In Europe if an airline does that they have to pay you a penalty of a few hundred euro.
But KLM (Dutch flag carrier) found a way around that: if a flight is overloaded by weight they will keep all the passengers on board but leave their luggage behind. There is no direct penalty for late luggage, so many customers will get nothing except perhaps a little free shopping if they feel like filing forms to reimburse for having to buy clothes at their destination. But that's cheaper than the penalty for not taking the passenger on time, so KLM "optimized" it.
Just the mention of pieces of hardware we don't really need anymore (sound cards, modems, etc) triggers a flood of nostalgia. I used to spend DAYs poring over PC part catalogues dreaming of my ideal rig. And brands like Hercules, Creative, Matrox all trigger the same feelings.
Crazy contrast to me having spent the past weekend wondering if cloud gaming services like Geforce Now are matured enough that I can fully move to a thin client - fat server setup for the little bit of gaming I still do.
The technology works, but the business model doesn't, so there's the eternal risk that it might get shut down at short notice with no way to export your saves.
Yeah, that's definitely a worry. Also, the fact that you're dependent on them for adding support for future games, and that (like any cloud service) it might not be available right when you want it.
Yeah that's the issue - nobody wants to just rent you a gaming PC in the cloud, they all want a cut of game sales/licensing. But if someone were to do it, the technology is absolutely there.
You don't even need to create any internal tech - Steam Remote Play already has everything you need, and I successfully used it to play Battlefield from an AWS GPU instance (was even good enough for multiplayer).
If that were true it would also apply to C and C++. I have used Valgrind with Python + Boost C++ hybrid programs and it worked fine after spending an hour making a suppressions file.
Yes, but in my past experience, one often wants to be edit those files to make them more generic… Valgrind struggles to distinguish which parts of the call stack are essential to a “known leak” versus which are coincidental.
I have never tried asking an LLM to do this-but it seems like the kind of problem with which an LLM might have some success, even if only partial success.
For me, a waaay outdated suppressions file for Qt + a rough understanding what syscalls and frameworks do is enough. If my app crashes in a network request and a byte sent to the X server (old example, I use Wayland now) is uninitialized, I know to ignore it.
Valgrind(-memcheck) is an extremely important tool in memory-unsafe languages.
If you care about network bandwidth you can compress before sending, as virtually all web applications do. Then you don't need to worry much about the space efficiency of the application format.
Of the wire format you mean? I compress it and still need to care about the space efficiency of the wire format beyond that. Compression ratio does improve a lot when not doing our own, end result is significantly larger. Also it becomes also significantly slower because more data to process which is possibly the bigger problem.
It's probably not like most web application, it's hardware data loggers that produce about hundreds of millions to billions of events per second (each with minimum about 4 bytes of wire format and maximum roughly 500 bytes).
Not about smoking but I recently stayed at a W hotel and was woken in the middle of the night by the room lights turning on. They used electronic push buttons and I turned them off. Seconds later they turned on again. This repeated several times until I was fully awake and called the front desk.
"We can come put tape on the sensors."
"What sensors?"
"There are sensors under the bed."
"Oh, so you already know about this problem but haven't fixed it. Thanks, please don't send anyone."
I then looked under the bed and sure enough there was a motion detector on each side. I removed these from their brackets and let them dangle facing the floor instead of outward. This blinded them and solved the problem. I guess they were malfunctioning or they were able to detect motion above the bed via reflections.
The next day I reported this to the front desk, who were unsympathetic and unhelpful. They told me it was for my own safety. Apparently at other hotels I have just been incredibly lucky not to have fallen down when getting out of bed.
I will not stay at a W hotel again unless I can confirm in advance that they do not have motion detectors under the bed which spuriously turn the lights on at night. Maybe I'll add Hyatt to the no-go list.
Why the hell would they put occupancy sensors below the bed that trigger the overhead lights, that’s an absurd solution to people tripping and falling at night, provide a nightlight that costs $2 instead of (2) $100 occ sensor/relays.
Possibly the issue was they used PIR/ultrasonic (aka dual-tech) sensors and the ultrasonic one was picking up vibrations, I’ve seen that happen in tenant spaces before and turning down the ultrasonic sensitivity fixed it.
I run electrical work and if I was asked to install these, I would’ve written a sarcastic RFI to make sure the customer actually wanted to do something this stupid and expensive vs a $2 nightlight in a receptacle.
Strange choice, fitting rooms with a novel device to annoy guests. Do you suppose it's because somebody fell out of bed and sued? And then maybe some other people thought that was a good idea, and they fell out of bed too, and now the hotels have to have the annoying thing.
I find it somewhat unlikely, as this particular W hotel was not in a country known for personal injury lawsuits.
More likely it was sold to them by some interior design firm as a luxury feature. Unfortunately it's only helpful if you're alone--even if it worked correctly you wouldn't want the room lights turning on just because your spouse got up.
Can easily see this as another profit centre. If you paid for single occupancy and call down because the lights come on every time your partner gets up, hit ‘em with a big fine.
Stayed at a Marriott property in Germany that had these. Got up in the middle of the night to pee and the automatic lights woke up my partner.
I carry black electrical tape whenever I travel. It's marvelous for disabling sensors and covering up too-bright LEDs that light up the room all night.
One could argue that I shouldn't because I'm "improving" their property but reasonable people could disagree about the definition of "improving." Bottom line is that it's their property and their rules but if I can make a nondestructive change to make the place more comfortable while I'm staying there, I will.
Oh wow I ran into problems with those too. When I brought my cats to a Hilton, they would get the zoomies and run around at random in the middle of the night, which would make the lights turn on. I think I found some way to block the sensor.
I'm staying at a Candlewood in southern California and they've got motion sensing light switches for everything. They mostly go to plugs, though, and whoever stayed in the room before me unplugged all the lights. I left 'em that way. The bathroom light came on automatically blinded me three times a night when I'd get up to pee until I got in the habit of reaching in and hitting the button before stepping into the bathroom.
Automatic lights in private spaces are just a hassle.
I haven't seen the makers of Aeron (or anyone else) claim it's "orders of magnitude faster than plain UDP." Do you have a link to something about this? It doesn't pass the smell test for me unless you're talking specifically about using Aeron within a single machine (where it uses shared memory instead of the network)...but you said "Within a datacenter" not "Within a computer."
It has been a while since I saw their presentation. What I remember is that Aaron has an insanely low delay even in the high percentiles, that is orders of magnitude better. Throughput for a large stream of data is probably similar to plain UDP.
Please correct me if I remember wrong.
It does have great tail latency. But it's not a silver bullet, but careful engineering. And you pay for the latency with spinning threads. It's the architecture that makes it to stand out. In the end, it's just the same old UDP sockets, not even io_uring at least in the free public version. But one can use LD_PRELOAD if hardware has this trick - but again, it's not specific to Aeron.
I came here to write something very similar. Visa and their competitors all have the ability to transport data about the payment. Most receipts are a fairly small raster image which could be sent as an extension. Visa could also offer a structured text receipt data format but the image may actually be safer to avoid a billion questions about how to represent unusual situations around surcharges, optional extras, custom discounts, split checks etc.
It would be important for the receipt to include the original broken-out list of items purchased, and not just the "credit card receipt" with the total paid. People need detailed receipts for returns and business expenses.
In many cases, the POS software would have to go out of its way to share the itemized receipt details with the credit card processor. There's probably an incentive to get paid for that data elsewhere that overshadows the benefit to the user.
Some of the payment systems do now allow you to associate an email address with cards which gets you itemised receipts at the cost of privacy.
I actually do use this a lot. Have a corp card and corp email and it’s setup with stripe, square, bunnings (an australia thing) etc. to email receipts. For the corp use case its great. Spam that one use corp email all you want, just send me the digital receipts.
But KLM (Dutch flag carrier) found a way around that: if a flight is overloaded by weight they will keep all the passengers on board but leave their luggage behind. There is no direct penalty for late luggage, so many customers will get nothing except perhaps a little free shopping if they feel like filing forms to reimburse for having to buy clothes at their destination. But that's cheaper than the penalty for not taking the passenger on time, so KLM "optimized" it.
reply