Re:ATM NFC Hack
by aliasEli on Friday June 25, 2021 @04:40AM
(#27628420)
Yet another buffer overflow. Probably the software is written by some C/C++ developers who swear that those never occur in their code. Given the large number of security problems that have been caused by buffer overflows and similar problems, I find it very sad that any programs that need security are still written in C/C++.
[ Parent
| on HN ]
Please don’t skip the prove hypothesis part when drawing conclusions. It’s at best disingenuous.
C/C++ isn't the issue, it's the lack of not using memory-safe features like smart pointers which you'll need C++ for.
Re:ATM NFC Hack
by nathanaldensr on Saturday June 26, 2021 @07:52PM
(#27647123)
While you're technically correct, your ignoring the pragmatic truth: most developers are terrible and should never write code. Because of this reality, C/C++ and all its footguns are something to be avoided unless there is no other alternative (usually this happens when integrating with legacy software or in industries like gaming where the switching costs are extreme).
So you're saying that since we're encouraging people to flood the market with little to no skill, we should instead accommodate them with Baby's First Programming Language and then force all of the other developers to follow the same methodology?
This makes no sense. It just sounds like Go.
Not the OP but we should weed out all the incompetent folk ASAP rather than hide it to maintain delivery speed.
I completely agree.
terrible programmers ain't nothing new though and large corp in critical fields are aware of that and have (or had) structural mitigations (slow and heavy processes, regulations etc)
This is sort of true, but show me a (non-trivial) C++ program that doesn't resort to naked pointers or references at some point.
In the real world, smart pointers help a lot and are a great idea but they don't fully solve the problem in the same way that Rust or memory safe languages do.
[dead]
Re:ATM NFC Hack
by catlifeonmars on Saturday June 26, 2021 @08:37PM
(#27647392)
Pointers kill people
Re:ATM NFC Hack
by scintill76 on Sunday June 27, 2021 @02:38AM
(#27649178)
“We kill people based on memory safety.” - Michael Hayden, probably
Re:ATM NFC Hack
by gabereiser on Saturday June 26, 2021 @10:31PM
(#27647987)
null pointers kill people
Re:ATM NFC Hack
by ndesaulniers on Saturday June 26, 2021 @10:59PM
(#27648133)
Dangling pointers kill people.
Re:ATM NFC Hack
by kumarjsingh6 on Saturday June 26, 2021 @11:19PM
(#27648251)
redditizing HN kills people
Re:ATM NFC Hack
by FounderBurr on Saturday June 26, 2021 @10:37PM
(#27648025)
If you can’t be trusted to handle basic, simple non-creative tasks like managing memory and buffers you are not actually a professional programmer, though you might have convinced an employer otherwise.
By this logic no C or C++ developer in history is a professional programmer. Approximately zero non-trivial programs have been written in C or C++ without memory management problems and/or vulnerabilities.
That's an ignorant and wrong assertion, spoken with an incredible amount of confidence.
Look up software verification. There are verifiers for C.
This is misleading. Formally verifying the behaviour of any large software is a really tall task, no matter what language they are written in.
That's not what the parent comment asserted though.
Often it's the employer encouraging the programmers to cut corners.
Re:ATM NFC Hack
by jbverschoor on Sunday June 27, 2021 @05:18AM
(#27649954)
With a lot of things I agree with this statement, but not about memory management in C.
It's the "fault" of the programmer, but the language makes it very easy to make these faults, so I'd say it's the fault of the language.
Re:ATM NFC Hack
by bottled_poe on Sunday June 27, 2021 @02:24AM
(#27649112)
No, it’s the market.