Re: Excessively energy-consuming software considered malware?

[Liliana Marie Prikler <liliana.prikler@gmail.com>]

Hi Maxime,

Am Sonntag, dem 20.02.2022 um 11:05 +0100 schrieb Maxime Devos:
> [CC'ing some people in Guix I know to be interested in cryptocurrency]
> 
> Hi,
> 
> Guix packages some cryptocurrency(*) software (bitcoin, monero, some
> people have been working on packaging ethereum).  So far, it only
> appeared that clients are being packaged.
> 
> More recently, a ‘miner’ for monero has been packaged
> (https://issues.guix.gnu.org/54068).  At least for bitcoin, mining is
> known to consume an absurd amount of energy (the footprint of a whole
> country, and 1 Bitcoin transaction is said to be equivalent to 735121
> Visa transactions)[1].
> 
> Guix has a policy against including malware[citation needed 2], and
> furthering global warming[3] (and energy prices[4], if [3] is not bad
> enough for you) seems rather bad behaviour to me.
> 
> Would these miners be considered malware in Guix?
> 
> TBC I'm not making a case for rejecting all inefficient software, only
> software that is absurdly inefficient by design -- a, say, math
> library not using vectorised operations might be quite a bit less
> inefficient than a math library using vectorised operations, but that
> can be resolved with some programming work and it would seem to pale in
> contrast to the mining situation.
I don't think there's a case that can be made from the FSF's point of
view against wasteful software if the waste is intentional (which is
sadly part of the point of cryptocoins).  

To make my point in a more accessible manner, `guix show stress' yields
(as expected)
--8<---------------cut here---------------start------------->8---
name: stress
version: 1.0.5
outputs: out
systems: x86_64-linux i686-linux
dependencies: autoconf@2.69 automake@1.16.3
location: gnu/packages/admin.scm:2214:2
homepage: https://packages.debian.org/sid/stress
license: GPL 2+
synopsis: Impose load on and stress test a computer system  
description: Stress is a tool that imposes a configurable amount of
CPU, memory, I/O, or disk stress on a
+ POSIX-compliant operating system and reports any errors it detects.
+ 
+ Stress is not a benchmark.  It is a tool used by system
administrators to evaluate how well their systems will scale,
+ by kernel programmers to evaluate perceived performance
characteristics, and by systems programmers to expose the
+ classes of bugs which only or more frequently manifest themselves
when the system is under heavy load.
relevance: 29
--8<---------------cut here---------------end--------------->8---

Cheers