[ art / civ / cult / cyb / diy / drg / feels / layer / lit / λ / q / r / sci / sec / tech / w / zzz ] archive provided by lainchan.jp

lainchan archive - /tech/ - 21



File: 1492104039537.png (18.34 KB, 250x250, 1468090273610.jpg)

No.21

We don't yet have any /tech/ board but I want to talk in lainchan about this.

After just a few years inmersed into programming and related stuff, I've come to deeply detest the whole state of the computer and programming industries in general.
To begin with, the Web. It is like kicking dead whales down the beach, web browsers are bloated, vulnerable, and constantly leaking memory (and possibly your data). Yet it's the main medium of communication available. If you want some communication with the world, including the access to a lot of useful information, we are trapped with this nightmare called the Web. Let's add to this a thing called massive user espionage and how hard it is to trust that your communications have barely reasonable privacy.
Then comes the insane complexity found in operating systems and hardware. A horrible architecture full of back doors, operating systems that can barely stand and have a soykafload of moving parts, all from independent vendors, and a bunch of proprietary drivers that further complicate matters.
Even the best case scenario implies using something like GNU on a x86 processor, usually with X and possibly even firefox or some of it's zillion clones that do not make things any better. (either that or something like webkit, lain forbid). All the time while on the lookup for another heartbleed.

I get more and more frustrated every day with the whole state of computers that every step of the way I just want to give it all up and forget about it. It's really a race to the bottom.
I have seen some people talk about isolating from the whole world, make their own architectures, and make their own operating systems. It really doesn't sound like a crazy idea anymore.
Right now I'm using Fedora for reasons, and I keep wondering how it is possible for software to suck so badly and keep pulling pranks on me all the time and be so fuarrrking sluggish. And the problem is, if I wanted to use my favorite system, I would be limited in some ways because proprietary drivers, GNU, etcetera. And if I wanted to have a decent setup, I'd be largely unable to use the Web at all because javascript and terrible formatting. I hate this, and often it makes me want to just drop it all and live as far as possible from modern technology.
It feels as if everything was being made soykaftier and soykaftier on purpose, just for the sake of annoying people because now everything is a nasty hack waiting to break. It really, really is like kicking dead whales down the beach.

I wonder how other lainons feel about this, it's hard to have a balance with usability where the whole ecosystem is a terrible hack just waiting to blow up.

  No.22

my thoughts exactly
that's why I'm giving up CS and going for theoretical mathematics or physics
f​u᠎​ck thi᠎s s᠎hi​t

  No.23

>>22
Coward. I'm in CS to make things better. I plan on going heavily into programming language design and implementation, and then operating systems, so I can get a picture of what exactly needs to be done to overhaul this soykaf.
I've been thinking about building a toy OS to see what general computing could look like beyond what we have now, but I don't think I'm quite skilled enough to execute it.

>>21
>the whole ecosystem is a terrible hack just waiting to blow up.
It's true. This much complexity isn't sustainable. We're never going to be able to teach everyone all of the horrific intricacies of so many broken systems. The question is, what will happen when it does "blow up"?

  No.24

I think about it almost every day. Security is porous-to-transparent because of complexity. It's unsustainable, and it's clear to me that in the next decade there's going to be a series of events that has us re-evaluate how we build systems.

I think we'll look back and see this time as one of the biggest bubbles ever: "The time when all software was trash, the time when everyone put everything on the web, and the time when companies didn't think about security until after the fact."

I think there's going to be a massive series of breaches/hacks/outages (maybe state-to-state) that will be sustained enough to shake the whole population's confidence in how they interact with the net. I'm not talking just getting off facebook, but people afraid to use ATMs, email, text, etc. It'll cascade and crash economies.

At that point gov't will step in and start holding companies more responsible in an effort to try to restore trust in systems, but that will initially just aggravate the crash, as it scares more companies out of business.

I'm probably just being paranoid. But the number of vulnerabilities out there... and the stakes... and the actors... it has to break at some point. Hopefully it won't be too terrible.

  No.25

File: 1492147792638.png (74.5 KB, 200x121, tpeye_01_small.jpg)

>>24
>I think there's going to be a massive series of breaches/hacks/outages (maybe state-to-state) that will be sustained enough to shake the whole population's confidence in how they interact with the net. I'm not talking just getting off facebook, but people afraid to use ATMs, email, text, etc. It'll cascade and crash economies.
There is a sci fi comic with a plot like that. "The cloud bursted" and the internet was taken down completely but the impact it left on society was so big that secret identities became a normal thing. The story follows a private detective/paparazzo who gets tangle between two terrorist organizations with clashing plans.
Pic related is what I'm talking about, if you're a weeb who has never read a western comic this one will surprise you.

  No.26

A part of me would like to believe that things will eventually get better. It's just that the ecosystem of software engineering seems to be too demanding for the average programmer.
Some people do it for the money, so they end up producing a hack of a final product and wait for their paycheck.
Other people love their jobs, but spend way too much time on it. They either get burned out and leave the industry, or their work ethic suffers.
Full disclosure: I'm still at uni, so I've really only heard these stories second/third hand, or I've witnessed soykafty work ethic in my fellow students (and myself ;_;)
I've taken the approach of informing people of how broken/insecure our internet and software is now, so that when soykaf hits the fan, I know none of my friends wrote the code that threw it.
I feel that the entire ecosystem of producing software will need to change before we start seeing change in the quality and security of software.
Most of my thoughts sprouted from this article: https://medium.com/message/everything-is-broken-81e5f33a24e1

  No.27

>>25
Working for a government is impossible to get people to realize the risks inherent with the cloud.

You warn and warn and warn, but they see performance improvements and cheaper operating costs as major pluses to good to miss.

Even though the advantages are not parented. Imagine government departments putting their ENTIRE financial systems in the Amazon cloud... its happening already. Imagine for a moment if that leaked, the databases leaked and you saw every cent they spent along side employee details.

Imagine the clusterfuck, they have the budget for it, they can afford to run onsite servers in secure rooms but they lack the expertise because their employees are fuarrrking stupid grads who have only working in the government and have Degrees that have nothing to do with ICT. I know people who are financial grads working in network security making some very sketchy decisions, but its ok, they have a degree!

We deserve this, the governments have played like twats for too long.

  No.28

Within the next quater of a century or so, I predict that there will most likely be some sort of massive system breakdown on a global scale. With the increasing globalisation of tech companies, and the faults in existing and possible future software and security architecture, some sort of hack, bug or an (as of yet) unknown threat will bring the system down to its knees. Seeing as companies and buisnesses and now governments are relying on a small number of companies, ie the giant megacorps such as Google, Microsoft and Facebook (which a lot of small buisneses almost entirely rely on), it is not hard to imagine the economical and societal devistation that such a massive crash could bring.

Its a matter of time lainons...

  No.29

>>23
Start by redesigning the Von Neumann architecture.

  No.30

>>27
soykaf, The Private Eye is more relevant now than ever, I should storytime it.

  No.31

>>30
yeah, if anyone has a .pdf of this, it looks very interesting.

  No.32

The web isn't inherently bad. It's javascript that's bad.
The web is really great once you learn how much metadata and semantic stuff you can cram into XHTML and HTML. A world readable by humans and computers. A world of knowledge.
I agree though that proprietary hardware (and thus proprietary drivers) are bad. Software is only as free as the hardware it runs on.
>Fedora
Try OpenBSD. I can't stand GNU/Linux, but the BSDs are all much more enjoyably Unix. But even Unix is something we wish would've been replaced with Plan 9 or something totally new some day.
You mention the web again: it's bad creators who are bad. Creators who can't be assed to make a compliant AND accessible site. Would you blame C or Lisp for a bad program? No, you'd blame the person who wrote it.
Javascript, though, I wouldn't mind someone saying is inherently bad. I think it's not every useful at all. I don't think CSS is useful either, but that's just my personal opinion of course.
>>22
>giving up CS
I wish I could afford to go to college. I want a PhD in linguistics or CS, and explore new worlds of math and understanding.
>>24
>be a series of events that has us re-evaluate how we build systems
No, I don't think this will happen. I think our world is more like the one in Brave New World or Fahrenheit 451. A world where bad things happen but everyone is already too happy and oblivious to notice.

  No.33

>>32
Yes, I was using fedora because of some circumstances that arised. I didn't say it, but my ideal setup is with OpenBSD. However, I'll be lacking wireless drivers there, so that's why I sticked with fedora, for the convenience. After I wrote this I wiped it with OpenBSD.
You know what's the problem? If I go ahead and install firefox in my OpenBSD box, then I might as well just stick to fedora, you know what I mean? Right now I'm posting from a Windows box because I don't see the point in ruining OpenBSD with horrible soykaf. I guess I could get lynx instead, though I don't really like it either because formatting and other things. But beggars can't be choosers I guess.

On the topic of HTML, there are things I don't like about it, part of it is that it just looks ugly and I'd rather have other markups (a lispy one for example) that look nicer. But I also think that HTML encourages an emphasis on superfluous things instead of the actual content of the site. That's one reason why lynx formats things so bad, because so much is formatted around menu bars at the top, tables all around, and so on, and the content gets obscured. Then if you go look at the source, it is a horrible mess, and you have to weed out analytics, banners, superfluous menus (things that look good when rendered but that take a lot of space when not rendered the way they're expected).

I think OpenBSD is the bestie of the stuff out there, and I'd be happy with it, but then again come some of my complaints with other stuff such as proprietary drivers that only make things hard for users, which is also the reason I can't install 9front on any of my computers.
And let's not get started with porting programs. I ranted this after looking at the source of GNU APL and seeing that besides heavily GNU idyosincratic, it is specifically made for debian (and some RPM thing, which I'm guessing is Red Hat). But again, beggars can't be choosers.

On the topic of college, there are a zillion books out there on the internet right now, if your goal is knowledge, you have no excuse.

  No.34

>>31
digital western comics don't come as pdf but as cbr or cbz.
It'll be more fun if I sorytime it here or in 4/co/ anyway. maybe this weekend.

  No.35

>>33
>wireless drivers
I understand your pain. I have an Nvidia video card since I built this computer back before I knew I'd ever use Unix. So I have no graphics, and I seem to keep using FreeBSD rather than OpenBSD solely for that reason, despite preferring OpenBSD.
>you know what I mean?
Yup.
>HTML encourages an emphasis on superfluous things instead of the actual content of the site
That's not true at all. In 4.01 strict you don't get much useless soykaf. But a lot of people, as I mentioned earlier, use 4.01 transitional or HTML5 or even fail to use anything properly at all.
But yes, many people don't know how to make a site readable as a document. That's what HTML files are: documents. It should be easy to read one from top to bottom, even in lynx. I lost my blog when I accidentally wiped my drive a few months ago, but it was an example of a good, readable website. I remade it very very quickly a few days ago, but it's barebones and not much of an example now.
>the source
Part of the problem is people automatically having sites generated for them. If people wrote websites by hand it'd probably mean more readable source: which, by the way, is a goal of HTML. Entirely human-readable source. A good website is one you can open in vi, ed, notepad, whatever, and understand, and mentally visualise.
I interestingly have have a problem with soykafty GNU code. I hate how they claim their code is for ``unix''. I want them to be honest and say that their code is for GNU/Linux only. I like portability, but honesty is even more important. mit-scheme doesn't compile on OpenBSD 6.1 and it used to in past versions but I can't find out what's going on, and I don't even think mit-scheme ever compiled with clang.

>On the topic of college, there are a zillion books out there on the internet right now, if your goal is knowledge, you have no excuse.

Correct. For knowledge itself, I have far fewer excuses.

  No.36

I'm with you, I loved computers and programming for years but gave it all up and my life ceased to be the nightmare that's romanticised as "student life", that's as far away from life as you can get. Rethink it all, I assure you it's worth it. Let's not fight this hard for an imposed need. <3

>It feels as if everything was being made soykaftier and soykaftier on purpose

It is. That's where the money is at, how would corporations make that much money in a secure, elegant, private, respectful, hackable system?

>>22
I'm with you too. Maths and Physics will never betray us. College though, it puts an awful coat of industry market over anything.

>>23
Even if you want to make things better, you should leave CS and study on your own or with a group you make. College teaches you how to work, you won't learn what you're after.
>what will happen when it does "blow up"?
My guess is single-corporation corporate internet, Foogle is already working on that.

>>33
>On the topic of college, there are a zillion books out there on the internet right now, if your goal is knowledge, you have no excuse.
I would go with libraries and people with the same interest, I think online books are overrated, unless you're getting into weird subjects or live in some relatively isolated place.

  No.37

As far as the Web goes, I think the main problems stem from the rapid commercialization of the Internet.
This probably dates back to the dot-com era, but it especially holds true for Web2.0.
Companies are racing to deploy their services, often over-relying on unnecessary frameworks and overly complicated development stacks. The main priority is how quickly a service can go public. If efficiency were the top concern, most apps would be built from scratch (or from simplified modules) using only what is crucial. These fast-paced production cycles have led to almost the entire Web ecosystem relying on code from a few tech giants.
Even services that build personal websites are bloated and poorly implemented.
The commercial possibilities of the Web are driving design philosophies faster than the protocols can manage.

Reliability has been traded in for convenience.
HTTP has been stretched beyond its original intention just to add new functionality to applications.
Also consider the IoT and how easily smart devices have been exploited.

There is little emphasis on secure, simplified code. Instead everything is outsourced to libraries of which developers have limited understanding.

>>34
I'd been interested to see it, seems cool.

  No.38

>>31

http://panelsyndicate.com/comics/tpeye

You can download it from the authors' page in pdf, cbr and cbz formats.

  No.40

>>21
tbh i'm kinda sick of people always making treads with this, whining about how soykaf everyting is but not working on solutions. most of you settle with installing the penguin flavour of the month and jerking about "muh, open source is the only path", not seeing they aren't helping anyone.

  No.41

>>40
>not seeing they aren't helping anyone
They're probably helping themselves, which is already one more person that helping noone.

I do agree with the sentiment of you post, though.
Our environment might be soykaf, but at the very least it isn't fragmented.
Imagine if every goof on here was to build their own systems, geared only towards their tastes, from ground up.
None of those systems could use any existing software and we'd be set back by about 50 years in progress.
That's not really worth it.

  No.42

>>23
>I'm in CS to make things better.

Sorry to say it, but, you'll not change the world.
It's too much work for one people.
The best you can do is find a way to audit/verify the trustfulness of hardware (check to hardware trojans). That's one critical thing that has very little studies.
Or, contribute to things like formal verification, dependent type languages, LowRISC, open source firmwares (coreboot), secure systems (seL4).
It evolves so many areas of study that would take too much time to change something.

  No.43

>>41
my main problem with OP's (and others like him) mindset is, that they SEE something is fuarrrked, but instead of starting to work on a deep level solution (like the founding of the various european pirate partys 11 years ago, before they got killed off by sjw activists), they rather discuss wich band-aid called linux or bsd they patch on their harddrives. they might help themselves on a surface level, at least since vault 7 it should be clear that now is the time for a movement, a political solution than to the same old penguin dance.

  No.44

>>43
>they might help themselves on a surface level, at least since vault 7 it should be clear that now is the time for a movement, a political solution than to the same old penguin dance.
I don't quite see how you made the mental jump from complaining about the hardware/software ecosystem to calling for revolution.

  No.45

>>44
because seeking a solution only on a hardware/software layer is on the long term masturbation.

  No.46

Write a web-browser for plan9.
This is the first step towards a better future.

  No.47

>>44
Actually, he's right. Capitalism will always promote bad hardware and bad software. Another poster said:
>It is. That's where the money is at, how would corporations make that much money in a secure, elegant, private, respectful, hackable system?
It's inherent to our way of life to make bad things, and as long as this is our life, bad things will be the norm. Better to pull out weeds by the root, rather than trimming their leaves.

  No.48

>It is. That's where the money is at, how would corporations make that much money in a secure, elegant, private, respectful, hackable system?
But that's just what people want.
99% of the users couldn't give a soykaf about all those attributes and wouldn't be willing to spend the extra money that they'd realisticly cost.

  No.124

The answer to this and all questions about the world is get off the rock. fuarrrk this planet, if you want to start over you have to get off this soykaf hole and build from the bottom up. You will necessarily have to use hacked and soykafty hardware/software to do it, but after you leave you can try to make a better system. Building the perfect system in a vacuum won't solve the underlying problems.

  No.140

>>37
This. Hype drives bubbles and confidence in high risks investments. The more risk, the less you'll want to initially invest.

The tech economy is a high-octane race to build the cheapest simplest products that simultaneously have massive reach and profitability. When infrastructure holes, sec, and dependency issues appear no one will want to fix them because it's a huge "sunk cost" on a few cheap experimental products. It's not the error of capitalism, but the obsession on ends without thinking about the means.

>>45
Long term is enough, as long as you leave a note for those in the future to eventually fix it again.

  No.145

>>48
That's what YOU want (me too!). But unfortunately most people don't give a rat's arse about "hackable" systems, or secure/private, or any of the above. Outside of the board and other boards and small tech communities, the numbers don't even come close in comparison to the amount of people camping out to buy the new iPhone.