Register a SA Forums Account here!
JOINING THE SA FORUMS WILL REMOVE THIS BIG AD, THE ANNOYING UNDERLINED ADS, AND STUPID INTERSTITIAL ADS!!!

You can: log in, read the tech support FAQ, or request your lost password. This dumb message (and those ads) will appear on every screen until you register! Get rid of this crap by registering your own SA Forums Account and joining roughly 150,000 Goons, for the one-time price of $9.95! We charge money because it costs us money per month for bills, and since we don't believe in showing ads to our users, we try to make the money back through forum registrations.
How many quarters after Q1 2016 till Marissa Mayer is unemployed?
1 or fewer
2
4
Her job is guaranteed; what are you even talking about?
View Results
 
  • Post
  • Reply
Kyte
Nov 19, 2013

Never quacked for this
It almost sounds like you're saying things were good when it was an all-american oligopoly until those asians with cheap products ruined everything.

luxury handset posted:

i really don't think it's productive to think of software updates as "loving people over". it's not a very compelling stance, more slogany than anything

I don't think you understand. The loving people over isn't in the updates, it's in the product design.

Like, Sonos could've designed the product such that the control board is a slottable board that can be replaced once the electronics get obsolete, without having to replace the speakers themselves. It'd still be exactly as plug and play as before, but now you only need to change a small part of it once every 5 years or whatever.
Would make it easier to service, too.

It's particularly easy in this instance because there is a very well defined interface between both components so you don't even need to worry about backcompat.

Kyte fucked around with this message at 15:02 on Jan 24, 2020

Adbot
ADBOT LOVES YOU

Kyte
Nov 19, 2013

Never quacked for this

luxury handset posted:

they could have, sure, but i don't think we can fault any manufacturer for not building their products to support post end-of-life reuse or a perpetual upgrade cycle. it's a valid design philosophy, but choosing a different emphasis on product design is also valid. like, the entirety of the apple product ecosystem is all about "difficult to upgrade or tinker with", and if we're going to fault that as just attributable to greed then how do we fault the consumers who prefer single use or single lifecycle products?
I can and will absolutely fault any product designer for designing their products without thinking in the long term use of their product. Disposable consumption is like 90% of the reason we're in such a lovely global state right now.

And yes, Apple is awful about it. And people prove they're idiots by buying into it.


luxury handset posted:

not really? there's tons of products out there which no longer have a viable maintenance/support system for them, the only variance is timescale. try taking a 1973 dodge to a dealer to get it fixed, or the lengths you'd have to go to find someone who can service a 486 running windows 95. eventually all things become subject to the degredation of time, and this point comes quicker the more complex or obscure a product is.
There's an incredibly huge gulf between technology creep and planned obsolescence. Yes, eventually things will become useless or unsafe or whatnot.
But products can be designed to last longer, at least partially. Modularity is an example of this. The example I gave before is like that. Separate the well-proven speaker from the quicker-evolving control unit, then allow the latter to be replaced.

quote:

i agree with you that many companies do design for products to die on a quicker cycle to generate revenue, this happens a lot and it shouldn't. but also, it's not a useful expectation for everything to be designed for maximal longevity given the pace at which consumer technology develops. for example, consumer safety in traffic would be lessened if everyone were still driving easily maintained vehicles built to mid 80's safety standards

i'm really not sure how you would write legislation to move product standards from "won't blow up randomly" to "will still be usable and able to interface with contemporary technology in three decades time" as a standard. like that would mostly kill all web development, streaming video, most of the internet today if it still had to run on computers built in the 90s - not that this would be a bad thing, but it would be a tough law to write
This is such a bad faith argument holy poo poo.

Kyte
Nov 19, 2013

Never quacked for this
The bad faith argument here is in immediately assuming a wide-ranging ban is the only option. "ban embedded computable devices which can't be separated from peripheral objects" is ridiculously vague.

Have consumer receivers like smarts TVs and audio players and such allow third party firmware updates.
Allow their boards to be serviced by third parties as well.
Recommend schemes that allow the peripherals to be kept while replacing the control innards.
Ban the use of firmware lockouts or authentication schemes that deny third parties from repairing systems.
Create labeling regulation that indicates relevant recyling, repair and production info to the consumer.
To elaborate on this one:
Over here in Chile all food must have special seals applied if a food product is high in salt/fat/calories/sugar, so customers can at a glance tell how bad for you
something can be.
Meanwhile domestic appliances have to apply an easy-to-interpret label that state their power efficiency.
Mobile phones must have labels that indicate network compatibility.
There can definitely be laws stating that products must have labels like "Cannot be fixed piecewise", "Manufacturer-locked software", etc.



And just toss out DRM, really. Half of the bullshit in consumer electronics has to do with protecting their precious DRM enforcing hardware.

Kyte fucked around with this message at 17:01 on Jan 24, 2020

Kyte
Nov 19, 2013

Never quacked for this

luxury handset posted:

i'd say most people who use a transit system or travel on foot use it frequently enough to not need cues as to when to embark/disembark or which path to take. the majority of daily trips tend to be repetitive, especially commutes as the largest individual component of trip category. again, this is just my anecdotal observations, each individual has their own perspective

I am constantly using the phone to anything but my daily commute. Street signs are often not legible from the bus and we don't have stops announcement. Not that I'd remember the stop name or hear it over the general noise, anyways.
Even in the subway, which does announce current stop/next stop, I often miss it 'cause the speakers are kinda crappy (depends on the line) or I'm just plain distracted and only the newest lines (which I don't get to use often) have little lights that show your current position, otherwise I have to rely on the signs within the station platform and those might not be visible depending on various circumstances.

It's far easier and quicker to just check your position vs the stop's position.

Kyte
Nov 19, 2013

Never quacked for this

AceOfFlames posted:

Oh no! Now who will pay for all those podcast ads?

Mayhaps I might interest you in a certain mobile game called Raid: Shadow Legends.

Kyte
Nov 19, 2013

Never quacked for this
I use Zomato and it works fairly well, and googling them for a bit have a surprisingly small number of controversies.

Of course, which service works best depends entirely on where exactly you're looking.

Kyte
Nov 19, 2013

Never quacked for this
I never super got along with my classmates since I lived in the nerd corner but I still felt emotional when it happened. It still represents a farewell to people you spent a big chunk of your life with* and the end of an significant stage of your life.
To just dismiss it as a waste of time is silly.

* To be fair due to the way the school was structured it was about 15 years for me and my classmates, so it was probably more significant to me than to most people.

Kyte
Nov 19, 2013

Never quacked for this
I've heard of this case in recent news and stuff, but there's one thing I'm not quite clear on.

Did Google only reimplement the Java API? Articles I've read seem to imply they also copied framework code as part of the reimplementation process, which would've definitely been a gross violation of copyright that they'd only get away with because they can afford the legal costs.

Kyte
Nov 19, 2013

Never quacked for this

Platystemon posted:

No.

The ruling says that even assuming the least charitable facts to Google’s case, in which copyright did apply, it would still be fair use.

Isn't that kinda huge for, say, emulators or knockoff consoles or such?

Kyte
Nov 19, 2013

Never quacked for this
Alright, thanks for clarifying.

Kyte
Nov 19, 2013

Never quacked for this

Motronic posted:

You are seriously asking this question. In this thread?

There is a world of difference between utility company owned and installed devices that do this versus your IoT garbage thermostat having third party access over the internet.

Not living in a country where AC is standard-issue, I'm a bit confused. Wouldn't a company-owned thermostat still end up being IoT garbage accessing through the internet?
And utilities don't strike me as the kind of company that properly secure their networks.

Kyte
Nov 19, 2013

Never quacked for this

CommieGIR posted:

We need Nuclear + Renewables. The idea that any energy storage outside of Hydro is going to be feasible in the long run is insane.

Its worth noting Australia is struggling to give up the addiction and this battery setup is largely a distraction from their massive carbon debt from being a net coal/gas exporter and user.
A shitton of places have a massive rain deficit. The storage capacity of hydro is diminishing.

Kyte
Nov 19, 2013

Never quacked for this

starkebn posted:

This is so stupid. Did they suddenly not think his account was actually run by Danny DeVito or is the blue check just a status symbol.

Rhetorical question - I know Twitter blue check is just a loving "I'm better than you" bullshit

https://www.nbcnews.com/news/us-news/danny-devito-briefly-lost-his-twitter-verification-it-wasn-t-n1277255

quote:

In a statement to NBC News on Friday, a Twitter spokesperson confirmed DeVito lost the prized blue check mark. But, the spokesperson said, "his account was debadged temporarily because the account’s information was incomplete."
Of course, it's just coincidence that it only mattered right then and there am i right.

Kyte
Nov 19, 2013

Never quacked for this

Tuxedo Gin posted:

At the end of the day we don't really know though. Look at what leaded gas did to boomers. Who knows what'll happen to us when (if) we grow old. It might not cause cancer but we are still woefully ignorant about the inner workings lf the brain and how environmental factors affect it.

Tetraethyl lead was originally sold as "ethyl fluid" and they had people go out and claim the additive was safe to use (please ignore the news about our manufacturing plants).
The people involved absolutely knew what was going with lead back then.

Kyte
Nov 19, 2013

Never quacked for this

Mister Facetious posted:

Thank Christ.

Also gently caress any reply system that doesn't use chronological order. Reddit can burn.

I don't understand this. I assume you're referring to top-level replies here, right? If so, why not just sort by new?

Kyte
Nov 19, 2013

Never quacked for this
https://twitter.com/spurtmagoo/status/1469677182481932288

Kyte
Nov 19, 2013

Never quacked for this
The video remarks that the original FastPass was a system made by Operations, whereas FP+ and successors were essentially Marketing projects. The former was made to solve a problem. The latter were made to make money.
I remember going to Disney when FP was still active and it was legit great. It helped I went during off-season but still.

Kyte
Nov 19, 2013

Never quacked for this
It's funnier than that.

So the token's contract had a little function that is called in every transaction (create coin, transfer coin, destroy coin). All it does it check if the destination is the owner's address and if it does, it errors out.
But the owner can change the ownership of the token.

So after a while, the owner changed the ownership to one of the biggest exchanges. Now nobody can cash out because part of the process involves a transfer to the exchange (which is now erroring), so the coins accumulate inside the exchange, at which point the guy can just drain the exchange of its coins.
(or something like that, I don't quite understand ethereum infrastructure)

Apparently the so-called verifiers didn't think the implications of the "programming error". Which are really loving obvious.

Kyte
Nov 19, 2013

Never quacked for this

CAPTAIN CAPSLOCK posted:

But why?

Rich people are insanely bored and stupid

Implanting my favorite amiibo to my hand.

Kyte
Nov 19, 2013

Never quacked for this

Owlofcreamcheese posted:

Eh, there isn't really a "the server youtube runs on" or anything. I'm sure youtube is just a bunch of docker containers or something similar kinda running on an as needed basis in endless random server racks serving a bunch of files from a bunch of weird collocation endpoints.

Which isn't to say that it is totally abstract and only exists mystically in "the cloud" but more like, I'm sure it's nothing special and just exists as some allocation on server racks in general more or less like all modern web stuff does. Like I'm sure most of the bare metal it runs on is just running various hypervisors and the hardware running youtube now will be running some business website in 10 minutes and a porn site 10 minutes after that and an MMO login server after that and so on and so on. Cloud stuff is mostly having tons of interchangeable hardware and running whatever whenever as it scales up and down. I'm sure youtube is just thrown in that sort of general population on google's cloud servers. I doubt they build much "YOUTUBE SERVER, HERE IS THE HARD DRIVE YOUTUBE IS SAVED ON, IT PLUGS IN RIGHT HERE" and more "On US east we'd like this function to scale from 800 to 800,000 servers, this function is serverless, we want to cache these videos here and those videos can go into cheaper deep storage"

It's not that simple. Youtube handles livestreams and the automatic archival thereof and live video delivery has much stricter timing requirements than regular content.
A stream with 100k CCV and rewind on has to be transported to 100k people but some of those are actually watching the vod which needs to be generated and delivered on the spot and actually it's not one stream but rather 6 to 9 different ones 'cause the source's being transcoded into all available resolutions and the vods too.

Which reminds me I can actually tell they've got lag in disseminating vods across their CDN because a newly made vod loads much slower (like, "buffers every 5s" slow) than one that's about half a day old. Interestingly livestreams work fine which suggest they use a different (and quicker) channel.

Motronic posted:

It's so much worse, as you know. "I want to put one server in the cloud". Well, ackshually you need to run that across 3 availability zones in at least two regions which means you need to rearchitect everything. Also, since you now have 6 boxes you can buy this HA/proxy product from us to make this actually work again.

Whoops.....what happened to my cheap cloud hosting?

There is absolutely a place and a time for it. It's super useful for spike loads and many other things. But "everything, all the time" is just stupid.

How is having one cloud instance different from hosting in your one local machine/datacenter? Everything you mention is a service level upgrade that clearly you weren't planning for 'cause that's why you only had one server to begin with.

Kyte fucked around with this message at 09:41 on Feb 4, 2022

Kyte
Nov 19, 2013

Never quacked for this

BiggerBoat posted:

Hm, yes, let's see... a tank of helium, 2 rolls of duct tape and an air tight duffel bag. Check.

People Who Bought this Also Purchased: Ammo, rope, razor blades, knives, contractor bags, Dark Side of the Moon CD, Bible, ammonia, air fresheners

Proceed to Checkout? Y/N?

*ads for estate planners displayed*

*No red flags raised by algorithm*

The algorithm doesn't know what things are used for. How could it raise a red flag?

However, it's the responsibility of Amazon to blacklist certain combinations of products/categories. Especially when it's an issue that's been raised within the platform already.

Kyte
Nov 19, 2013

Never quacked for this

BiggerBoat posted:

The same way it determines and knows everything else I might want to buy or what my habits are. You just have to put as much as effort into identifying problematic combinations as you do "man, this buyer sure does purchase a lot rope and books on lynchings". So instead of automatically suggesting instructional manuals on building a gallows, maybe program it to identify serious issues like the article talked about. It was popping up suggestions for other products that help facilitate suicide. Yes?

Not sure I follow you here.

I am 99% certain the suggestion algorithm is much simpler than you think it is. It's just correlations between product IDs (and possibly their descriptive text), it doesn't have any understanding of the meaning of the products.
If a number of people buy a bunch of products together, then the algorithm will consider them correlated (might do two or three degrees of separation, who knows) without understanding why they're being bought together. If it's a low-traffic product, then it'd need only a few purchases to influence the suggestions.

Kyte
Nov 19, 2013

Never quacked for this

Sagacity posted:

It's in Amazon's best interest to fix this issue, because people that have committed suicide do little to no repeat business at Amazon. This is the only reason why this will be addressed.

I couldn't find any sodium nitrite when searching beyond curing salt that includes "X% sodium nitrite" in the title so I guess they did something.

Kyte fucked around with this message at 10:06 on Feb 7, 2022

Kyte
Nov 19, 2013

Never quacked for this

Detective No. 27 posted:

People killing themselves will affect the immediate quarterly profits, so they care. Amazon doesn't care about getting long term business seeing as no one can afford to have children, ensuring they won't have customers years from now. Hell, they're so anti-children they programmed Alexa to teach kids how to kill themselves.

No they didn't, Alexa just did the google thing where it sifts through search results and answers based on whatever sounds related based on keywords. Sometimes those results are awful. Also, it relies on Bing.

For the electrocution one, the kid asked for "a challenge" and Alexa returned with a TikTok challenge and those are like 10% legit 40% stupid and 50% self-destructive so it was just a matter of odds really.
For the greater good one, the person asked about the cardiac cycle and Alexa returned with Wikipedia which had, at the moment, been vandalized.
You can even see copycat vandalism in the page's history.

E: some of you need to understand that, in the majority of cases, companies like Amazon aren't actively malicious, they just don't loving care. And their algorithms don't understand enough to be able to care either way.

Kyte
Nov 19, 2013

Never quacked for this
Following the toothbrush rabbit hole, I came upon this:
https://twitter.com/chadloder/status/1491487441605128197

Kyte
Nov 19, 2013

Never quacked for this

TACD posted:

Is it really that difficult to take the touchscreen ordering thingies they already have indoors and put them within reach of vehicles??
Imagine how grody those things would get.

Also I wouldn't be surprised if exposure to the elements fucks with their function, which means they either become a financial burden with replacement/repair or a throughput nightmare without.

Kyte
Nov 19, 2013

Never quacked for this

HootTheOwl posted:

Someone was playing code golf and did
code:
if(IsValid == IsGuardian == True)
and then felt really clever for getting to check all conditions in one line.

if(IsValid && IsGuardian) is shorter though? Then again that's probably the joke and I'm just missing it.

Kyte
Nov 19, 2013

Never quacked for this
Even if it doesn't trigger the actual ASMR response they often make for nice white-ish noise for working or sleeping.

Kyte
Nov 19, 2013

Never quacked for this
Do encrypted PDFs work? It's how I usually receive bank and pension statements over email.

Kyte
Nov 19, 2013

Never quacked for this

goatsestretchgoals posted:

hardware mute can be anything from "hi operating system, please mute now" to a physical disconnection

assuming its the latter, it should be okay

When Windows mutes the mic the software will be incapable of picking up anything unless you start messing with things at a deep enough level where all this metaanalysis is moot because you could just use proper surveillance tools instead.

Kyte
Nov 19, 2013

Never quacked for this

shrike82 posted:

i've always wondered about the employee psychology behind that mindset - there's a divide between tech employees that buy into the "we're on a mission to change the world" tech cult and the hyper-cynical folks that are just there for the paycheck

it happens even in super-staid places like Google and Microsoft

people want to believe their work is fulfilling?

Kyte
Nov 19, 2013

Never quacked for this

Arsenic Lupin posted:

As we said in the software industry, bullshit in => bullshit out.

You cannot quantify race. This software quantifies race on a scale from 1 to 10, with "Arab" giving you a 7 on race and "Egyptian" giving you a 7 on culture. As has been pointed out by various people, the robot has a higher race score than some of the characters of color. To be data-driven, you have to actually have data, not simply make it up. Assigning something a number does not make it data. Several of the other 1-10s are equally bullshit.

One of the devs at the company has already posted saying she's never seen this chart, and it has to be an upper management thing, because at her level, they just use their eyes. (What she doesn't say is, "and are then overruled because the execs say 'Can't she be a little lighter?'".)

Sigh. Bad snipe, no biscuit.

https://twitter.com/_mlktea/status/1525508828627841024
Have the tweets for context.

tbh, "super dumb poo poo meant to boil things down to Numbers to appease the executive class" isn't unprecedented.

Kyte
Nov 19, 2013

Never quacked for this

x1o posted:

nodejs and npm enable a special kind of stupid which is hilarious to watch from a safe distance. My favorite npm issue was when they released a new version which would mangle your filesystem permissions if you ran `sudo npm help`

I know of the leftpad incident, but I haven't heard of this one. Mind sharing?

Kyte
Nov 19, 2013

Never quacked for this

His Divine Shadow posted:

I loving hate that pixel and google crap, I tried removing it and other tracking code temporarily and the responsiveness of the site almost doubled in pagespeed insights. Pagespeed insights is reporting a lot of issues that could be improved and it's basically all stuff to do with googles own tracking code that's suboptimal. Console says similar things too. Google is the biggest hurdle in getting your websit to do well on it's own metrics...

Probably so they can push for AMP.

Kyte
Nov 19, 2013

Never quacked for this
For my internship I had a stint at the software development corner for the revenue management dept. of a certain airline and I remember you'd have webapps running in C# or whatever. To get flight information from SABRE, you'd connect to a remote terminal session and parse out the output. I made a really nice state machine for parsing that output. Obviously there was no docs so I had to figure out what meant what from looking at it and asking people what they knew things meant.

Also IIRC these sessions were limited per department, so you'd need to contact a Java web service that handed out session tickets from a department-wide pool. The joke was even if you properly released the ticket the Java application didn't actually have a way to actually end the terminal session so even if the ticket returned to the pool if you did things too fast the sessions would be exhausted (or rather you'd get an invalid session) until they timed out on their own and SABRE gave fresh ones and then the whole department's apps would start having trouble.

I discovered this after I tried to optimize an application that needed to request flight route data for dozens of different routes but each session could only do one at a time so it'd take like 30 minutes for the thing to complete. I did a simple parallelization and it worked great until it didn't because the sessions ran out because releasing them didn't do poo poo. Oops. IIRC the solution was to just tune it down until I stopped using up terminal sessions faster than they timed out. I think the magic number was 3 or such.

What's double funny is that somebody had already tried that but a bug in some XSD meant C# XML deserialization didn't work so somebody patched it to use a static debug variable to get the raw XML and parse it manually. Obviously this meant locking the requests so despite being ostensibly parallel it still ended up doing everything serially. Fixing that involved getting the generated source for the deserializer, stepping until I found the offending line, discovering that Sabre was providing a flawed XSD, raging at them for a while, then installing a transformation thingy to add the missing namespace to the incoming XML.

Anyways airline core systems might be solid but everything else is sticks and gum. A classmate that also interned there made a script to reboot a server every 30min because it'd crash otherwise.

Kyte
Nov 19, 2013

Never quacked for this

Roadie posted:

I find it bizarre that there are people in this thread who think the only options are "give all your data to a US bigcorp" or "don't have computers". That Danish school system would have been fine with the "one guy running an Active Directory system with a single server rack" solution that my bumfuck nowhere high school had in the pre-Google days.

"One guy running AD on a single server" doesn't cover giving hardware to students, though.
And I imagine selling your soul to google comes with additional e-learning perks like collaboration platforms and whatnot.

Kyte
Nov 19, 2013

Never quacked for this
What's wrong with those?

Kyte
Nov 19, 2013

Never quacked for this
Ah ok so it's just a bad fit for the use case. Cool, for a minute I was worried there was something terrible about Pis I'd missed.

Kyte
Nov 19, 2013

Never quacked for this

Motronic posted:

Of course it was. If it wasn't why did they open the window? Why did they not drive away?

followup1 posted:

For everyone asking why I didn't drive off, I DID the part in the last video was right here. Where the police were finally here.
She shows the cops on the first followup

followup2 posted:

the dog followed my car yall. Biting along the way. I wanted to run over the dog so many times. I really think I did when he screamed but I really was trying to keep him away from the kids coming home from school. And no I did not want to kill the dog, I was in my car safe. I just cannot belive him eating my car
The second followup shows her maneuvering around

followup3 posted:

The owner was bit. The attack started outside because he didn't have the collar on. I was able to run to my car with the dog and close the door but he got my window before I could get it up. He bit her and she ran inside.
The windows started open. Warning: injuries here.
I mean it could be faked but I don't see why. Pitbulls are not safe dogs.
Either way dog was put down by their owner.

Adbot
ADBOT LOVES YOU

Kyte
Nov 19, 2013

Never quacked for this
Dwarf Fortress has demonstrated it's entirely possible to make a game that's 100% procedurally generated, but each step of that procgen was specifically tailored to its task. There's no need nor benefit to adding AI to it.

Maybe for character portraits and similar assets, but the cost/benefit is hilariously bad.

  • 1
  • 2
  • 3
  • 4
  • 5
  • Post
  • Reply