• 1 Post
  • 200 Comments
Joined 1 year ago
cake
Cake day: June 30th, 2023

help-circle


  • What kind of tv? For webos it’s potentially a bit complicated but also potentially stupid easy depending on which version of webos your tv has

    https://www.webosbrew.org/rooting/

    I would strongly suggest avoiding nvm even if it’s supported unless you’re very comfortable with hardware hacks. The others are all software and fairly easy to do if you’re capable with following instructions. The most recent, dejavuln, is fairly simple but can be a bit finicky (you may have to try a bunch of times) but lg is also rolling out patches for it so if your tv is updated you may be out of luck. It’s hard to say because the patches aren’t rolled out unilaterally. Webos is a bit confusing and there are many “branches” that all have similar features but wildly different numbering. If your tv is patched block updates by either disconnecting from the internet or blocking the above sites in your router and watch the webos homebrew discord (linked on that site). There are people actively researching new exploits and if one pops up it’ll be discussed in the discord first (and if it’s a big deal, like they expect it to be patched, they usually ping everyone to let them know to do it asap)



  • In some less common scenarios you may need the ftdi or ch34x drivers. Support for these is usually already in the kernel but if not the ftdi comes with the arduino sdk and the ch34x drivers are available

    You would generally only need this for older or super cheap printers, and even then you would generally only need this if you need to reflash the firmware for some reason or refuse to print from an sd/usb. but don’t refuse to do that, even back in the days before wireless printers it was a dumb idea to print via usb



  • that’s why the internet fucking sucks now. Everyone’s too afraid to make their own site and lazily relies on the conglomerates of social media, which has reduced the internet into like 5 websites that repost each others content

    Also seeing isn’t converted to income, especially on instagram. The more hardcore comic fans may want to see your work in full, may want to follow the story if there’s a narrative to your comic, see your arts evolution, etc. and they’re probably the ones that are far more likely to drop cash on merch for a series they enjoy. That’s why you combine the approaches, post comics on instagram or whatever to get the word out, and have a site so your hardcore fans can easily browse your work (with the added bonus of letting people who don’t fuck with social media also see it)

    Again, or don’t, I don’t care. Post everything to facebook and twitter, make the internet just 2 websites instead of 5, refuse to have control of the primary platform your work is shared on, whatever


  • If they don’t want their comics browsed that’s their choice of course but it seems pretty silly. What if instagram deleted/bans their account for nonsense? Goodbye audience and archive of everything you’ve ever done

    Used to be you got a free webhosting account and posted comics to a gallery on a shitty handmade webpage until you built up an audience or gave up. If you got bigger you’d move to a better site with a custom domain and new readers could catch up if they were interested. Achewood, gunshow, dinosaur comics, questionable content, xkcd, penny arcade, nedroid, etc all started about this way and many of them continue to this day. Use social media for promotion, not for archival

    My guess is they don’t want to bother with people who aren’t willing to fuck with facebook, twitter, pixiv, etc. or they don’t know how to make a free website. Whatever, just means they lose the audience of people who refuse to use facebooks bullshit


  • Hosting is cheap and there is free hosting available if you don’t care about having a custom domain and have a limited audience. If your audience is big enough to go past that bandwidth you can probably monetize somehow and cover server costs easily, even if it’s just selling a few pieces of merch. My website is $80 a year with a custom domain and I get unlimited transfer/bandwidth. It’s shared hosting so over about 1k visitors per day means it’ll get slow but if you’re getting that kind of traffic you can probably sell more merch and get a vps

    Use twitter/pixiv/ig for promotion but if you don’t do the above your locking out anyone who refuses to make account. All of them won’t let you look at more than 1-5 images before locking you out entirely with account nag screens that can’t be bypassed. Or just stay on those platforms, I don’t care, I’ll just never read your comic






  • Just fyi for anyone who would care about this: while hue bulbs are built well they are moving towards a model that requires you to put them on “the cloud”, even though they were sold for years and years without that requirement. The update will be mandatory whether you want it or not as part of Philips security being integrated into the app. It’s unclear what will happen if you don’t create an account and sign in at that point

    So if you’re like me and put all your iot shit on an isolated vlan without internet access they may not be the best option for you. Or if you just don’t want to support a company that wildly changes the tos years after purchasing their (expensive) product. I don’t want my home shit on the internet, I don’t trust Philips to put enough cash or effort into securing their servers, etc.

    The bulbs do work with zigbee though and that seems to be a viable alternative to using their hub/app although I haven’t tested it fully. This also means if you’re using them via HomeKit you’ll need some kind of bridge like home assistant




  • Clicker training is just paired stimulus to provide an easier and cheaper mode of reinforcement, training a behavior is separate

    like you do the clicker training by associating a noise (or whatever) consistently with a positive stimulus. what stimuli you use to prepare depends on the learner as different stimuli have different potency depending on the learners preferences. eg you can say a blanket “I’ll use food” for your dog and for some dogs this is fine. mine certainly seems to be food indiscriminate with no serious preference and very few refused items. But even with that I still need to take care not to use the non preferred (he spits out lettuce and celery. otherwise literally anything gets him salivating)

    but then to change a behavior you’re still relying on operant conditioning which would be something like upon exhibiting the desired behavior provide access to reinforcement consistently and then fade it out as the behavior strengthens. Operant conditioning is much more complex than this of course but this is a pretty standard jumping off point.

    That said there are pros and cons to clicker training humans. This is something that is practiced and even has some evidence behind it. The clicker solves a lot of potential reinforcement issues: it’s far easier to deliver immediately (which matters a lot). But I worry about the potency loss translating a strong reinforcer to a clicker. If you pair it with a food you really love or something it will potentially be effective but never as effective as the food itself.

    This is still potentially worthwhile as food reinforcers are often problematic (increasing caloric intake, often food reinforcers for people aren’t healthy options, promoting unhealthy eating habits) and reducing it to a click eliminates those issues. But if the behavior you’re trying to create is particularly difficult or aversive the reduction may mean the potency is no longer high enough to motivate.

    Often this can be countered by making the behavior less complex and working up to it (eg instead of learning a complex task in its entirety breaking it down into more manageable chunks). In practice this may look like just initiating the task at first and providing reinforcement, then as comfort increases raising the bar for reinforcement. Eg I need to keep my room clean but I hate cleaning so to start out I provide reinforcement for just picking up one item/small area. But then when I do that consistently I raise the bar and now I have to pick up 2 areas. Etc. or you could approach as a tolerance thing, I start by cleaning for only 3 minutes and reinforcing, then 5, 7, 10, etc. numbers are arbitrary and depend on the learner. Approach depends on the learner too, the toleration approach makes more sense for most people but if you do a bad job cleaning and need to develop the skill of cleaning thoroughly the first can make more sense. Then reinforcement is not time based but quality based, Eg did you clean the area sufficiently even if it took you 8 minutes. Drawbacks and positives for every approach

    And of course there’s the issue of delivering your own reinforcement. If you control access what’s to stop you from just taking the thing even though the behavior wasn’t exhibited. These strategies typically work better with external control of r+, but some people do have the self discipline to do it alone.

    There’s a LOT more to conditioning and reinforcement but I’m getting bored of this lmao. Also you may notice I didn’t describe anything about punishment. That is intentional because it is generally at a much higher risk of creating adverse effects and some studies suggest it is not nearly as effective as reinforcement based strategies wrt general population (and some specialized populations)



  • They’re following the model of the tech industry, which makes sense because there’s a lot of crossover there

    I fixed an acer laptop yesterday. It was a gaming one, like a $700 laptop. Wouldn’t turn on. Acer said the motherboard had to be replaced. When I got it I found a blown capacitor shorting the main power rail, replaced it, and it works fine now. A part that costs like 3 cents in bulk. Repair was roughly 45 minutes including diagnosis.

    For this one a motherboard swap isn’t the end of the world but the additional point is that for many of modern laptops and for all phones this results in a superior repair. This laptop in particular had removable nvme storage but tons of laptops have the ssd soldered directly to the motherboard so swapping the motherboard means you lose all your data. No one ever has backups lmao

    But acer, apple, Lenovo, hp, etc all do this. It’s much easier to train their techs to just do board swaps, it’s much more lucrative to make repairs a several hundred dollar endeavor instead of the pennys it would cost to replace passives or basic ics, etc. they then send the “junk” boards off to the manufacturing depot in sea to actually get fixed and then sell them again as refurbished