←back to thread

666 points jcartw | 9 comments | | HN request time: 0s | source | bottom
Show context
SwiftyBug ◴[] No.43620583[source]
I've been living in Brazil for the last 20 years.

Pix revolutionised the way we transact in Brazil. I've used Pix to pay for things that cost only cents, and I have a friend who bought her house using Pix. The system just works for any transfer amount. And it's so easy to use.

Its speed is truly baffling, and so is its reliability. Never have I failed to make a Pix payment because of downtime. I never cease to be amazed by how fast money arrives in my Brazilian account when I make a withdrawal directly from my EUR wallet on Wise. I receive a push notification from my Brazilian bank before Wise finishes running the animation of confirmation of withdrawal. It's like magic.

And it's so widespread that nowadays I don't even question whether someone accepts Pix. When I get in a taxi, no matter how old the driver is, it's certain that they take (and prefer) Pix.

I've even had homeless people ask me for Pix instead of change on multiple occasions.

Cryptocurrencies don't stand a chance.

replies(32): >>43620618 #>>43620630 #>>43620653 #>>43620658 #>>43620741 #>>43620818 #>>43620824 #>>43620883 #>>43620954 #>>43621164 #>>43621188 #>>43621341 #>>43621460 #>>43621833 #>>43621851 #>>43622093 #>>43622233 #>>43622479 #>>43622501 #>>43622627 #>>43622888 #>>43623111 #>>43623692 #>>43624345 #>>43624424 #>>43624538 #>>43626843 #>>43627268 #>>43627624 #>>43628278 #>>43629343 #>>43630864 #
yetihehe ◴[] No.43620653[source]
> I receive a push notification from my Brazilian bank before Wise finishes running the animation of confirmation of withdrawal. It's like magic.

After I had to add a special animation for one email system so that user was sure that "the core functionality of encrypting" was indeed working (it took milliseconds in reality), your experience doesn't surprise me that much. But, in my "IoT" system we have a mix of devices. Our service can handle most requests in sub millisecond, but some devices (gprs) need at least minimum 1 second (20sec is still within time limit) to respond only because of slow connectivity. And then I have a parking ticket machines where you press button, wait 2 seconds, it beeps, then after 2sec it changes screen to "printing ticket", then after 2s you get the ticket, where everything can be a local action (free ticket without payment). Technology is wild.

replies(3): >>43620821 #>>43621654 #>>43631265 #
seszett ◴[] No.43620821[source]
The parking ticket machine might make things deliberately slow because the printer needs to warm up or something.

Maybe it needs up to 5 seconds to warm up if it's in deep sleep, so splitting this into three 2s periods provides the least frustrating user experience.

As soon as you need to deal with real hardware things always start to get complicated.

replies(1): >>43620992 #
1. pjc50 ◴[] No.43620992[source]
More likely it's warming up the mobile comms state machine, without checking if it's actually needed. Unlike mobile phones which try to keep their data connection somewhat live, IoT things often drop back to the lowest state to save power (and possibly SIM cost)

https://www.sharetechnote.com/html/Handbook_UMTS_RrcStateCha...

replies(3): >>43621064 #>>43621112 #>>43626674 #
2. actionfromafar ◴[] No.43621064[source]
And programmed on BASIC Stamp on some godforsaken discontinued hardware. :)
3. yetihehe ◴[] No.43621112[source]
More likely it was written by some cheap interns and requires getting unique ticket id from server for "controlling" purposes. Then there is one part time employee (met him, small talked a little) who goes from car to car with terminal and checks if those tickets are valid. I have some experience with gprs systems here, so probable flow:

- press button

- gprs roundtrip about button press with "no payment, free ticket" (2s)

- machine shows "printing ticket", asks server what to print (aka the idiotic unnecessary step)

- gprs roundtrip (2s)

- printer warmup? (?s)

- prints ticket

> to save power (and possibly SIM cost)

Nope, costs per sim are monthly per card, until you hit the data limit, then per MB. Those machines typically have enough power to keep connection alive.

replies(1): >>43621515 #
4. anal_reactor ◴[] No.43621515[source]
More likely a parking machine needs to be accessible to all users, and some people get confused when technology works too fast.
replies(1): >>43621741 #
5. desiderantes ◴[] No.43621741{3}[source]
Perhaps if you get confused by fast things you shouldn't be driving?
replies(1): >>43622163 #
6. inetknght ◴[] No.43622163{4}[source]
Perhaps if you're driving, the things around you need to give you time to react to other things around you. Fewer things are more frustrating than getting honked at because you pressed a button, then got distracted by a car pulling up which you needed to look at to be aware of, then missed the printer asking if you want a receipt, and then having to press another button to talk to someone to ask for a reprint which, of course, holds up the line of cars growing behind you while someone gets paged to come to the kiosk.
replies(1): >>43627045 #
7. dmonitor ◴[] No.43626674[source]
The cell providers also get really opinionated about how much / how often your IoT device talks to the cell towers when they seek to approve your device.
8. desiderantes ◴[] No.43627045{5}[source]
So you wrote this new scenario where the parking ticket machine does NOT print a ticket unless you confirm it (after you already pressed the button)? And you get... mildly inconvenienced by some honking. Yeah you shouldn't drive.
replies(1): >>43627872 #
9. inetknght ◴[] No.43627872{6}[source]
No, not printing the ticket isn't the scenario. Not printing the receipt for payment is the scenario.

It's not a new scenario. It happened to me.

I'm not mildly inconvenienced by some honking. That's just aggravating. I'm inconvenienced by having to call to get a receipt.