HNHacker News
TopNewBestAskShowJobs

bayesianbot

660 karma · joined June 20, 2019

submissionscomments
bayesianbot··on PowerShell: The object-oriented shell you didn't know you needed
And slowness, which might be related to the size. I was using PowerShell for months and loved it, until one day I needed to start fish shell and it felt amazing, the experience was just so lag-free compared to PowerShell. I didn't notice it while using PowerShell but after that I could never change back. Also shell startup speed felt quite meaningless before but after getting used to few second startup delay with PowerShell, it was so nice to just open terminal and start typing instead.
bayesianbot··on Guy Callendar, the engineer who discovered human-caused global warming
It's also quite interesting window to thoughts of other people. I find it extremely mind-boggling that we live in times of detecting gravity waves, building particle accelerators etc. and some people live in reality where we don't know how much energy we're getting from the sun or co2 from volcanoes.
bayesianbot··on I made a thing to help you make a thing out of a keypad
Huh, weirdly I was just searching for solutions for this just few days ago - I've got some ESP32 projects with keypads and LCDs, which usually are wired little further away from the ESP and other attachments, so the wiring is somewhat PITA. This thing could help it go from 12 wires total for both LCD+keys to only 4 using shared I2C wiring, quite nice!
bayesianbot··on The argument over a long-standing autism intervention
I think a lot of traits or experiences of autistic people are found in many neurotypicals as well, but the levels/intensity actually make it completely different. I definitely see your society box, but it can be quite different experience when the things you're hiding are main part of your being, brain and thoughts, for all your childhood.

I only have my experience - it made me extremely stressed for over a decade. It felt like anything I wanted to be or do was probably wrong in some way that I could never predict, as often things I thought would be perfectly fine turned out to be wrong, often for reasons that I really couldn't understand why or base general rules on. I was basically scared around the clock that somebody catches me behaving in wrong way, and I was always prepared to stop everything and start masking - looking like and doing things my peers expect me to. I went from talkative happy kid to extremely quiet person who probably always looked like I'm ready to soil my pants. I spoke really little for a ~decade, feeling like every time I'd open my mouth I'll say something wrong and it can be used against me. Later on I stayed the nights awake quiet in a dark room and slept when other people were home and awake, because then I could maximize the time being just myself without worrying about who I was expected to be. I pretty much still don't have a good relationship with my parents, they've never known who I am as I just learnt to hide myself. I think they're good parents, they just didn't know what to do with different kind of kid and made some terrible choices in a conservative time and place where having a "non-normal" kid was not acceptable outcome. And I wasn't really good at communicating what I was feeling or the kind of damage it was doing to me.

I know I took it in a really hard way, I knew even then. As autistic person I can have quite intense feelings. Like I said I'm sure there's autistic people who have the opposite opinion and experience. But I know I'm wasn't some special case, I'm quite sure there's a lot of trauma associated with a lot of ABA'd children. I can definitely understand the parent's viewpoint and also that there's wide spectrum of completely different cases, but I really don't think this should be ABA-parents vs autistic kids disagreement, we both would have a better life if we'd have some better therapies for autistic people - I just think the best starting point would be understanding what causes the stress/anxiety/discomfort/meltdowns and trying to fix the underlying issues, not from trying to hide the symptoms like we're praying gay away. The goal should be a child with as good, healthy life as possible, not some expected behavior. Maybe ABA style therapy is good for some and I expressed myself in kinda extreme way - there's some history behind that, that I really haven't fully processed in a healthy way yet - but parents should understand when it's used in the wrong way or wrong cases it can go really bad, it's not something that should be just done because good options are lacking.

bayesianbot··on The argument over a long-standing autism intervention
The argument against ABA is exactly that it looks like it works from the perspective of the (neurotypical) parents, but the kids feel like they're basically tortured until they behave like their parents and society expects them to. To be something they aren't, be forced in to a box where you just don't fit, until they do.

I haven't gone through official ABA therapy but I had extremely ABA-like childhood and I'd have to agree. When I was kid and didn't know what's different about me I used to think at least there's probably millions like me who don't have to go through that. When I found out about my autism and later ABA, I was horrified to know it's official therapy. But I'm sure there's autistic people who'd disagree with me.

bayesianbot··on Important Update on GGPoker Security
Backstory:

GGPoker is currently the biggest online poker network. In January, someone wrote an article in CardPlayer Lifestyle[0] about how their client is quite badly programmed (with Adobe Air) and sends data as clear text with real name information that shouldn't be shared with everyone. It seems GGPoker didn't really take it that seriously, but at least enabled SSL for their connections.

Lately a bunch of players noticed player named "MoneyTaker69" was playing with extremely loose strategy (which usually would be losing big time), but winning at insane 90 big blinds / 100 hands rate (good players might win from few to maybe 10+ big blinds per 100 hands on average). Lots of hands played were really suspect, and definitely anyone playing hands like that should never be getting so lucky that they'd win during thousands of hands. Also the most winning leaderboard is full of people who won at really, really good rate for ~15 000 hands, and then just disappeared - which would be strange to do for any gambler or even a profitable poker player in a great winning streak.

So this leads to this statement by GGPoker. I linked the archive link as currently the real url is leading me to another page, so it might be removed. GGPoker has a feature that when players are all in, their winning probability gets shown, and apparently also some thumbs up / down button that depends on the odds. It seems like player(s) were running a modified client that could request that information at any time in the hand, even when the hands are not shown yet or the players all in. GGPoker also said they returned $29 000, while "MoneyTaker69" won $48 000 in a single tournament, even without counting any of their cash came winnings.

It isn't the first time this has happened - in 2007 Absolute Poker and UltimateBet were found to have "superusers" that knew cards of the other players[1], that time they were insiders in the companies. They were found out in the same way - crazy hands and massive winrates that didn't make any sense, and an absolutely wild call in a tournament that the superuser "POTRIPPER" went on to win.

I thought this might be interesting for HN - mostly for how the biggest Poker Network has handled their security.

[0] https://cardplayerlifestyle.com/ggnetwork-acknowledges-fixes... [1] https://en.wikipedia.org/wiki/Cereus_Poker_Network#2005%E2%8...

bayesianbot··on Gentoo goes Binary
So apparently it would be possible to use mostly binary installations but compile some libraries, interpreters and cpu hogging apps with -march=native - and that wouldn't cause any problems with ABI or anything?

Would it be hard to set up a build server on another machine? What if that other machine was running Arch or some other system?

I remember when running Gentoo ~20 years ago, it was quite often that package compilation failed and I had to go fix some package settings or something, don't even remember well except how disappointing it was seeing a failed compilation again after waiting a long time - is this still common occurrence?

edit: at least the build server doesn't look too bad and the running other arch could be solved with a WM / container.. Feeling really like wanting to give it a try, still remember the good times I had with Gentoo way back (even with the compilation problems)

bayesianbot··on Obsidian 1.5 Desktop (Public)
You can use git with it. It automatically commits at configurable intervals, and with few hooks[0] you can make pushing automatic and also pull changes made elsewhere (which then get instantly shown on a running Logseq desktop instance).

The default git configuration was kinda weird, but I think I initialized the git myself and then added it in Logseq before adding the hooks and it's been good experience.

[0] https://github.com/CharlesChiuGit/Logseq-Git-Sync-101

bayesianbot··on A new approach to domain ranking
There's details in the first linked See Also -post: https://www.marginalia.nu/log/69-creepy-website-similarity/
bayesianbot··on LibreChat – Enhanced ChatGPT Clone
I have been wondering why I haven't really found good self-hosted UIs that could also use plugins / functions. Just installed this (which was quite easy) and it seems to work very well! Definitely gonna replace the one I was using before.
bayesianbot··on Ask HN: Why do half of Internet users think we are living in a simulation?
I think it's fun thought to play with from technical perspective. How would you make an ethical simulation with limited processing power in 1000 years?

- Dark energy would make it less compute-intensive towards the end of the simulation. If you knew where your observers are you could simulate less and less over time, while also reaching an ethical endpoint for the simulation

- and with Double-slit experiment, we kinda know that universe knows when things are observed

- Cosmic Microwave Background would hide a lot of signals further away, giving a possibility of aggregating or dropping signals further away from observers

- Diffraction Limit would limit the resolution of observations we can make from further away, limiting the resolution of the needed simulation.

- Quantum uncertainty principle would be easy way to make the simulation non-deterministic by just adding some jitter / variance.

I might have written some things that are wrong. Also it doesn't really answer your questions as this doesn't really lead me to believe we surely live in a simulation, it's just something I like to think about sometimes.

bayesianbot··on Americans' trust in scientists, positive views of science continue to decline
Doesn't really mean they should proudly talk about scientific subjects like they really know their stuff when they mostly don't know the things you'd learn from first chapters of your first science book, does it?

It's not like this is only related to Covid. You can see people "doing their own research" on climate change as well, without knowing the basics of science, physics, systems science or actual climate science, but they've watched a podcast or read a blog. Covid just made this bro science stuff so mainstream that it's hard to have a beer in a bar without hearing about it.

bayesianbot··on Americans' trust in scientists, positive views of science continue to decline
My opinion on extended lockdowns or anything else Covid-related really doesn't have much to do with my opinion that when forming strong opinions on complex subjects, especially ones that differ from consensus of experts, you should have good understanding of the basics of the subject as well as what are the experts basing their opinions on and why. Especially in science where the whole basis of the whole thing is evaluating how well your model performs predicting real world, throwing that all out and feeling things out really isn't as great method as many people seem to think.
bayesianbot··on Americans' trust in scientists, positive views of science continue to decline
After covid my trust in future of humanity has plummeted. It feels like every time I see people they're talking about "science" where they've actually thrown the science part out and are having a lot of feelings how things are. Apparently everything is impossible to know, even all the things that are extremely well studied and really simple, and you can replace all the modern science by reading some tweets, listening to bro science podcasts and feeling them out, and always ending up the result you were looking for anyways. It's hard to understand how people's understanding of complex subjects can be on so ridiculous level but still be thinking they know everything they need to know. Maybe my view is biased but it's always the same story everywhere I go, though I've stopped going out as I find it so frustrating and depressing. I even meet people who have been "studying science" for years but don't really know how scientific method works and obviously don't even know what scientific paper looks like.
bayesianbot··on Privacy is priceless, but Signal is expensive
https://archive.is/k90dC
bayesianbot··on Ask HN: What are you passionate about at the moment?
It's just hundreds of variables of data and some calculated features - basically from current point in time and some configurable points in history + some aggregated history data, fed to ML.NET FastTree / FastForest. I think LSTM / some Python frameworks could have been better, but learning C# better was one of the main reasons so ML.NET algorithms it is. I get quite reasonable predictions, though I need to do some more to get better at predicting the end of an activity because of some limitations how I'm giving the data to avoid overfitting. GPT is just going to be last step of the process, being fed todays actions and we'll see if I can get it to blurt out anything interesting based on it.

I just checked my manual tags in the training database and I've been on this for 197 hours and 37 minutes since since 12th of October. No wonder I feel like I need a break..

bayesianbot··on Ask HN: What are you passionate about at the moment?
I've been collecting data from my phone and some different "sensors" (food delivery orders, bank payments, Activity Watch statistics, I'm adding a video camera soon. For phone about every sensor I have in my phone like ambient light, wifi APs, GPS, rotation, steps, when it was last unlocked, what app is in use, what app / artist is playing media etc.) and I'm using ML to predict activities I'm currently doing. I guess the end goal is to automate my diary writing to GPT as a joke..

Started as a small project on wanting to learn C# better and being interested about how much sensors collect information about us and how we could analyze that, but it really got out of hand. I've been working on it for a ~month (with some breaks), sometimes all day, built a lot of tooling (like webassembly & android app to tag current activities for ML training data) and analysis things for it and am nearing a burnout point :P but been fun and I've definitely learned a lot about C# and ML.

bayesianbot··on Show HN: I made a ChatGPT UI that looks like Slack
I've used about $70 of GPT4 API this month for coding help and chatting, and tried a ton of different self-hosted UIs which I feel like all are lacking and usually end development quite soon after beginning. That being said, this looks really great and more like something I've been looking for. I'll give it a try, but it has to be really good to overcome the negative of not having a possibility for self-hosting.

I was little lost there for a moment, I didn't add a project but went straight into settings, which brought up an empty page (with navigation though) with apparently nowhere to go. Only after logging in again and adding a project made it work, so it was little confusing and maybe it should steer you to right direction if you end up on that empty settings page.

bayesianbot··on How to find a lost phone in a no-cell-coverage camping site?
I use this sometimes when I can't find my daughter in a large shop - just turn on my wifi tethering and when I have a device connected, I know she's nearby.
bayesianbot··on Marcel the Shell
Have to agree with the slowness, I just used powershell for a month or two and thought it's great.. Until one day I accidentally started fish and it felt so blazingly fast that I couldn't go back to pwsh anymore..
bayesianbot··on Build your own BitTorrent
I wonder if you could insert your own downloader to pacman / whatever without touching the source.. I might try to do this at some point if I have time. I often get over 100MBps from torrents but only like 5MBps from package repositories (though on good days that might be up to 40MBps)
bayesianbot··on RealFill: Image completion using diffusion models
I got really good compression using this technique with JPEG XL, I'm sure there's even a good reason why it works so well but it's been a long time and I don't seem to remember why.
bayesianbot··on First Impressions with GPT-4V(ision)
IIRC it wasn't that impressive in the end as instead of recognizing the places the AI apparently learnt to recognize subtle differences in street view cameras used in different locations? I might be wrong / thinking of the wrong model l and I'm on mobile without my browsing history so hard to check, but I think it was putting a lot of weight on some pixels that are noisy
bayesianbot··on Webb Discovers Methane, Carbon Dioxide in Atmosphere of K2-18B
404 not found, did they maybe remove the article for some reason?
bayesianbot··on 9/11 in Realtime
I just love these real time real life retrospectives, yesterday I thought I'm not going to spend time watching these again but here I am.. Any other good ones around? I've watched the Estonia[0] a few times (easier as a Finn), I think I've seen one about the 2004 Tsunami one but can't seem to find it (or might remember wrong), and Apollo 13[1] as well.

edit: On second thought I don't think the tsunami one might make that much sense, probably was just a collection of videos and news captures.

[0] https://www.youtube.com/watch?v=V5tbah19qo8 [1] https://apolloinrealtime.org/13/

bayesianbot··on UPI Payments: 10B transactions a month done, next stop 100B
Thailand has a similar system - you use your banking app to scan a QR code, either the code is generated and has the amount included or it's printed static paper and you write the amount yourself, then show the receipt to the merchant or they check the notification in their phone. Started in 2017, nowadays it's everywhere, I can go to local market and pay $0.5 products with it. My 12 year old daughter opened her own account as well and didn't need my help, only installed an app and went to ATM that has a special slot for ID card to verify identity. Quite convenient as no need to carry cash anymore.
bayesianbot··on Ask HN: What is your policy regarding smartphones for your children?
I was quite similar, asking just programming books for birthday/christmas presents since I was 10, but my parents took a different route. I'm neurodivergent so I behaved different from other people, and coming from small town conservative family they just wanted me to be "normal" and blamed the computer on my differences. For me it was crazy as it was obvious I'm different way before the computer and I just couldn't understand why they wouldn't want a kid that's spending his free time studying programming (and science). The constant stress of somebody coming to my room and whining that I'm sitting on computer (and behaving weirdly) made concentrating on programming really hard and my life really anxious and stressful. Some point at my teens I noticed that going to drinking with my friends was way more relaxing, even though I hated drinking and just wanted to study. The anxiety+stress+drinking really left a mess in my life.

Now I have my own kid and I'm finding it hard to put limits on phone usage as her peers are on tiktok etc. all the time. But the most important thing I'm trying to do is have these discussions and then that's it, try to act on what we've agreed, but not arbitrarily whine about too much usage or anything else that would make her constantly stress about it. And try to guide her to spend the screen time in better ways, like spending time playing Factorio / Minecraft multiplayer with her so we get creative time together instead of her looking short videos.

bayesianbot··on Some flying robot behavior improvements
You can set the logistics request maximum amount of the item to same value as minimum amount - though then if you want to pick up some more you have to adjust so the bots won't come take them away, but for me it's much preferable to inventory being filled from tiny extra stacks.
bayesianbot··on Record low sea ice led to catastrophic breeding failure of emperor penguins
That's in 2022, doesn't sound good as 2023 Antarctic sea ice extent is lowering old records by a huge margin so far.

https://earthobservatory.nasa.gov/images/151692/exceptionall...

bayesianbot··on Antonmedv/walk: Terminal file manager
+1 for broot[1], it works so fast and can be used even to fuzzy search sub-sub-directories. I'm also bit lost on the features but the help is quite good and mostly I use it just to find a directory and enter it, which requires only one keyboard shortcut.

[1] https://github.com/Canop/broot

← PreviousPage 4 of 7Next →