The other two comments are partially correct, but it’s not just layer time.

The plastic we print shrinks when it cools. Solid layers shrink all the way to the center, pulling in on the outer walls. Thin walls cool and shrink, but without the extra pull from plastic next to it shrinking, so it shrinks less.

Reducing layer time helps a bit, but you actually need to change your model to fix this. A chamfer or fillet on the inside lower edge will help a bit, especially if it’s wider horizontally.

Even better if you can include a small section with 0% infill to decouple the internal shrinkage from the outer wall. I don’t really know how to do this, myself—I think it needs to be accounted for in the model, but I suppose it might be possible with modifiers on a region of the model? (Like, maybe no infill on a cylinder that maps to the inside of the box, so only the chamfer in the bottom edge has no infill?)

Is your startup still in the VC stage, or are you looking to IPO soon?

Meh. I like seeing it. I don’t spend a lot of time thinking/worrying about it, but it’s a nice data point to keep up with.

As with most “computer geeks” forced to use Teams/M365, I have a strong reason to want more people comfortable switching to open source. The faster we can ditch American property slopware, the better.

CrealityPrint resets the prime tower every time I change anything. I’ve had so so many plates of prints with a model or two fused with the prime tower because I forget to move it.

Forgetting to reattach the PTFE tube and feeding endless filament into open air.

Printing directly onto the magnet, with no plate installed—weirdly enough, the print came out okay-ish, with only the bottom layer not great.

Using metal grease instead of thermal grease when installing a nozzle—pretty sure that’s what eventually lead to my hotend failing.

Not setting purge volumes correctly, and ruining full plates with massive colour bleed.

Thinking I can skip the prime tower, and that it will save filament (it doesn’t), since it purges anyway. Then not even noticing how badly filament bubbles without priming.

Not realizing that the printer purges multiple poops, so seeing so many clean colour poops means I’m purging too much. (I wasn’t.)

Not noticing that a 2-colour print was alternating colours in the infill, and that’s why the colours were always muddy, and that increasing purge volumes was just wasting more filament.

Never calibrating a filament, or even changing settings to match manufacturer settings, leading to printing fast PLA way too cold, and turning to spaghetti from poor interlayer/bedplate adhesion.

Not removing a colour-change-by-layer when scaling a model, leading to an entire plate with the wrong colours. (The top layer on a flat print, of course…)

Selecting the wrong filament colours when sending a print.

… I could go on. At length.

They say you learn the most from failure, so I’m pretty sure that makes me an expert.

What’s doubly ironic (and evil, for that matter) is that FDroid apps are so much safer than Google Play apps. Open Source apps verified and compiled from source by a trusted repository is the gold standard.

Meanwhile, malware stealing all your personal info, apps lists, and contacts are rife on the Google Play store. I avoid it as much as I can.

Supports are not a one size fits all.

Exactly.

This isn’t the perfect tool for every model, but there are definitely situations where a support fin would be perfect, and they’re hard to get with most (all) slicers?

Right, but nobody goes to these sites independently, it’s all to plaster so much spam that they get search traffic.

Try searching for any niche topic in the form of a specific question, and you’ll get search results that are a barrage of slop.

The article addresses that. These are not numbers released by the company; they’re from independent analytics companies’ estimates.

an Uber spokesperson said, "We strongly disagree with this decision and disproportionate fine, which we will appeal. The AP examined historic policies that were discontinued years ago. …

So, basically, “we broke the law 4 years ago. That’s like forever ago, so you can’t fine us.”

If that’s their legal defense, LMFAO.

Seems fine, for what it offers. It’s only 1 layer that slows down that dramatically, then it should be back to regular printing speeds.

You wouldn’t want to do lots of huge overhangs, still, but this method will definitely have some really cool applications.

Just FYI, since nobody responded, the steps to check:

1. Search if the phone model can have the bootloader unlocked. If no, then you’re likely unable to do anything. Note that in some cases, this can vary by region. (Ex. the US version could have a locked bootloader, but the international model can be unlocked). The bootloader is basically the software that loads your phone operating system, and it needs to be unlocked to allow the OS to be changed.
2. LineageOS is probably the easiest to check for a reputable replacement OS, for non-Pixel phones (for which you can get GrapheneOS.)
3. XDA Forums used to be the #1 spot for phone “ROMs” (alternative OSes); it’s changed a bit, afaik, but I think they’re likely the next easiest spot to look for a custom OS, if GrapheneOS and LineageOS isn’t an option.

A few Linux phones work, but there are only a handful of models supported. If you don’t have a Fairphone 5 or a phone specifically marketed/sold as a Linux phone, you’re most likely out of luck, unless you’re willing to make major compromises, like non-working cell connectivity, camera, etc.

A more balanced requirement would be replaceable storage or removable storage support.

If you can install your own (bigger) drive, then that’s a fair alternative, imho. Like a 64GB–>1TB Steam Deck.

And USB is reaching 10-40 GBps transfer speeds, so swapping cards doesn’t offer the same speed savings as it used to.

Most (all?) Canadian carriers shut down/are shutting down 2G and 3G, so older phones don’t work at all.

It also means that locations that have spotty 4G connectivity can have big dead zones now, for non-5G-enabled devices, that used to be covered by 3G bands.

Even some 5G phones are affected. Like, I have an international model phone that has most but not all Rogers 4G/5G bands that used to fall back to 3G, but now is just worse.

Really? I was just watching some of their newest videos with my son and we thought they were great. I haven’t noticed a drop in quality.

That quote I strongly agree with, but I think you completely misunderstand what the professor means.

A huge problem in education right now is students using AI inappropriately, in many different ways. As an extension of the “digital literacy skills” education has been focusing on teaching in the last ~1-2 decades, educators are now teaching “AI literacy” skills, because students are using it wrong all the time. And it’s unacceptable both academically and professionally.

AI literacy skills are mostly about teaching students how bad AI is, what it can’t do, and why they should not rely on it. Like, understanding how an LLM works means that students will understand that they’re just fancy word predicting machines, not “intelligent” in any way. Students who understand this are much less likely to trust LLM output.

AI literacy skills are not teaching students how to vibe code faster.

Yeah, fair. I was more thinking in terms of a neurodivergent special interest, but desired to “hobby” to be more inclusive of neurotypicals.

Eh… I think it’s fine if some people want to subscribe to a car. Some people really like cars, and want to have the newest model. They can lease a car, then return it (lightly) used for someone who is price conscious to get a cheaper car with warranty protection.

I buy old, used, cheap cars because I don’t give a flying fuck about what I drive. The cheapest I can get my $/km, while having a reliable vehicle, the better. Sub $10K CAD direct buyer sales are great, for me.

I think it’s a waste of money, but I’ve spent $thousands on my hobbies, so who am I to judge someone else for their hobby spending, assuming they can afford it.

Even without a tire over them, they weigh more than the heaviest tire on the market. So, yeah. Exactly.

This was my thought, too, especially with how stable Linux traffic is relative to other OSes.

It might just mean that more people are web scraping, and that web scraping is easier/cheaper on Linux.

The Steam numbers are more interesting to me, as a measure of consumer usage, since they should mostly represent people actually using their computer.

PornHUB puts out annual usage stats, too, but idk if they include OS. (When I used Reddit, their stats would come up on dataisbeautiful frequently enough to come to mind years later—they had some interesting stats, like average visit duration and that the most Republican states were the most likely to search for gay and incest porn, but I digress.). Would bots make up a high volume of traffic for them? I imagine not, since there’s little economic value in scraping it (to train LLMs, in particular — I think it would poison the training data rather dramatically!)

ikr? My video card died last week. RIP my wallet.