Memory safety does not matter

monkyyy crazymonkyyy at gmail.com
Sun Jul 28 19:54:44 UTC 2024


* 1000 messages in a thread this week

* the cia pushing rust several months ago(making the news every 
day since)

* @live

* 2 decades of debating about the d's gc

* arguably nan init

Allot of time and attention has been going towards an unsolvable 
and irreverent problem. With imaginary nonsense solutions like 
rusts bower checker promising it will fix everything.

There *cant* be a complete solution, if you could decide where a 
pointer could point with 100% accuracy why not decide the halting 
problem? While there trivial partial solutions like... *drum 
roll* slices, note slices are data structure and not "code 
coloring" and deal with indexing.

If you want safer code, *provide nice to use data structures* 
that provide clever, *partial* solutions on opIndex. Not @safe vs 
unsafe vs @live vs whatever code coloring hell.

I *do not care* and will never lift a finger to use safe code, 
and if you quickly survey what people *do* and not *say*, we are 
no where near mathematical perfect safety, instead my passwords 
are trivial to find online, people rarely care about strong 
password, users are a weak point, and credit card numbers end up 
as plain text by fortune 500 companies paying thru the nose for 
fake security.

99.999% of physical locks are trivial to pick, and your life is 
harder to back up and arguably more important then your data; yet 
society functions and moves on. Why is that, if I "need" a 16 
character password and 3 factors of authentication to protect my 
10$ on paypal?

Please dont spend any more time on code coloring.


More information about the Digitalmars-d mailing list