You don’t believe how often people leave secrets in the app or use webviews and iframes badly, misconfigure OAuth in client side and so on. There are many issues where secure API does not help.

Not the end, just google bleeding itself out as a search engine while LLMs take over. Chatgpt often links me random, actually decent sites that I wouldn't have gotten from a simple google search on first page, because it would've been exclusively seo slop

ChapGPT (and all other hosted L To me) still use a search engine behind the scenes whenever you ask a question. Often, it uses those seo slop sites as primary sources when formulating an answer. Which is why, even though hallucinations are now somewhat rare these days, you _still_ should independently verify their output for anything important.

How many random parasites has modern medicine removed from you? Likely you never even had one. It's not just medicine, we also have abundant, safe food, clean water and warm housing

>we also have abundant, safe food, clean water and warm housing

for now

Yes! I think it’s really easy for us to take for granted the long list of incredible things we take for granted. Every time I hear someone lament how horrible the world is today, I wonder when they’d rather have been born. Especially as a random human like this.

How would you count?

"you don't have to, they'll tell you."

(+1)

Then they deserve you taking a photo with a second phone

Well, I did, after the fact, but it's only because I had a work phone on me (that doesn't yet actively block sharing to non-work devices).

I doubt most people have a second phone on hand, ready and able to capture actual screen shots when your phone is preventing screenhots.

There might be a market for a USB-C camera to take phone screen photos to immediately upload back into the phone.

Yeah, few moments ago it hit me there could be a market for a phone case with in-built camera(s) that photograph your screen at a shallow angle on demand, and transform that into a screenshot automatically.

However, my idea happens to hit a very new limitation: for some reason, "privacy filters" are suddenly a thing. New phones have them built-in as some new magic display function, and for everyone else, there's this mad marketing push to get people to buy and use privacy foils in place of glass/foil screen protectors.

IDK what's up with those. Since when "shoulder surfing" with smartphones is a real problem, and why do marketers and product designers bet there's so much latent demand for it?

Recently handled someone's phone who had it. Neat technology. But the use case for it doesn't make sense, for the reason you specified.

The only thing I could think of is if you want to browse TikTok or Instagram in public and don't want people seeing the weird shit that's on your algorithm, but even then, that's kinda weird.

That'd be either analog computing or you're implementing floating point number math in each neuron. Maybe some kind of PWM calculations would be doable, do we have it researched or would be a new kind of math?

That center pin looks fragile, especially if it's printed along Z axis. How about printing teeth that'd mesh with the belt?

Printed teeth are weaker than the belt's own teeth.

Conversely, with this design the center pin could break loose and the belt would still hold, because the pin just has to stop the loop from fitting through the slot.

Timing belts are crazy strong. From my FRC work the amount of abuse and weight they will deal with is handy.

Belts will burn through polycarb if it isn't tracking properly. They are strong enough to deform any pulley mount on a tight fit.

Only minor complaint is that the teeth will slip if you don't have a ton of them engaged (or they are excessively loose) but otherwise they are amazingly useful.

For FRC 2010, we built four custom gearboxes using timing belts. It was a pod with two wheels driven by a CIM motor to get a crazy amount of torque to get over the humps in the middle of the field. At any time, whether on flat ground or when tackling the humps, we would always have four wheels touching the surface. The most forward and rearward wheels would only touch the inclined . We couldn't go through the tunnels because our endgame mechanic was a big hook on top that would pivot out, latch onto a vertical pole, and pivot the bot sideways using a winch. It made for an amazing photo since few teams at that first regional event were even attempting the endgame.

That’s exactly it… if you tolerance the mouth reasonably tight, the centre pin doesn’t actually carry much, if any, of the load. Friction in the mouth carries a bunch of it, especially if you add the zip-tie clamp from my sibling post.

The pin doesn't even need to be captive, if you tug on the belt it'll jam in the slot kind of like a climbing nut in a crack.

I agree that it looks fragile, but my experience with a similar kind of design in practice has shown that it’s not actually as big of an issue as it seems it would be.

A few other things that can help:

- for the “it can work itself out” problem, I’ve had pretty good luck on other projects by making a standardized “dual dovetail” (bowtie) male/female slot with clearance tuned so that the male piece fits tight into the slot. Printing the male part flat on the bed means that the filament is oriented correctly for strength

- integrated zip tie slots. One design I worked on that was similar to what OP did but needed better grip… I made the slot where the folded over timing belt meshes longer, removed the printed centre pin, and put two integrated zip-tie slots: one near where the centre pin was, so I can use the zip-tie ratchet to actually pull the belt backwards and put tension on it; the other was near the mouth of the jaws, so that after tensioning it I could use a second zip-tie to “lock” the jaws onto the belt

- a third one: you make a little block with dovetails that looks like the “circle plus mouth” piece from this post, but also include a pocket for a nylok that you can use to add tension to the belt after getting the belt secured inside the clamp.

I have tried printing teeth that would mesh with the belt but didn’t get particularly good results from that. Possible that newer printers have better resolution and could print the teeth better, but my test with that just ended up with a little bit of a rougher surface and the teeth couldn’t actually mesh.

I have printed GT2 teeth on a prusa mk3s+ that mesh perfectly

Awesome! I’ll have to try it again. Last time I tried was on some Wanhao i3 clone like a decade ago and it was not great, but the printer itself was not great all around other than the price :).

It looks like the pin doesn't take much load. It looks like the pin holds the teeth open, while the load is on the shoulders of the entryway.

If you printed teeth, you'd have to make sure the teeth in the top loop are precisely aligned in relation to the teeth in the bottom loop based on the length of the belt, right? That seems like it could be limiting

You don't need a loop, just a toothed slot.

A small bolt could easily replace the pin, this would also allow a washer to be added preventing any chance that it working loose.

You dont actually need the pin

> wearing the hard drive

Do reads wear out nvme? Iirc only writes do

only writes and honestly that’s a bit overblown they last a long time

How about being truthful? Say "I think A, but I'm not really an expert, so also try asking your favorite LLM"

Maybe if it's a business level tool made in house with claude vs contracted to a software house in India. Never for a compiler, who even is that person?

> Never for a compiler, who even is that person?

Why do you care?

It's a stranger from the Internet. Same as with any other software you're using. In particular, same as with any OSS tool you're using.

> > Never for a compiler, who even is that person?

> Why do you care?

> It's a stranger from the Internet. Same as with any other software you're using. In particular, same as with any OSS tool you're using.

Don't you think that even if you don't know someone personally, you can calibrate your trust in their ability based on reputation, track record.

Yes. But we usually start with a high default, and only recalibrate if there's a specific reason in a given situation (e.g. there's been a specific user or sets of users involved in spreading malware, and software in question could've been targeted).

> you can calibrate your trust in their ability based on reputation, track record

rarely. How many of the libraries you pull from package managers do you actually know the authors of?