mihvoi

Indeed, one should choose the the lowest consistency guaranties that are required - to favor scalability and availability.

Some more thoughts on minimum requirements for a read-after-write consistency: https://software.mihvoi.ro/2026/04/2-flow-consistency-read-y...

An intuitive and simple Phone application/theme on Android smartphones - especially for old people.

The main thing that keeps me using Windows and not Linux is a tool to index folders and files from all my DVDs and external hard disks. It also accounts for total size of each folder and browse the archive.

I only use the search file/folder name feature, I don't even need any content indexing that this tool has. It just needs to scale to Terabytes size (like millions of files/folders) and optionally do de-duplication.

If you know such tool on Linux, it would really help me to be less dependent on Windows. Or maybe someone you can build such a tool on Linux. I tried to create my own tool in Java but it took too much DRAM to index the many files I have. This Win tool indexes all my files and folders in a file of 1-200MB and searches all in 10-30 seconds. No server/database in involved, just a standalone tool.

I share the tool just for reference. The tool didn't had any update for years, I hope the programmer is still alive and healthy. https://www.whereisit-soft.com/

Make multiple copies in various places, including read-only like CD/DVD. Re-copy at regular intervals.

Do not underestimate the resilience of paper format, however it's harder to move it to digital again.

More ideas that you could also use for programs: https://meaningofstuff.blogspot.com/2015/05/backup-your-smar...

I only "believe" in the computable numbers - that can be defined by an algorithm at arbitrary precision.

That includes all rationals along with irrational that we have a way to approximate by algorithms - like sqrt(2) and PI. It does not include the incalculable numbers like Omega.

Being linked to algorithms, such numbers are only countable many, so they cannot contain "most" of irrational and therefore they lack most of "Real numbers".

There are mathematicians that take this view seriously, like "Constructivists" and "finitists". For example: https://www.youtube.com/watch?v=REeaT2mWj6Y

This crypto madness kept way too long. I say it from 2017:

https://meaningofstuff.blogspot.com/2017/12/why-bitcoin-will...

Dogecoin was created as a joke. The founder removed himself from the project, other people are probably use it for scams. https://www.cnbc.com/2021/01/29/dogecoin-cryptocurrency-rise...

Regardin all crypto: https://meaningofstuff.blogspot.com/2017/12/why-bitcoin-will...

They printed a lot of fiat money without loaning it to people that can produce goods. The result is not enough loan pressure to "buy" the fiat papers over goods. On the good side, I guess they have some enforcement and a lot of violence :(

A limited supply coin does not guaranty a need to exchange goods over that particular coin and not another.

The good thing is that you can find more easily good people at low work cost. However, it might be hard to evaluate the value of the contribution of each "employee". What if some people throw unreasonable estimates?

One idea to overcome this is: you work by small increments of features. You put each feature on the table and people bid the time to complete it. You assign the task to the one with the best reasonable estimate and with good track on the delivery - depending on the risk of failure. People that don't offer reasonable estimates can be replaced. The thing is that you know that without understanding the code, because someone else will bid less time and do it.

A variant is per-project, "last feature is free". After each feature, you pay the previous delivered feature, based on hourly time counted by the programmer. When you find the bill to be overcharged, you just pay the previous feature and close the collaboration. You might guaranty to pay a reasonable but fixed sum, in advance or at the end, to reduce the risk of the programmer to not be paid after too few tasks.

People should also have some stock or bonuses by global performance, otherwise they would not be also team players. And more than this, they should share a minimum motivation to be proud of the global outcome, for example because of how their work is making the world a bit better. Money if not always enough for people to really want to make it work.

It's hard to get a programming job without having practical experience. It might be easier to apply to a Testing/QA job and gradually transition to programming when opportunities arrives (test automation -> small tools -> small features -> bigger features...)

nice. so true...

Using the "float" or "double" data types could lead to unexpected results if done without enough precautions. For example incrementing a "float" variable could leave the value unchanged, for relatively small values — like 20 millions. Do you think that these dangers are well understood by the average software developer?