blog/content/links/20200304.md

+++
title = "Links for 2020-03-04"
date = 2020-03-04

[taxonomies]
tags = ["links"]
+++

Interesting links for 2020-03-04.

<!-- more -->

# [Leaked Document Shows How Big Companies Buy Credit Card Data on Millions of Americans](https://www.vice.com/en_us/article/jged4x/envestnet-yodlee-credit-card-bank-data-not-anonymous)

Alright, this is stupid: "We removed the user information and put an unique ID
in place" is completely bullshit. I know it, you know it, but it seems it is
the general way of selling user information around.

Even if someone removes the unique ID but keep other information, things are
still traceable: For example, how many people do you think are white, male,
above 40, not married, living in a specific suburb (I won't even claim the
street itself) that bought some product? I bet you could pretty much find me
with just that, 'cause that description would fit 0.00001% of the city
population, anyway.

Maybe we need to rethink this "sell information" thing, since privacy is long
gone. We need laws that, if you want to sell collected information, you need
to put one single vector per information. That means that, if you want to sell
that information, you need to have one dataset for skin color, which says a
white person bought some product; another dataset for sex, which says a male
bought some product; and so on. No unique IDs.

It's not perfect, but I believe that would get more traction than simply say
"don't capture any data", 'cause, seriously, it is not working.

# [Aerogel from fruit biowaste produces ultracapacitors with high energy density and stability](https://www.sciencedirect.com/science/article/pii/S2352152X19309077)

Hey, that's cool that we are finding ways to recycle stuff like biowaste for
something like capacitors, but I have to wonder if that, in our current world,
that means people will produce durian and jackfruit just to let it rot to
create biowaste for the capacitors, instead of something simple, like feed
people.

# [IBM To Transition Their z/OS, POWER + AIX Compilers To Being LLVM/Clang-Based](https://www.phoronix.com/scan.php?page=news_item&px=IBM-Will-Use-LLVM-Clang-Stack)

I have to wonder if the move is something like this: IBM bought RedHat; RedHat
sells Linux solutions; IBM stronghold are the mainframes, which are dying; by
changing the base system from their compiler to Clang, they are also saying
"hey, people who build stuff for our platforms, you should use Clang now!",
which would make transition those codebases to Linux a bit simpler.

But yeah, just wondering if that makes sense.

(Also, Clang _Fortran_?)

# [Microservices – Combinatorial Explosion of Versions](https://worklifenotes.com/2020/03/04/microservices-combinatorial-explosion-of-versions/)

I have a strong problem with this post: Basically, the problem it points
happens if you keep two different versions of the _same microservice_ running.

That's not how you should build your fleet -- even if it is possible and not
that hard.

One point of upgrading a microservice is keeping backwards compatibility with
their inputs; if you deploy a new version what receives a complete different
input, you can either add both input interfaces in the new service or just add
a brand new microservice that just converts the old input to the new one --
which increases the fleet, but it's not that a big point.

One thing to keep in mind with this upgrade policy is that you _need to
monitor the inputs_. Once the old input version is not being send anywhere
else, you can remove the input/disable the microservice -- with the second
option being less intrusive, as you won't need to redeploy the updated
microservice with the old input support.

# [Stop using Material Design text fields!](https://www.matsuko.ca/blog/stop-using-material-design-text-fields/)

Or you could, for instance, to drop Material Design completely.

Yes, I'm not a fan of Material Design. I'm not UX expert, but I feel that it
fails in a lot of places -- not only text fields.

# [The Self-Attribution Fallacy](https://www.monbiot.com/2011/11/07/the-self-attribution-fallacy/)

"Intelligence? Talent? No, the ultra-rich got to where they are through luck
and brutality."

"If wealth was the inevitable result of hard work and enterprise, every woman
in Africa would be a millionaire."

I keep saying this around, and one thing that resonates with what I say would
be: It's not that the ultra-rich are brutal, brutality is what made them rich.

# [Void: terminal-based personal organizer](https://github.com/void-rs/void)

Not sure if it falls into any of my workflows, but HOLY COW, graphs in the
console?

# [Building Rust for Multiple Platforms Using Github Actions](https://medium.com/@jondot/building-rust-on-multiple-platforms-using-github-6f3e6f8b8458)

While focused on Rust, because the focus are the Github Actions, one could
"easily" port for other languages (for different levels of "easy").

# [Web crawler in Rust](https://rolisz.ro/2020/03/01/web-crawler-in-rust/)

"How to Draw an Owl", but for Rust crawler. :)

I got most of the stuff, but I guess anyone learning Rust would suddenly see
the later listing and think "This is nuts!"

# [The Cost of Indirection](https://www.joshmcguigan.com/blog/cost-of-indirection-rust/)

"How a newbie, a seasoned and a veteran Rust coder solve this problem", but...
for Rust.

I'm not that interested in the times, but in the different implementations of
the same thing in the same language. May be interesting 'cause some
implementations may make more sense than other to some people.

# [Toss a Coin to Your Bitcher](https://uncannymagazine.com/article/toss-a-coin-to-your-bitcher/)

A disabled person commenting on the way "The Witcher" series deal with
disabled people, in this case, Yennefer. Honestly, even when I watched, I
couldn't see Yennefer as disabled, 'cause... well, one 'cause although I think
about my own disability (glasses), I don't see that much different than any
other "normal" person (although I still think Gordon Freeman is the _only_
short-sighted "hero" I ever saw), and I know what happens with Yennefer in the
future (thanks for the games).

Still, it's eye-opening to see this kind of post to get a view of how disabled
people see disabled characters.