Some helpful advice for all you software devs:

- You cannot solve social problems with code

- You cannot improve a system until you understand why it works the way it does

- You cannot produce tools for someone else without understanding how they work and what they are trying to accomplish

(these are all actually the same advice)

@tess old trope and factually dubious. Code is just an implementation of a set of processes, and processes can solve social problems. The fact you solve more social problems by removing than adding software is more about the fact people working in software and systems ought to be qualified in systems practice.
If you think software can't help solve social problems then remove all the software from the water supply system and see how long society lasts.

@tess Also you can improve a system without really understanding how it works. You do it every time you read a book and learn something. Indeed Ashby's law says you can never understand yourself - yet self improvement is possible

You can design effective systems when you don't know how people will use them It's harder and requires expertise. However the statement is mostly a misunderstanding. Systems evolve and the software design, implement, deploy, run away model is what is broken

@tess Actually bullet two is the key of learning and understanding software and becoming a upstream contributor in most of major open source projects :-) You improve the system patch at a time until you understand why it works the way it does.

Thus, 180 degrees disagreement on bullet 2.

@tess

Some helpful advice for all you software devs:

- you can cause social problems with code
- you can degrade the system if you don't understand how or why it works
- you can produce obstructions for someone else without understanding how they work or what they are doing.

@tess Wow, the number of mediocre reply guys thinking they're smart being contrarian... 🤦

(1/?)

As for this post, the first part is a bunch of strawmen slaughtering.

@etchedpixels
> you can improve a system without really understanding how it works

What @tess said was;

> You cannot improve a system until you understand *why* it works the way it does

(Emphasis mine)

Understanding how and why are two very different things on dancing with systems.

@strypey @tess I possibly should have used the word "why". However I will simply note that the statement I made is equally true if you substitute the word why for how.

We don't understand how or why our brains work in full (and Ashby says we probably never can), but we still know how to change them by observation of what happens when we change the inputs and outputs.

(2/2)

> You can design effective systems when you don't know how people will use them It's harder and requires expertise

This is hubris. Comparing years of watefall vs (genuine) agile practice strongly suggests otherwise. If you are a representative sample of the software user yourself, then you know how people will use it. Otherwise, best to talk to people who are, right from the start of the design process.

(1/?)

@etchedpixels
> Code is just an implementation of a set of processes

True. But beware the 'all fish are trout' fallacy. Just because some processes can be expressed in code, that doesn't mean any process can.

> processes can solve social problems

Not the kind you can automate, or set and forget. Which is the kind you can turn into software.

@tess

@strypey @tess I don't follow your reasoning here at all.

Code can implement a subset of all processes (the computable subset, in reality the computable in reasonable space/time subset).

Systems based upon code can and do implement things that change over time and can evolve. The "fire and forget" model is just broken software design as I said earlier. The code that solved the "we can't scale with telephone operators" problem is not the code we use today nor could it be. Systems evolve.

(2/2)

> If ... software can't help solve social problems then remove all the software from the water supply system

That's definitely at risk of an 'all fish are trout. Just because removal of software creates certain social problem, that doesn't mean all software removal does. As you say yourself, the opposite is true at least as often.

More importantly, removal of infrastructure software creating certain social problem isn't actually evidence that software can solve social problems anyway.

@strypey If you think water supply was not historically a social problem then I am not sure how to have a rational debate with you.