• 0 Posts
  • 60 Comments
Joined 2 years ago
cake
Cake day: June 30th, 2023

help-circle
  • Except using developer mode means you trade away support. Why pay all that money for a bambu if you’re not going to get support? Might as well build your own printer at that point if you’re going to have to problem solve all the issues yourself anyway

    Also “This is beta testing, not a forced update”

    Beta implies that at some point this will no longer be beta and will a mandatory update.

    They’re testing the waters, thankfully the pushback may have them reconsidering. It probably just has them reconsidering the rollout/timing though. they may do something like Philips did with hue: announce the cloud integration into a product that did not require an internet connection or cloud integration for over a decade, get a bunch of backlash, then not implement the cloud part while the heat dies off while still fully intending to do so. The hue cloud was announced as a mandatory change in September of 2023 and still hasn’t been implemented but there is a reminder with each app update reminding you it’s eventually going to be necessary if you have not done it yet.





  • Mario games were always kind of tech demos for Nintendo innovation and the innovation is kind of boring now

    In the 90s it was graphical fidelity and gameplay mechanics, smb1-3 on the nes, smw on snes, m64 were serious contenders for “oh my god this shit is crazy and plays like a dream”.

    Sunshine is around where the graphical fidelity started to lag but the physics were still nice.

    Then they started to innovate in other ways and galaxy brought in motion controls which were well done (arguably) and the (by then) distantly lagging graphics were almost a feature

    But odyssey showed that doesn’t work anymore. Everyone has motion controls now, even phones, and they’re often just annoying. The switches innovation isn’t something that really applies directly to gameplay itself, so ultimately odyssey stays just a standard game, with solid graphics thanks to great design but noticeable slowdown in several areas because it’s running on an outdated tegra tablet, and the same old platformer gameplay because there’s nothing new to do with it.

    They did have some cool level design that helped it feel fresh and interesting at times but it ultimately felt like it had similar problems to sunshine: not a bad game, but not an experience the way mario 64, mario world, or mario 3 was. Maybe I’m just too old


  • Saw this from the moment they did the rfid nonsense, doubled down on my beliefs when they started burning cash on advertising like crazy. Tons of youtubers and such shill this shit

    They had a series b round in 2023 of an undisclosed amount with several chinese vc companies and they’ve had investments in 2021 and 2022 as well. I don’t know how chinese vc works but I assume it’s similar to american vc where there is a strong demand from the backer(s) to monetize in this fashion

    Why do you think reddit went to shit? Series b in 2014. Those people drop serious cash. Reddits seed round in 2005 at y combinator was for 100k. That’s serious money to you and me, but to vc people that’s not worth getting out of bed. Reddits 2014 series b was fifty million. They suddenly had a gigantic influx of cash to grow infrastructure and compete with the big dogs like meta and twitter. They were fairly successful with this. They then raised 1.2 billion over 4 rounds from 2017-2021. That’s why they had a relatively quick turn to shit; that money was to try to make the site bland and profitable in preparation for ipo. It worked out because the stock made investors a ton of cash at the expense of making the site dogshit

    Bambu will have a similar trajectory. Investors will give them an amount of money that is frankly obscene, they will use that money to develop (and probably to party, ridiculous salaries and/or fluff jobs, and have really fancy offices), then they will actively make the product worse. 5 years from now they will have used that money to entrench themselves in the market space. Don’t be surprised if the average person thinks “bambu” when they think “3d printer” because they pissed away 10s of millions on advertising. But their printers will have more consumer hostile bullshit (finally fully locking out 3rd part filament instead of just requiring you to do a pain in the ass respooling seems inevitable) like this and it would not be surprising to see the build quality suffer too.



  • Yeah absolutely, and any bugs that are found are permanently gonna be there, stuff like that. It sucks. Thus the “I probably wouldn’t bother either”. But if you’re really passionate about it I say go for it, just cover your tracks really well, only work with people you trust implicitly, and don’t popularize the project until you’re ready to be done with it. The moment it gets hyped and picked up by ign/kotaku/etc you’re done.

    I think you can still get the word out to get other modders though. These things don’t seem to get squashed at the planning stage, otherwise publishers would be spending tons of resources going after people who have maybe done no actual work. It seems like what happened with this mod, when they make socials and release ongoing updates, screenshots, videos, etc, of their work, that’s what gets people hyped and all the attention (and eventual hammer) on them

    Of course this means you get no credit for your work, which sucks. If you’re undertaking such a massive effort and getting 0 credit I could understand why you wouldn’t bother (although it’s kind of badass to do it for no clout)






  • Most people won’t close their facebook account. if it ever does happen it will be because the accounts are purged (highly unlikely given facebooks raging hardon for data) or that the site undergoes a massive transformation after losing a ton of value and rebranding ala myspace.

    That likely wouldn’t happen for many many years until a solid competitor arose that grew enough to overtake them and that’s real challenging given their size. It was one thing when user counts were in the millions or even tens or hundreds of millions but facebook has several billion users. That’s like a sizable chunk of the entire population of earth.

    It would take a very novel approach to overcome those numbers and then you also have to consider momentum: at this point there are a great deal of people who consider facebook “the internet”. Like they open their browser/phone and that’s what they do. It is their habit. Then in second place you have instagram so even if you knocked facebook off meta would still ultimately be ok. And with the incubation period of social media they’d probably have another one up and coming long before your threat became viable that would have the benefit of starting in like 7th place simply because of their massive market share (see: threads). By the time your social network had the 10ish years it takes to get to hundreds of millions of users they’d potentially have that one at a billion, or have pulled the plug and moved on to another trial with a massive head start

    Doesn’t mean to give up on the fediverse stuff, just that the gross corporate social media likely wont go away for a very long time, if ever. Barring outside influence like regulatory change of course



  • The audio improvements are more impressive than the video improvements but they’re niche

    8k is a boring upgrade. We’ve officially reached the point where pixel density is like “well that looks better but only a little”. 1080p was dramatic, 4k was solid, 8k is eh

    The audio improvements are really for people who have external sound processing and want ultra low latency. This is cool but again only impacts really niche use scenarios. They use the example of AVR and headphones. I will use the example of gaming with an avr and games that rely on on low latency sound like bemani/rhythm games.

    I have an avr and playing rhythm games on it requires calibration every time, and often changing settings about muting player sounds so I don’t disorient myself with sounds that are triggered late by my own playing. The game I’m currently into shows an audio delay of just over .1ms with the calibration tool, mostly imperceptible for media watching and playing easy songs but a nightmare on hard difficulties.

    Certainly not worth upgrading my tv, avr, and cabling though



  • Moving to a rack is nice, I love my rack. If you’re in or near a city I suggest keeping an eye on Craigslist and ebay (search by distance nearest and lowball ones that have been sitting for months) because it’s not uncommon for nice racks to go real cheap as long as you come get them. I got my rack realllll cheap ($40, 42u, fully enclosed with massive pdu) because it’s a 90s ibm rack and it’s welded steel so it’s like 450lbs. Moving it was a nightmare but it’s real sturdy and I’m never moving it again now that it’s in my basement

    For my goals in the short term I have to replace a sas cable that caused a crc error on one drive, it only happened once per smart data but still want to get that done asap. I also have another drive that’s beginning to show some smart issues; it’s on the same sas cable so it may be related because the errors didn’t increase (they all were related to an unclean shutdown, confusing things) but it’s old anyway so better safe than sorry I guess.

    Medium term I want to finally upgrade my ups. The one I have now is not a rack mount which is part of what led to the unclean shutdown. It’s also a bit undersized. I have a generator for my house so I don’t need something massive but the one I have is 450va and several years old so with the tired battery I only can get about 5m of runtime. It’s more than enough to cover the transfer from power cutting out to generator power but I want something that’s a bit more reliable in case of generator failure. This is pricey though because my array is pretty huge so it’ll probably be held off unless I find a good deal on a dead one that has cheap batteries available

    I also want to put the rack on its own circuit. This is something I should do asap because it’s cheap, just gotta find time and rearrange my panel a bit because it’s pretty full. This would be the other part of the unclean shutdown as the outlet would be in a much better location and I could also install a locking outlet

    Would also be nice to pick up a super cheap monitor locally, like something for $15-20 from a pawn shop or Craigslist or something for the rack. Earlier this year I had nginx crash on my server and the webui became inaccessible, I had to drag my nice and kind of large desktop monitor down to the basement to solve the issue, would be nice to just have a shitty small monitor on the rack for that

    Speaking of nginx I keep meaning to setup some kind of reverse proxy or mdns for all my dockers so that I can just do whatever.whatever instead ipaddress:3993 which makes my password managers barf but I’ll probably just be lazy and edit my hosts file

    Longer term I want to add a secondary low power server that can run something like pfsense to handle my routing, then turn my current wireless routers into access points because they kind of suck as routers.

    And of course the array could always be bigger, especially if drive prices fall

    I will probably realistically only do the drive and cable replacement, the circuit thing since that’ll be like $40 and a half hour of work, the monitor if I can find one, and maybe the hosts file thing. If I run into cash (unlikely) or a crazy deal (you never know) the ups would be my next priority but there’s a million other things going in life (deductibles just reset for health insurance, hooray)




  • Theoretically possible but someone would have to reverse engineer the protocol and write a new service that works based on that

    There are some projects but all that I’ve seen are open source head units that either: are terrible and aren’t android auto/carplay (eg they only have the most basic of integration, can show notifications and play music via Bluetooth) or they emulate android auto/carplay without the necessary licensing/hardware (so you still ultimately rely on google/apple)