(Thread IKs:
Platystemon)
|
Jon Do posted:"On Something Awful, we ruled with photoshop and troll power," the Moderator said. "Here, we must scrabble for Twitter power. This is your inheritance, C-SPAM." He who controls the ratio controls the twitterverse.
|
# ? Feb 2, 2020 08:53 |
|
|
# ? May 27, 2024 16:58 |
|
Inceltown posted:https://video.twimg.com/amplify_video/1220752921962405893/vid/1280x720/16WISiyOQtiHD699.mp4?tag=13 His masters porn?
|
# ? Feb 2, 2020 10:21 |
|
Centrist Committee posted:Look at this social butterfly!
|
# ? Feb 2, 2020 11:40 |
|
*insert gif of yang dispensing whipped cream into the mouths of supporters while his campaign manager balks n looks resigned here*
|
# ? Feb 2, 2020 11:58 |
|
https://twitter.com/kmlefranc/status/1221869659139366912?s=21
|
# ? Feb 2, 2020 12:46 |
|
Eye‐searing illuminated billboard at W Washington Boulevard and S Hill Street in Los Angeles
|
# ? Feb 2, 2020 12:59 |
|
|
# ? Feb 2, 2020 13:20 |
|
Platystemon posted:
What an absolute nightmare to have that constantly flashing into your windows
|
# ? Feb 2, 2020 13:57 |
|
Enfys posted:What an absolute nightmare to have that constantly flashing into your windows They're planning to build this just round from my flat "Imagine about 42 IMAX movie screens all tiled together in a giant see-though sphere" Needless to say there's some objections from residents...
|
# ? Feb 2, 2020 15:22 |
|
https://twitter.com/stevenjcrowley/status/1223977380794064897
|
# ? Feb 2, 2020 16:17 |
|
https://www.geek.com/tech/this-bag-of-tostitos-is-also-a-breathalyzer-that-calls-an-uber-if-youre-drunk-1686713/
|
# ? Feb 2, 2020 17:44 |
|
ekuNNN posted:
Apparently they're making around 25k of these things. Lots more useless electronics in the garbage!
|
# ? Feb 2, 2020 17:48 |
|
StashAugustine posted:butlerian jihad now
|
# ? Feb 2, 2020 18:45 |
|
https://twitter.com/GalaxyKate/status/1223641049748230145 https://twitter.com/GalaxyKate/status/1223643462483886081
|
# ? Feb 2, 2020 19:22 |
|
Alexa, we need guns edit for youtube version: https://www.youtube.com/watch?v=7xbglr1pEiM ekuNNN has issued a correction as of 23:48 on Feb 2, 2020 |
# ? Feb 2, 2020 22:31 |
|
Security guard at work today could not shut up about how cool this was and how he needed one, "To keep the Mounties from finding my toys" more proof that the only thing worse than a cop is a wannabe cop I guess.
|
# ? Feb 2, 2020 23:08 |
|
can you copy paste it here? I can't read it without an account (lol)
|
# ? Feb 2, 2020 23:09 |
|
Bert Roberge posted:Apparently they're making around 25k of these things. Lots more useless electronics in the garbage! oh sweet they took some time off supporting nazis and shooting black people to make a funny joke about stupid cyberpunk nachos
|
# ? Feb 2, 2020 23:12 |
|
actionjackson posted:can you copy paste it here? I can't read it without an account (lol) found it on youtube https://www.youtube.com/watch?v=7xbglr1pEiM
|
# ? Feb 2, 2020 23:45 |
|
Bert Roberge posted:Apparently they're making around 25k of these things. Lots more useless electronics in the garbage! not gonna lie it seems bad design to make the “don’t drive” image a literal steering wheel
|
# ? Feb 3, 2020 00:08 |
|
Scrub-Niggurath posted:not gonna lie it seems bad design to make the “don’t drive” image a literal steering wheel Yeah, also if that image is really indicative of how it looks it's a massive UX failure all around; it should say "don't drink and drive" all the time in unfired/greenlight state and change to "you're drunk don't drive you moron" when redlight... Though it also requires assuming breathalyzers are accurate (which they aren't) and that some arbitrary BAC score is really indicative of one's impairment (which it isn't) or that drunk morons are gonna go out of their way to make sure it detects them blowing hot (which they won't) and that anyone who fails the test will take it as a serious indicator of whether to go driving (which they wouldn't need to do if they have good judgment and are in a rational state of mind...) so all it does is muddle the decision-making for people who are already not in a good state to be making decisions about where to fling their ton of steel. I'd hesitate to say 'the cops are right' but in this case it's more like 'the marketing gimmick is bad pseudoscience masquerading as personal responsibility'
|
# ? Feb 3, 2020 00:30 |
|
How long before someone uses one of them and then drives only to get busted and then sue Tostitos?
|
# ? Feb 3, 2020 00:33 |
|
Inceltown posted:How long before someone uses one of them and then drives only to get busted and then sue Tostitos? That’s the good outcome for Tostitos. The bad outcome is that a customer drives drunk, crashes, and the drunk’s and victim’s families both sue.
|
# ? Feb 3, 2020 00:47 |
|
https://twitter.com/johnjcook/status/1222906507517407232?s=21 https://twitter.com/tomscocca/status/1222908054867529729?s=21 ugh
|
# ? Feb 3, 2020 02:26 |
|
cross-posting from the modern Star Trek threadAntherUslessPoster posted:If someone wants a somewhat realistic portrayal of UIs and poo poo they can watch The Expanse. adaz posted:so uh, about that. it did make it in and it did kill people. I would highly recommend that article to anyone interested in computers, failure engineering, UI design, or general how hard can it be to design a good ui for ships???
|
# ? Feb 3, 2020 02:53 |
|
Honestly? Not surprised. Got a loaner from the dealer when I brought my car in for some recall work. Why on earth you would put in a control system (that giant dash-filling touch display that seems to plague every goddamn modern car these days) that requires you to do unsafe things to manipulate it, then have something (a driver monitoring feature) yell at you for doing unsafe things is completely beyond me. I think I've come to the conclusion that I will never be buying anything except the shittiest of Japanese shitboxes, where they can't afford to put in non-physical controls. SwissArmyDruid has issued a correction as of 03:44 on Feb 3, 2020 |
# ? Feb 3, 2020 03:13 |
|
SwissArmyDruid posted:Honestly? Not surprised. Got a loaner from the dealer when I brought it in for some recall work. Why on earth you would put in a control system (that giant dash-filling touch display that seems to plague every goddamn modern car these days) that requires you to do unsafe things to manipulate it, then have something (a driver monitoring feature) yell at you for doing unsafe things is completely beyond me. lmao from the article on how the Navy had to fork out money to change back to physical more intuitive controls after the accident.
|
# ? Feb 3, 2020 03:16 |
|
Teslas design aesthetic is a cancer that's spreading to rest of the industry.
|
# ? Feb 3, 2020 03:37 |
|
Nothus posted:Teslas design aesthetic is a cancer that's spreading to rest of the industry. so they weren't kidding when they mentioned tesla was an inspiration in that awful star trek movie
|
# ? Feb 3, 2020 03:44 |
|
Which awful star trek movie? IIRC the most recent TOS-inspired ones had physical controls... I vaguely remember a making out the shape of a giant shiny throttle at the conn past all the lens flare.
|
# ? Feb 3, 2020 03:46 |
|
galenanorth posted:cross-posting from the modern Star Trek thread Actually reading that article makes it sound a lot more like the people were never trained on the software rather than the software being bad. The software definitely wasn't great, but not knowing how to tell what console is controlling what when it says that at the upper left and whether the shafts are ganged or not by both the indicator of such and the GIANT GREEN BARS not obviously matching each other are evidence that the operator has no idea what they're doing. I've never driven a warship and even I can tell how to do all of the functions they hosed up just by looking at the layout and labelling of the buttons. You're telling me the guy who's job it is to do this stuff can't? Either they were never shown or they were incompetent or both. Anyone who's worked anywhere at all has dealt with really lovely internal software that's poorly documented and has weird special cases that you don't see every day that only some people fully understand. Still, it is your job to learn it if you're going to be using it. That software seems a lot more intuitive than the software I use at work if I'm being honest.
|
# ? Feb 3, 2020 04:32 |
|
SwissArmyDruid posted:Honestly? Not surprised. Got a loaner from the dealer when I brought my car in for some recall work. Why on earth you would put in a control system (that giant dash-filling touch display that seems to plague every goddamn modern car these days) that requires you to do unsafe things to manipulate it, then have something (a driver monitoring feature) yell at you for doing unsafe things is completely beyond me. 2019 and newer cars are required to have a backup camera which in essence is a requirement for them to have screen. It saves the manufacturer a lot more money to have commonality in software for the screen than it would for the entry level cars to have a completely different control scheme when they're already mandated to have the screen. The only difference between the low end and the high end would likely be the quality of the screen because you've got to figure everything else is a commodity component. In other words, I've got some bad news
|
# ? Feb 3, 2020 05:12 |
|
LastInLine posted:2019 and newer cars are required to have a backup camera which in essence is a requirement for them to have screen. It saves the manufacturer a lot more money to have commonality in software for the screen than it would for the entry level cars to have a completely different control scheme when they're already mandated to have the screen. The only difference between the low end and the high end would likely be the quality of the screen because you've got to figure everything else is a commodity component. Shrug. It just means that I'm never getting off the car I'm on until the day that I stop driving entirely, whether by autopilot finally being good enough, or we finally get our heads out of our asses and fix public transportation.
|
# ? Feb 3, 2020 05:20 |
|
LastInLine posted:I've never driven a warship and even I can tell how to do all of the functions they hosed up just by looking at the layout and labelling of the buttons. You're telling me the guy who's job it is to do this stuff can't? Either they were never shown or they were incompetent or both. lol come on how can you say that with confidence if you've never used the software? Maybe a problem with the UI is that the intuitive sense you are getting looking at it is completely wrong like this post has a real "pssh I could have made that free throw" energy
|
# ? Feb 3, 2020 14:11 |
|
Peanut Butler posted:lol come on how can you say that with confidence if you've never used the software? Maybe a problem with the UI is that the intuitive sense you are getting looking at it is completely wrong Did you read the article? I was going into it with the attitude of "This is internal corporate software but worse! This is an inside look at government internal software and how bad it actually is!" The article itself actually shows you with animations what it does and how to work it as you scroll.
|
# ? Feb 3, 2020 14:26 |
|
The article does mention a lot of contributing and mitigating factors, like the inexperience of the helmsman and the fact that the helm split wasn't something the guy was used to, but those are human factors, not software deficiencies. The article also mentions that the software is a piece of poo poo, crashing when fed too much data and when different features reached arbitrary limits, but remarkably none of the ways in which it is a piece of poo poo contributed to the accident. The accident was 100% caused by every person involved not knowing what they were doing, screwing up what they were supposed to do, then being unable to look at the screen and determine what it is they screwed up and reading between the lines, the one guy who could do all of those things didn't act fast enough and was covering his rear end in the inquiry. I will concede that maybe the software just straight up fails and reports incorrect data in which case no one could do it but that's not what the Navy investigation or the Propublica investigation found. If that happened you'd think someone would claim that it did and no one is.
|
# ? Feb 3, 2020 14:36 |
|
LastInLine posted:The accident was 100% caused by every person involved not knowing what they were doing, screwing up what they were supposed to do, then being unable to look at the screen and determine what it is they screwed up and reading between the lines, the one guy who could do all of those things didn't act fast enough and was covering his rear end in the inquiry. all of these things are signs of software deficiencies in systems of this scale and importance and complexity, an important design goal is to minimize the opportunity for human error and maximize the opportunities to recover from human error, while also avoiding imposing unnecessary cognitive load on the human operator obviously, no amount of interface design can compensate for someone who is absolutely determined to do the wrong thing at every opportunity. but it's crucial to make the right thing easy and obvious to do while making the wrong thing easy to recover from, so that even a inexperienced operator who's tired at the end of a late shift and panicking because they're ten seconds from crashing can handle it
|
# ? Feb 3, 2020 15:23 |
|
Main Paineframe posted:all of these things are signs of software deficiencies Yeah but the defense industry only drains a quadrillion dollars per year from more worthwhile causes like actualizing Space Jam, how could they ever possibly afford all the engineering and QA that goes into that kind of package?
|
# ? Feb 3, 2020 16:54 |
|
Main Paineframe posted:all of these things are signs of software deficiencies Exactly, and what silentsnack adds is crucial, they have an unlimited budget to get it right. I just think the Navy investigation was mostly correct: the operators weren't trained properly. The key element which leads me to say that the interface was good enough is that, "an important design goal is to minimize the opportunity for human error and maximize the opportunities to recover from human error" and that was definitely done. The causative action was assigning the propulsion to the Lee Helm who in accepting controls decoupled the drives. When they attempted to slow the ship by dragging one of the indicators down, the starboard side didn't go with it. That caused the turn. The interface could've made it clearer that accepting without checking the box would decouple them. It might default to the way a user wouldn't want, I don't know. The interface does, though, immediately show the user that they've only reduced one side if they aren't coupled. The article mentions that *exactly* at this time a different alarm distracted the user and that a monitoring officer "eventually" reached over and recoupled the drives. As you say the shifts are brutal in the Navy and were found to be causative in the accident two months prior. I don't doubt the user was at his wits end but there's only so much the interface can do and would seem at least in this case it was clear to the operator whose only job was to make sure propulsion was as ordered wasn't done correctly. Again, his only job was to look at those two lines and verify that they *both* were at the level ordered and he didn't. It's hard to say it's cognitive overload when that was the only thing he was tasked with especially when the first guy to look over his shoulder saw the problem immediately.
|
# ? Feb 3, 2020 18:12 |
|
|
# ? May 27, 2024 16:58 |
|
LastInLine posted:Exactly, and what silentsnack adds is crucial, they have an unlimited budget to get it right. I just think the Navy investigation was mostly correct: the operators weren't trained properly. the story you tell here does not match what's in the linked article, so I don't know where you're getting stuff like the bit about someone supposedly looking over a shoulder and immediately seeing the problem, or how it only happened because someone didn't check a required checkbox but indicators and warnings and options are not a substitute for good design. in fact, they can aggravate bad design by overloading a user's attention capacity during information-dense moments the system should default to the more common, more usable, and more intuitive option. the more complicated and dangerous option should always take more effort to choose, so that if someone isn't paying attention and misses a checkbox, the resulting human failure fails safe. the fact that the default workflow for transferring propulsion control was to divide control of the individual propeller shafts is more than enough to say that the software was a significant contributing factor, because that is an opportunity for human error. sure, the problem could have been avoided if the person doing the transfer wasn't distracted when the system was in a dangerous state during an incomplete transfer of control, but the distraction was only a problem because the software unnecessarily required the human operator to temporarily put it into that dangerous state
|
# ? Feb 3, 2020 19:01 |