Re: Have not found a way to set glass on fire.
I believe it was 'lab assistants', not people in the original write up. It's an important distinction to make.
535 publicly visible posts • joined 2 Mar 2007
It is possible to detect resistive faults such as a cable not completely plugged in or a bad connection from FOD (both seen as problems with the new connector). There is no need to give any warnings at the driver level, simply turn the card off (or don't even start it) when a fault is detected.
The first thing you (should) do when troubleshooting a graphics card problem, especially a newly installed one as would be the case for almost of these faults, is check and reseat power connectors.
Of course the hardware to detect those faults isn't present on the 4090 board, so we're back in redesign territory.
It was a problem with some of the more poorly designed powerline adapters in the early days of them. The competent designs use passwords of some sort to configure.
Now of course it's all push-button-to-sync stuff and while it's theoretically possible that you and your neighbor are doing your network setup at the same time and you end up cross connected, that really is quite unlikely.
As a veterinary assistant, I've had quite a few long haired cats that required anesthesia to trim because they wouldn't allow the owner or trim salon owner to do so while they were awake.
We'd often remove multiple cats worth of hair from the cat. At one point we even jokingly placed the pile of removed cat hair in one of the kennels when the owner came to pick them up, which did briefly fool the owner until the cat had enough of our shenanigans and loudly voiced its disapproval from the lower kennel.
As a cat owner myself, I regularly clean the filters on my PC that help prevent cat hair from getting to the internals and it has helped massively.
Yes, for safety and ease of maintenance reasons. Where we take the efficiency losses for granted. Also, shavers, toothbrushes and the like use a lot less power than a smartphone and those chargers are typically designed to specifically fit the one device they're charging, allowing for much tighter magnetic coupling than a generic inductive charging device and thus better efficiency.
Yes and no. There's probably a lot of Linux hosts without OMI running, but more importantly, in its default configuration the OMI host is not exposed to the internet at large. Only to machines within the same (virtual) network.
So for most people, this gaping hole is covered up by the fact that you can't easily get to it from the internet. It still needs patching, but it's not as absolutely disastrous as it would be if the default configuration was open to the internet.
That of course also means it's difficult for researchers to tell just how many vulnerable installations exist. Most of those installations will be invisible to the researchers.
We have a nightly compile job that starts at 21:00 and would sometimes terminate abruptly at midnight because that is the time the server administrators had chosen to reboot all of our machines.
This caused several days of lost work for 30+ employees over a period of weeks before our department manager asked the CFO if he could bill other departments for time spent on their behalf. The CFO agreed this was fine and our department manager dropped a bill the size of that server admin's team entirely monthly budget on their desk.
The next week our build servers were set to reboot after midnight, 5 minutes after the scheduled build task was completed and our problems were resolved.
Turns out policies that can't be changed can suddenly be changed awfully quickly when the people responsible have to start justifying massive bills from other departments for loss of work.
Mythbusters tested this with a speeding camera in the US and managed to beat it at 245 mph, considerably faster than this airplane. In their earlier testing they got up to ~160mph which is was not fast enough to beat the camera, so for their second take they brought out a jet engine powered car.
There are quite a few solar collectors that work for FAR more than 12 hours a day. Some even manage 24/7 operation.
Yes, those collectors are up in space, which means powering your home with them is a little more difficult.
There are designs for massive solar collector farms in various orbits, but they often have issues with power transmission to the ground. Laser or Microwave transmission to the ground is relatively lossy, but more efficient than solar panels inside the atmosphere would be, but people are for some reason worried about GigaWatt lasers or microwaves being used as weapons or causing disasters.
T-4 was meh. T-4B was an improvement. T-5 was under construction in 2019, but I see nothing more recent.
MIT and Commonwealth Fusion Systems have also recently announced they're building a prototype compact fusion reactor and that they were aiming for 270 MegaWatts of net power by 2025.
They started in 2017 and the most recent news from that is late 2020 with construction underway.
There are ways to extinguish a burning battery with powder or foam extinguishing agents. Unfortunately most fire departments don't carry said chemicals around with them because prior to the proliferation of electrical vehicles, they simply didn't need them.
Any battery fire they did encounter they simply extinguished by dousing the (small) battery with very large amounts of water. Much easier to do of course when it's a phone or laptop battery.
Our fire departments here are actually starting to carry the necessary hardware now, so they'll knock the fire down with water, then spray foams/powders over the burnt out wreck to keep it from reigniting immediately so they arrange a container to immerse the battery in water long enough to make it safe.
My (manual transmission) car won't even turn over if you don't have your foot on the brake pedal and the clutch. It's an idiot protection feature, as in it prevents idiots from trying to move the car with the starter motor or accidentally throwing themselves off a cliff/into the car in front while trying to start the car. I'd have expected that kind of feature to be standard in automatics where it's even more important to hold the brake.
We have an admin-credentials required set of tests. Since we don't have admin credentials on our testing VM for reasons, we need to make a ticket for IT to start such a program.
So we made an agent instead. IT starts the agent, the agent starts our jobs whenever we need them. This of course completely circumvents the controls IT put in place.
The problem is that the agent would regularly get killed or fail to start on a reboot of the VM, so we'd have to bother IT.
After a particular week with daily outages on our remotes (which would take hours to fix and render us unable to work) our team leader decided ended up writing around 200 manhours onto a timecode for IT related overhead.
We now have admin credentials for our software testing VMs and no longer require IT intervention.
Edit: The restrictions on admin credentials where put in place because of certain people in marketing/sales/HR. Everyone in QA and development still has local admin rights
IIRC both sensors were present on every plane delivered. But without the optional AOA element, the AOA display wasn't part of the instruments on the display and because of that, the AOA disagree warning wouldn't show up.
In other words, the hardware was there all along. The software to display this warning was there all along. It just didn't work unless you paid for an optional flight instrument.
If they had included the AOA disagree warning on every plane, we may have never found out how flawed MCAS really was, or at least it would have taken a lot longer.
In a true end to end encrypted system, provided the meta data itself isn't considered overly sensitive, you can use third parties to reduce the bandwidth requirements and still be sure that this party can't read your information.
A, B, C and D decide on an encryption key together, this is the really hard part because you need to be certain only those four parties have the key.
Then each of them can connect to server X and send their encrypted video and audio streams there. X will forward those streams to the other parties in the chat, so upload speed requirements are the same no matter how many users there are.
X doesn't have your encryption key, so they're sending this data back and forth blindly, they can't tell what's in it.
If the meta data is important and you don't want anybody to know who is taking to who, things get a lot harder.
The reason most of the open source solutions have those huge bandwidth requirements for multi person conversations is because they don't have a server X to bear that cost for you, you have to send the stream to every person separately.
We routinely make other testers angry when our automated tests run. Usually when we're doing single machine test runs to verify our automated tests are good before putting them in the weekly run over the weekend.
Since the transition to the cloud servers, we've routinely soft locked machines over the weekend and come back to find over half our tests have timed out.
This was a problem with the Radeon 6990s when I got one on release. I went through 3 cards before I had a functioning one. The first one I could swap at the retailer I bought it from, when the second one also turned out to be a dud it had to be shipped back to AMD, took a few weeks before I got a replacement that did actually work.
Not mine thankfully, but a friend of mine had a (legitimate) folding@home setup running on their test rigs overnight, when no overnight tests were scheduled.
This setup happily ran for a few years without issues and with approval from management. Then they decided to go cloudy for these test rigs and despite warnings from the test teams not to, IT copied the existing setups straight onto cloudy machines. And then set the limits for cloudy instances up appropriately generously so that the multi threaded memory heavy tests would run well.
And then they got hit by a stupidly high bill in the first month because of course the folding@home jobs had gotten copied across and were configured to use every bit of CPU and memory they could. This cost rather a lot more in the cloud than it had previously in power bills.
One of the things we had at the veterinary clinic I worked at was a strip reader for a multi-test urine strip. Dip strip in urine, wait 5 minutes, put strip in reader. Reader pulls the strip through and dumps it into the trash container.
That thing came from human medicine and was probably 20ish years old. It was toaster oven sized. There wasn't much incentive to make them much smaller though.
At school we had a few spots were a switch was installed with 1 cable in, 1 out, because the 4 put switches and the unshielded cable did not approval, but the shielded cable necessary to cover the same distance without switches or other repeaters was too expensive and would have to go through the (lengthy) tender process.
The fact that this solution was well over twice the cost didn't matter because the individual parts were below the threshold so could simply be bought off the shelf.
Theft is taking with the intention of depriving the owner of the thing stolen. Software or music piracy does not remove the thing copied from the owner's possession. More to the point, a lot of forms of software piracy or music piracy in the US don't (and shouldn't) count as piracy elsewhere, such as ripping CDs to MP3 (for personal use) or downloading videos/songs from YouTube (again, for personal use).
But they're proposing cutting the 9 month journey in half, when we have designs that will straight up cut travel time to 4 to 6 weeks. That we could build and launch with today's technologies.
It's called NERVA. Nuclear Engine for Rocket Vehicle Applications. And while some Nerva designs are T/W > 1, none of the mission designs made ever featured a fuelled Nuclear rocket being launched, they were all intended to launch the nuclear rocket, propellant and nuclear fuel in separate launches for payload mass and safety reasons. Once fuelled the rocket would make multiple trips to Mars and back, only needing new propellant and payloads.
Good LEDs use between 25% and 50% less power than CFLs. But obviously that's a MUCH smaller savings than between CFLs and incandescent bulbs. A good reason to replace CFLs early however is that a lot of CFLs suffer from fading, which hurts both light output and efficiency.
My experience with LEDs so far has been that the early ones had a relatively high early failure rate, usually within the first 2 years. If they didn't die early, they'd last a good 10 years.
I can answer that one. Performance and features. Vulkan performs better, is cross platform and implemented a bunch of new features. Not to mention that DX12 was late to the show. Vulkan APIs were out on win7 before DX12 APIs were available.
And then the few games that implemented both Vulkan and DX 12 found that the former ran faster on the same hardware.
Recently had a fun time with backing data up. A semi competent computer user asked me to help him get a Linux instead going on one of his PCs. Having set up a previous machine of this with Linux in the past, I said sure, helped him decide what distro he wanted and came over armed with a live USB.
I checked with him that he had everything important backed up, he had. So I plug the USB stick in and let it run the pre configured install while we talk.
I walk him through the steps so he can do it himself next time, including the part where the drive gets repartitioned.
At the end I said we were ready to put his backed up stuff on the machine and asked him for it. D:\backup he says. He'd created a new partition to put his back up on. It was of course nuked during the install.
I've learned to always ask for the backup before even starting the machine.
First time we went on holiday and brought the cats, one of them came running in with a young rabbit. She was in a particular hurry because mom wasn't happy about the cat abducting her babies. Fortunately none of those involved were seriously injured and we managed to reunite the rabbits.
The same cat was harassed by a big fat pigeon a few years later. A couple of days after she came strutting into the bungalow with a pigeon almost as big as she was. The pigeon was very much dead already at that point.
The two biggest problems for the F-35 in current counter insurgency operations are speed and its cannon. The much slower A-10 has a much easier time acquiring targets (especially infantry) and the gun is both significantly more powerful and carries a substantial amount more ammo.
For reference, the GAU-8 fires a 30x173mm round, at 3900 rounds per minute with 1350 round carried aboard the A-10.
The GAU-22 fires a 25x135mm round at 3300 rounds per minute and a mere 180 rounds aboard an F-35.
The fact that the F-35 can't loiter as long didn't help of course, nor does the extremely low availablity of the F-35.
Speaking from the far off lands of the Netherlands, but also available in some other European countries, we have a wonderful system called iDeal, where a transaction is stated in the vendor's environment, transferred via a single use number negotiated between vendor and bank and then finishes the transaction in the bank's environment.
We also use IBAN, which can be used to transfer money to an account and identify the account, but cannot be used to charge the account without jumping through a bunch of hoops at which point a large part of the responsibility lies with the bank.
Obviously this is less attractive for Amazon and all, because they can't use their one click purchasing with these systems.
Have I got a great example of that one. The social security managing software that the company I work for makes has some very interesting 'features'. For example, the client in a dossier and his partner are stored in 2 separate fields in said dossier.
When we make a new dossier, we check the existing ones to see if there are any incompatible dossiers around and still active. If there are, we show an error. This check of course only checked for clients, not partners.
So you could give client A and partner B their social security check and then give client B and partner A their social security check as well, paying them twice because partner B didn't already have a dossier as far as the system was concerned.
It only took a year to get it accepted as a bug and fixed, because of course the user and their processes wouldn't allow that situation to happen in reality.
You're missing mental healthcare in that list, I'd argue that the access to guns is the least of the problems in that list, with education and mental healthcare taking the 2 top spots.
Of course, improving education will take time and even if it happens today, it'll be a decade or two before the results start to come around. Improving mental healthcare would produce results faster, but the US seems allergic to any kind of affordable healthcare and has a cultural problem that has only recently started to shift of people not seeking help.
So access to guns and restricting speech that incites others to violence and discrimination are the lie hanging fruit. For certain values there of.
No, you take the new password which hasn't been hashed and the original forgotten yet and compute a whole bunch of likely variations and run those against the hashed old password. If any of them match, the new password is too similar to the old one.
If you don't get any matches, create the hashed password and get rid of the plaintext.