Hi Florian.

I'm happy to engage in this discussion with you, but one thing to set up straight, you keep referring to this as "your package", it's not mine, I did not create it nor am I a contributor.

Don't shoot the messenger :)

Regardless, you stated that "In software engineering there isn't at risk vs. benefit situation. Either your code is safe and respect the standard either it doesn't." and this statement is just false.

Risk vs. benefits is the spine and bones of any field of engineering, and this type of false dichotomy that either your code is "good" or "not good" will lead astray.

Any thing has pros and cons, any new line of code can lead to unintended consequences, every bug that you solve can create new ones, every technology that you add might has some vulnerabilities (log4j anyone?), every commercial solution you use in your organization might go bankrupt next year.

You always evaluate risk vs benefits with the information you have available.

If you think that the risk of compromised readability out way the benefits of using pointers, that's a legitimate concern, but it should be contextual and be evaluated per project.

especially when the alternative is to use a low level language which is again part of the false dichotomy mentality of "all or nothing".

PS - it seems like you believe that pointers are some how suppose to enhance performance, and this is a strategic error, pointers simply allow you to use addresses instead of the actual values, and they have nothing to do with speeding up your software.