>an engineer who is designing a car makes a mistake and one

http://www.independent.co.uk/life-style/gadgets-and-tech/news/man-accidentally-deletes-his-entire-company-with-one-line-of-bad-code-a6984256.html

>>55947974
Do you want good enough software with fairly quick fixes when bugs happen or do you want to pay for the highest level of unit and interaction testing that takes 100% more time to handover?

File: REEEE.png (873KB, 910x993px) Image search: [Google]

873KB, 910x993px

>>55948015
>this fucking site

>>55948117

>"""study"""
>""""""""""suggests""""""""""

I'm trying to work out if they could possibly be any more vague.

>>55947974
>Really makes you think.

What really makes me think is why aren't mods permabanning meme spamming underage fucktards like you.

>>55948015

>I was only pretending to be retarded hahahahahahaha

>>55947974
What about a DRM that bluescreen computers?

http://www.xtremesystems.org/forums/showthread.php?276698-A-fix-for-Valve-CEG-games-causing-BSOD-on-Bulldozer-issue/page4&

At first someone made a soft fix that had to be applied on every restart.

The permanent fix was to update your motherboard bios.

Valve should have gotten more shit for shit.

Reminds me of this.

http://www.theregister.co.uk/2015/01/17/scary_code_of_the_week_steam_cleans_linux_pcs/

>people who make things follow a specific, known set of procedures to make something that should be fucking foolproof
>they fuck up and they deservedly get fucked

>programmers CREATE A FUCKING NEW UNKNOWN GODDAMN THING FROM GODDAMN NOTHING EVERY SINGLE GODDAMN TIME
>mistakes are made and a good faith effort is done to fix it and everyone understands

Why are people so fucking stupid when it comes to actual technology knowledge?

>>55948157
well you could read the fucking paper in question, for one

Has a bug in an application ever killed someone?

I imagine if something like that did happen then someone will lose their job and more.

>>55948283
>AMD

try to mess up some space rocket code and see the little damage it does

>>55947974
>The company loses a shit ton of money in recalls, people won't buy its cars again.
Uh, no.
>an architect who is designing a building makes a mistake and the roof leaks
Sounds more like the contractor's problem.

>Really makes you think.

Only one logical response:

>stupid fucking people will be the death of us all

>>55949047
Google Therac-25

>>55948973
The fact that you think that it's a "known set of procedures to make something that should be fucking foolproof" literally proves you're a retard that doesn't know what he's talking about. Engineers make NEW things using existing technology.

Programmers(computer science graduates, not code monkeys) have no idea of the real world. I've worked together with quite a few and they're so fucking self-absorbed and arrogant. They expect everything to be perfect and deterministic.

>>55947974
>an architect who is designing a building makes a mistake and the roof leaks
No. If your building leaks water more people than just the architect fucked up.

>Program fucks up, blame the programmer.
Oh I see why you over simplified everything, otherwise youre memes wouldn't be epic.

>>55947974
>one person is responsible for everything
this kills your entire argument

also, pinpointing whom exactly to blame for a fuck-up of any kind is often very difficult, precisely due to nothing being a one-man-job

>>55947974
>an engineer who is designing a car makes a mistake and one part is prone to breaking
Wew, glad I got a refund on that car

>an architect who is designing a building makes a mistake and the roof leaks
Wew, the construction company sure fucked up, I won't buy this house


>a programmer makes a mistake and introduces a bug in an update or releases a program riddled with flaws, making the life of his clients miserable
Wew, this is a shitty program, I better uninstall it and look for a new one

>>55949047
Plenty of times.

>>55948015
>"""""""code""""""""
>actually a command

WEW

>>55947974

1. With large enough programs, it is virtually impossible to remove all bugs before release.
2. Programs aren't physical objects, and are significantly more malleable. There is an expectation for them to be regularly re-engineered over time. If it is not to fix bugs, it will be to add more features and improve performance. Of course, this will likely introduce new bugs.
3. Buggy software is often significantly less likely to induce serious damage to people. Where it is capable of causing damage, such as when used in nuclear power plants and what not, we might see lawsuits if the software was not thoroughly tested.

>>55949003

Someone already went over the paper and found numerous flaws in it, and also found out that the various media sites that were reporting on it were doing so without checking the fact that it wasn't fucking peer reviewed.

http://slatestarcodex.com/2016/02/12/before-you-get-too-excited-about-that-github-study/

>>55949739
>an architect who is designing a building makes a mistake and the roof leaks
>Wew, the construction company sure fucked up, I won't buy this house

Architects have no concept of maintenance.

>>55948117
It's talking about traps, we all know traps make the best programmers :3

>>55948015
it was a hoax scrub

>>55947974
Code generally won't kill people.

You fuck up a line in a truly critical system and you will have a bad time.

Anyways, the equivalent of what you are talking about is an SLA. When your company buys software from, they must be diligent in the kind of craftsmanship they expect and the penalties for any flaws.

However, from the consumer end, America has favored the producer far more than the consumer; the American market does not bother to place almost any consumer-favorable contractual agreements in paid software licenses.

It is your responsibility as a consumer and as a consumer market to not accept such bad agreements.

Unfortunately, the market as a whole is going the other way; expect the examples you listed to become more like the consequence-free scenario that you are complaining about.

>>55949236
A large source of software bugs come from the considerable dynamicism and radically changing requirements during development.

When you have a product that can take the traditional waterfall method and has the leeway to demand more time to get it right (in the same way that construction can go over budget and over schedule) then you can produce more robust software.

The problem is that it is not profitable and is very hard to estimate. Software a few months or a few weeks late can be useless.

>>55949047
https://en.wikipedia.org/wiki/Therac-25

for those lazy enough not to search for it like the other anon said

>>55947974

Let me put this into better context for you:

>and engineer who is designing a car does it correctly. Ships car to customer. Turns out customer is in Australia, driving on the opposite side of the road, and is using the compact car to tow a 20' boat. Also, customer expects to take the car off sweet jumps.

>an architect who is designing a building runs into problems when he realizes that the location is under an insane number of conflicting building codes which restrict his choices. Ultimately, the project fails because all you can provide in that city (platform) is a literal cube.

So yeah. Software is hard to get right, particularly on the sort of development timeline that the modern world expects.

If you want to make ANY kind of comparison to other forms of engineering and construction, you have to look at very rigorous areas of programming, like medical radiation machines, nuclear bombs, dam control systems, etc.

>>55948117
Just gonna put it out there that most of the people coding anything are men. Most women go into shit like gender studies because they're retards who are afraid of math. There are some extremely intelligent women out there, as rare as they may be. If you compare the work from 5 of them to the work of 1000 /g/tards it will probably be significantly better.

TLDR there are a few smart women who do a good job and the study doesn't really prove shit

>>55948973
>inventing a new building and new car with new design, material and features
>not making. Something new

Crap, you're retarded. Difference is since the designers are responsible they try to acctually make quality things.

>>55951628

He's not retarded, he's correct.

Structures, in particular, are mostly built according to a well established system. It's a BIG FUCKING DEAL when someone wants to try a new method or structural element in a building.

Here's an example:

http://www.willamettelive.com/2010/news/courthouse-square-falling-through-the-cracks/

IIRC, an underground parking structure was built. They used a new composite construction system that used steel *cables* embedded in concrete, rather than the usual rebar.

Well, one day there was a loud BANG. After awhile, they realized that the structure was failing.

The head engineer had skipped several critical things that had to be done to implement such a construction method properly.

But if you're building a regular stick home around here, wham-bam-thank-you-mam you're done in under a month, following what is essentially an algorithm of building codes, generally turned into modules, assembled together rapidly. They're pretty failure-proof.