Once upon a time, straight outta school, I wrote up solutions to problems from the Serway physics book. For Chegg. $5 per problem.

Fuck Chegg.

And they are all welcome back if they can satisfy the Linux Foundation that they’re not affiliated with a sanctioned entity on the SDN list.

Compare with the yearly release cycle on cars.

Like any other convention, it’s not really a big deal either way. Fortran gets along just fine with 1-indexing.

802.11ax, clients just… (essentially) wait for a random amount of time, listen for a break in the signal, and take a leap of faith.

Ethernet originally worked the same way, back when it competed directly against token ring. Ethernet won by being as reliable in real world scenarios while being cheaper to build out. Gigabit Ethernet was the first standard that insisted on full duplex only.

Half duplex mode with the collision avoidance is still actively supported for 10/100, but it is becoming very hard to find an unswitched hub. So you may have to write up your own twisted pair cables.

Really I worked a project once that just had post-its stuck to the wall. It worked as well as Jira does.

Why not just have the app dynamically generate the static with random numbers every time. There is no video file of white noise, and bonus the bumper intro is never exactly the same twice.

Here’s my guess. I don’t know anything about this particular device, but I have worked with medical devices.

A powered exo-skeleton sounds like it might be a class II medical device. Being a medical device, the OEM was required to produce a safety risk analysis per ISO 14971 in the EU and 21 CFR 820 in the US. I don’t know what all was listed, but probably one of the safety risks was thermal runaway from the (assumed) lithium ion batteries.

Lithium ion battery packs have a well known problem with occasionally overheating and catching fire. This famously delayed the launch of the 787 Dreamliner. This is also why you can’t put your phone or laptop battery into your checked luggage.

In the original risk analysis, there will be a number of mitigation steps identified for each hazard. For the lithium thermal runway, these probably include a mix of temperature monitoring, overheat shutdown, and passive design features in the battery pack itself to try to keep the impacts of over temperature and fire away from the patient.

So how does the price get to 100k? It could be some kind of unique design features that are now out of production and the original tooling is not available. The 100k cost is probably something like to redesign the production tooling, particularly if you have to remake injection molds.

You can’t just use any off the shelf battery pack, because that would invalidate the risk analysis. You’d need to redo the risk analysis, repeat at least some amount of validation testing, and possibly resubmit an application to the FDA.

TLDR: you can get some MEs and EEs together to solve this problem, but once they’re on the case, you can blow through 100k real fast.

Medical devices are required to comply with 21 CFR 820 in the United States, which establishes quality management standards. This includes minimum standards for the software development lifecycle, including software verification and validation testing.

In the EU, broadly equivalent standards include ISO 13485 and IEC 62304.

If an OEM wants to do a software update, they at minimum need to perform and document a change impact analysis, verification testing, and regression testing. Bigger changes can involve a new FDA submission process.

If you go around hacking new software features into your medical device, you are almost certainly not doing all of that stuff. That doesn’t mean that your software changes are low quality–maybe, maybe not. But it would be completely unfair to hold your device to the standard that the FDA holds them to–that medical devices in the United States are safe and effective treatments for diseases.

This may be okay if you want to hack your own CPAP (usually a class II device) and never sell it to someone else. But I think we all need to acknowledge that there are some serious risks here.

If wasn’t full garbage collection in the spec. It was some infrastructure support in the spec that would make it easier to write garbage collectors in C++.

I thought they catch fire and burn down slowly.

Correct. Both the recent pager and radio attacks, and the 1996 cell phone attack, were performed by planting military explosives inside the devices in advance.

There is no magical way to hack the electronics to make a lithium battery straight up explode.

Also, Israel already assassinated someone by exploding their cell phone way back in 1996.

Technically I think that’s still “put us first on the search bar” money. You’re giving the real under-the-table explanation.

criminal lawsuit?

What? Do they mean “criminal prosecution?”

Well it probably wasn’t a Vic Mackey-style rubber hose attack, because it sounds like this chump is getting hauled into court.

I’m thinking of highly niche industrial and embedded products who are likely to be left behind.

A major traditional selling point for Windows has always been the backwards compatibility.

Yeah. This sounds a lot like some PM type thinks they’re gonna get rid of control panel, and they just don’t know what all is actually in there.

And not to mention the custom control panel applets hanging around out there from who-knows-what vendors.

But the ads on search are the big revenue driver for Google overall. Presumably those stay with the Google Search subunit, and they would have plenty of cash to do whatever?

But if they’ve only been found to monopolize search, how does that remedy the search monopoly? Presumably the new separate Google Search company would still have a search monopoly.

There’s also D. You could just upgrade to D.