a small difference between what this post is saying and what you’re saying: “24x - 1400x fafster depending on what you’re measuring” is quite different to “[the changes are] not really helpful”. Regardless, I’m glad to see that they’re embracing Rust in some way, distancing themselves from javascript. I hope the SDK will be good enough for a bustling ecosystem of efficient clients (and servers!). I’m also glad that people like you and I are making it clear that we value the reduction of sludge, we value efficiency and therefore inclusion for everyone, including those who have devices that aren’t that powerful.


Let’s evaluate your analogy’s internal logic. Take this website. If we average all the reported speeds of slugs, we end up with 0.004925727 m/s.

That is, we add up all the speeds 0.013+0.000023+0.0028+0.013+0.0024+0.00086+0.013+0.0024+0.0018+0.0034+0.0015 = 0.054183 and divide them by 11: 0.004925727.

Now multiply that by 1000: 4.925 m/s, which is 17.73 km/h. That’s double the speed at which I run, but it’s slow compared to car or train speeds.

The upper bound is 13 m/s (0.013 m/s * 1000) or 47 km/h, a respectable increase. That’s as fast as the speed limit in my neighborhood.

The lower bound is a meager 0.023 m/s. That’s 0.082 km/h, a pathetic result.

Your analogy’s internal logic is valid.



I have a smartphone. My friend has a fliphone. How to easily encrypt-decrypt our SMSs?
I am aware this is a meta-data disaster on par with Whatsapp. Such is life.
fedilink



Are Lemmy posts crawl-able by search engines?
Discussions here are often very interesting, and at times incredibly helpful. If I had no clue about Lemmy, but I searched online for a topic that happened to be discussed in Lemmy, will that discussion appear in the search engine? As a related question, do you think the discussion example would show up in the search results in the most informative way? I mean in an search engine optimization-kind of way.
fedilink

Federated, free, and/or open source software will not be chosen over proprietary software unless they overcome this hurdle.
# Innovation requires resources Innovation happens wherever we put our efforts towards. The space race and the technologies it gave us is an example. Finding ways of reducing the cost of production of commodities is another. The green transition is another example. These are the bounds that determine how innovation happens in federated social networks. They will only innovate when there is enough effort put towards them. Beyond technical innovation (perhaps in a TOR kind-of-network way, or in a Git kind-of-version-control way), a full-fledged piece of software that is effective and attractive enough for people to use, takes resources. These resources largely refer to labor power. Remember, we're not talking about maintenance costs, but development costs. Programmers require money to survive. # FLOSS is no exception Yes, FLOSS software can survive with volunteers, but even those volunteers have to pay their bills. Yes, the internet plus (as Bruce Schneier defines it), copyleft licenses, and already-existing technical know-how reduce the cost of production, but the costs are there nonetheless. Someone needs to do the work. I think about open source projects I admire. Diaspora received exceptional funding, relative to its goals. Signal was heavily funded by donations. Element has a business model that lets them work towards improving Matrix. Linux has many companies that depend on it and are able to finance innovation on it. This argument, applied to Lemmy, makes me wonder. How do Nutomic and Dessalines handle it? Are they precious exceptions that drive Lemmy forward because of their personal values and their willingness to use their technical know-how for the development of this platform? Will Lemmy thrive without them? This brings me to another point: FLOSS and federated software has the advantage that many people are willing to spend their time working towards laudable goals. They are not motivated by profit as much as improving humanity. # FLOSS has a problem that others systematically address: making things attractive The issue with FLOSS projects is that they mainly respond to thought-out worldviews. That is, people are willing to engage with this kind of software because of logical reasons (*logos*, as Aristotle and those guys would call it). It is rarer to have people engage with FLOSS and federated software because of emotions (*pathos*). Unfortunately for humans, we are emotional creatures. I get put off by how unattractive the Free Software Foundation's website is, despite loving the values that the foundation stands for. I get put off by using terminals, despite the fact that plenty of FLOSS software does not have GUIs. I hate Thunderbird's calendar, despite using it daily. Companies with investors and FLOSS projects with enough funding know this and therefore pay graphical designers, user experience experts, and sometimes market researchers to make products attractive. This takes money. # Conclusion And that brings us back to my point: for free software to be *chosen* by most people, it has to have enough labor behind it to make it both effective and attractive. This is the hurdle it needs to be overcome. ## Notes on my sources These are reflections that arose after a series of conversations with a friend who works at an 'innovation office'. His job is explicitly to design an 'innovation ecosystem', which attempts to create innovation with minimal investment. Everyone at that office knows this is bullshit. Innovation rarely comes without money. Therefore, they basically look for investors for projects that don't have enough money. That's it. This view, that innovation requires investment, is shared by Anwar Shaikh and classical economics. However, it's more complicated with the research behind innovation. Let's take 'platforms of innovation'. For example, cosmopolitan cities, the internet, and universities are hubs of innovation. However, it's a mistake to think that these are 'neutral' in terms of costs. All of them require operational costs. All of them imply costs of technical training. Even here, there are costs that cannot be ignored. If we look at mission-oriented innovation, the situation becomes clearer. Universities doing cancer-related research, States doing green energy-related research, or companies doing market-related research all clearly align with the argument I made above. Finally, it's perfectly possible that the argument I made above is not at all original. In fact, I doubt it is. If anything, it could be similar to a high-school student discovering their own proof of the pythagorean theorem: it's not new for the community of knowers who already know it, but it's new from the point of view of the student. At least I get to share this with you and hear your thoughts about it. Oh, and given that Lemmy doesn't have terms of service yet, I wanted to make sure I could share this in the future. I licensed it under a [Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License](https://creativecommons.org/licenses/by-nc-sa/4.0/). Weird. I know. Oh well. At least you get to share it without fear!
fedilink

The [release post](https://matrix.org/blog/2022/06/16/matrix-v-1-3-release) specifies the spec changes. I'm not entirely sure what the entirety of the change regarding metadata reduction is, but this MSC seems relevant: - Deprecate the sender_key and device_id on m.megolm.v1.aes-sha2 events, and the sender_key on m.room_key_request to-device messages, as per MSC3700. (#1101)
fedilink

IMAP email setup in Android: Better to use phone-wide accounts for OAuth or to use one-time passwords?
What considerations should privacy-minded people take into account to make this decision? For context, I'm using FairEmail because K-9 doesn't seem to be able to *move to a desired folder* multiple emails at a time. K-9 doesn't use OAuth, so I don't have a choice to make there. However, FairEmail *does* use OAuth. And, when reading about OAuth, it apparently is safer than the alternative. This alternative is either using the main account (with no 2FA) or using an app-specific password (with 2FA activated). Hearing this, it would be a no-brainer for me to choose the OAuth, but the issue is that Google only lets you do OAuth if the app is downloaded from the Google Play Store and if the account is set up as a phone-wide account. Ouch. And yet I wonder if the security of OAuth is so much greater that I should forget about the alternative.
fedilink

How is China's internet regarding net neutrality, privacy, and anonimity? Are some websites slowed down? Can I browse with Tor, chat with E2EE, and use other measures to achieve privacy and anonimity?
Given that the NSA and that surveillance capitalism is everywhere, and given that I really dislike that, would it make sense to seek what I like about the internet (the potential for openess, anonimity, and privacy) in China without worrying about technical or political pushback of any kind?
fedilink

Linus might have crashed the Framework website with his Framework modular laptop review
Edit: The Linus forum agrees: https://linustechtips.com/topic/1359778-a-completely-upgradeable-laptop/ I'm trying to configure a laptop right now and I simply can't! The server isn't managing!
fedilink


Buying a laptop. M1? Framework? Something else?
Hey. My brother will buy his very first laptop soon. He was saving for a MacBook Pro, but hearing me go on about Apple being PRISM-compliant and about how open source software is awesome, he's open to new options. His main argument to buy an M1 is that there is currently no chip nearly as good (in terms of energetic efficiency). And I see that he has a point there. However, I was also kinda hoping he'd use his savings for a Framework laptop running Linux. Regarding those computers, my biggest hope is that they'll eventually run good RISC-V chips, chips that can be easily be changed with a simple module change. But that may be a *long* time from now, maybe decades. Another option I thought about was him buying the M1 and fighting his way to install a Linux distro that supports all the M1 MacBook hardware. He'll have a really fast and efficient chip, as well as a good system! But the main objection for this is that the M1 is not really future proof... like, it is *guaranteed* that in the next two years the much better M2 will be put into the MacBook Pro. That improvement isn't trivial; it'll be a 20% reduction in transistor size. But apart from quick changes, it's possible that the novelty of the M1 is problematic. For example, I was reading about a vulnerability in the M1s because of not having adopted a particular instruction set in the very basic operations of the chip. It's almost as if this M1 is an early-adoption technology, if that makes sense. Anyway, those are the considerations that I have about my brother's computer... hopefully we'll have more clarity by the time his classes begin. Do you have anything that could help us achieve that clarity? Or even muddle the waters a bit more in an interesting way 🙃? Edit: Thanks for all the comments! They spurred lots of discussion and some changes of hearts! So, I was really looking forward to getting a Linux-first machine, but two things happened. One was that there were few options (due to the chip shortage probably?): System76 Pangolin not available, TUXEDO quite expensive (and only integrated or Nvidia graphics), Slimbook Titan quite expensive, Slimbook X15 without dedicated graphics (or Nvidia I forget which). The other thing that happened was a friend having us consider the possibility of getting a pure-AMD machine. Since AMD has open source drivers (unlike Nvidia), they will probably work with Linux without much of a hassle. He'd also keep having the option of a dual-boot with Windows, to work with non-Linux software (in case he needs that for school). Such computers could be those with the 'AMD advantage' (AMD CPU and GPU), though they're a bit pricey. Yet this is his money and he's very excited about gaming in them! This is the most likely route. So, no longer Apple. I would've liked to support Linux-first machines, but I guess AMD was the winner here?
fedilink

Request: FLOSS, simple, and 'safe' ranked-choice voting
Y'all may make my day. # Here's the situation: I am part of a club in which we publish stories. We don't have enough money to print them all. So we vote. Currently, we can only vote with only a checkbox ☑️ per option using Framasoft's polling. Not optimal because... # I'm interested in ranked voting. - How can I make a dead-simple (from the coder and user perspective) ranked voting mechanism that also avoids them having to log into places or create accounts? We need to account for old, computer illiterate people. - How can I assure them integrity in the sense of their votes being counted once, in the sense of the votes being cast properly, and in the sense of people feeling that the solution is 'safe'? I say safe because I created this community's website pseudonimously and people loved it, but I once stupidly published fragments of someone else's stories and they got properly pissed at the pseudonymous site creator. They were asking questions about the safety of the community, allegedly seeking legal advice, and even doubting the security of the website. This lead me to make the website a static, open source site with checksums, and a password for some encrypted posts. I delivered the password through the community's official email list, which incidentally is one of the ways I've thought of verifying voter identities. - Could the solution piggyback on existing solutions? # Solutions I've considered: - Hardcoding Google Forms. I set up radio buttons to rank the first option (e.g. "Fries - 2". Then, depending on the rank they chose there, I make a new page for the next option with every rank except the one they chose (e.g. "1" and "3" because "2" was chosen already). This effectively creates a 'dynamic' (hardcoded) solution. - Pros: This seems simple. They can claim an identity (e.g. "Partners in Crime", a couple that writes together). - Cons: With more than 3 options, this insane. The amount of pages I need to hardcode with this method is 'n!'! Slower feedback in case there's double submission ("Partners in Crime" vote twice, or someone manipulates the election). - Giving a markdown template of the ballot and have them fill it out and submit through a Matrix room. - Pros: I can verify the ballot and, if incorrectly ranked (e.g. two options were ranked as "2"), have them correct it. They can edit the message. People can see each vote and verify the election. Minimal programming time. - Cons: Non tech-savvy people, in their ignorance, could think I'm demanding them sign up into a sketchy service. I know, this sounds absurd, but this group, while mostly sensible, has some tin-foil hat people. Formatting could be harder than clicking for older people. Anonymity is lost, in case that's a desirable thing (which I'm ambivalent about; ultimately, people could chat with my pseudonym privately). However, **the biggest con** is that, if not enough people sign up for this, the legitimacy of this election would be compromised. Some people may not sign up because it's harder than simply clicking a link and then immediately clicking on options (believe me, stuff like this matters). The last con is that, if someone finds out that in my private life I use Matrix regularly, they could out me as the creator of the site. I don't want to end up getting sued... - Using a proprietary solution like Survey Monkey and Qualtrics. - Pros: Known brands. I can sell the brand as a safety feature for skeptical older folks. Probably has a simple interface. Fast programming. Probably has a way of verifying identities through an email list. - Cons: This is proprietary software. I'd love *not* to use it. They may harvest our data and sell it, as it goes all-too-often. The email list thing on its own could be an issue. Plus, my current 'PR strategy' is to emphasize the openness of FLOSS, and how it's compatible with pseudonimity... - Rather than a static website, set up a database-driven website - Pros: The frictions regarding registration into Matrix would disappear. I can create this exactly as I want, with less redundancy than in the Google Forms hardcoded solution. - Cons: This is out of my skillset. This is not as easily verifiable as a static website with checksums or a Matrix room. - Hold a Jitsi meeting and vote 'face to face' - Pros: Election integrity is probably assured. - Cons: I'd probably be outed as the creator of the website, since I've blogged about how we should use ranked-choice voting. It's also not easier to 'cast a ballot' in a Jitsi meeting. We still need a ballot-collection mechanism. Otherwise the whole thing ends up as a procedural hearing with people dictating their votes... ugh... # The issue with email verification These folks even heavily questioned the fact that I had their emails to begin with (to be sure, this, in the middle of other comments that lauded the website, which is why I'm not giving up on this). The thing is, I got their emails from the publisher, but it isn't clear if the publisher should've done that to begin with. I found out about this because, when people were threatening to sue because of 'exposing private information' (an email), I got a lawyer and was told that this is a gray area; there are some cases that consider emails sensitive information and others that don't. This is why I'd prefer not to get those who are pissed even more pissed by using their emails again. But anyway, avoiding emails is why I thought it would be easier for people to *claim identities*. That is, they submit their vote, but also say "I'm Partners in Crime" or "I'm Mike", and at first I trust them. However, if, at any point, someone else claims those same identities, then I ask for email verification *only to them*, on the grounds that there is an issue in their identity verification. This would be especially useful in a public Matrix room, where everyone can verify that, indeed, there was a conflict in the claiming of identities. What makes this method better than the 'emails by default' solution, is that verification is their prerogative. If they don't want to verify it, the vote is not counted. If they want to, they can. And email is not the only way. I'm open to alternatives. Either from them or from you. But so far it's the only way I can think of, other than selfies (like Reddit does it at times...; not a fan of this method). # Why bother This is a beautiful community at its best. We have published loads of amazing literature and changed for the good each other's worldviews. It only helps that now have, as a writing collective, a good reputation. Apart from that, I have basic skills that I'd like to hone and make useful in making this ranked-choice vote happen. Ranked voting would improve our current voting system. And if I can make it with FLOSS, all the better!
fedilink