Date: Thu, 04 Sep 2025 14:45:27 -0700
> On Sep 4, 2025, at 1:59 PM, David Brown <david.brown_at_[hidden]> wrote:
>
>
>
> On 04/09/2025 15:33, Marcin Jaczewski wrote:
>> czw., 4 wrz 2025 o 14:51 David Brown <david.brown_at_[hidden] <mailto:david.brown_at_[hidden]>> napisał(a):
>> My point was that you could manually detect overflow if its not UB and
>> react accordingly.
>> UB sometimes makes it very hard to do some naive checks as they are UB itself
>> and can be removed by the compiler. And this is probably the biggest gripe that
>> people have with UB.
>
> That's just nonsense. Seriously - there is no need to have wrapping integer overflow in order to detect problems.
I am getting tired of this.
Your replies on this list are needlessly disrespectful.
If this comment was a single data point I would have ignored it, but you’ve been writing things like this in response to many other participants, myself included. You have been dismissive of any opinion that does not match your own, and your responses are increasingly tending towards simply insulting the opinions of others, if not the people themselves, for being different than your own.
We understand that you don’t believe that anything other than UB is valid. You have made that very clear.
The fact that you believe UB is reasonable, does not mean that all contrary points of view are unreasonable.
There is objective evidence that demonstrates that many cases, including overflow, where labeling behavior that has definite semantics in the real world as being undefined has resulted in security vulnerabilities that only occur due to the compiler being able to pretend overflow is impossible. Hence, it is a reasonable position to take, that the specification should match the reality of integer operations on a computer, rather than the theoretical idea of infinite precision.
This most recent comment from you, is to tell a person who is literally stating objective facts: testing for overflow is trivially easy if overflow is well defined with even “naive" mechanisms, and the specification labeling operations that have well defined semantics as being UB comes up in every discussion of the problems with C++ with C++ developers, security professionals, and government agencies. There is nothing in Marcin’s comment that was their opinion: they were reporting objective information.
You replied with “That's just nonsense. Seriously - there is no need to have wrapping integer overflow in order to detect problems."
The first part of that is directly claiming Marcin’s objectively true comment was false, in an a fairly condescending way, then continuing the condescension in addressing his comment about detection with a “there are other ways” (that is true, there are other ways, they are much more annoying, and are only needed due to overflow being UB), and then ignoring the rest of their - correct - comment.
I’ve put up with these replies from you, because for the most part you were arguing with my _opinion_, but this is your response to someone presenting a more or less entirely objective statement.
Treating other committee members as if their opinion is inherently wrong, or unreasonable, simply because it does not align with your viewpoint is uncalled for.
If you are finding this thread stressful, or frustrating, or whatever - it happens - just take a break, cool off, and come back to it then.
—Oliver
>
>
>
> On 04/09/2025 15:33, Marcin Jaczewski wrote:
>> czw., 4 wrz 2025 o 14:51 David Brown <david.brown_at_[hidden] <mailto:david.brown_at_[hidden]>> napisał(a):
>> My point was that you could manually detect overflow if its not UB and
>> react accordingly.
>> UB sometimes makes it very hard to do some naive checks as they are UB itself
>> and can be removed by the compiler. And this is probably the biggest gripe that
>> people have with UB.
>
> That's just nonsense. Seriously - there is no need to have wrapping integer overflow in order to detect problems.
I am getting tired of this.
Your replies on this list are needlessly disrespectful.
If this comment was a single data point I would have ignored it, but you’ve been writing things like this in response to many other participants, myself included. You have been dismissive of any opinion that does not match your own, and your responses are increasingly tending towards simply insulting the opinions of others, if not the people themselves, for being different than your own.
We understand that you don’t believe that anything other than UB is valid. You have made that very clear.
The fact that you believe UB is reasonable, does not mean that all contrary points of view are unreasonable.
There is objective evidence that demonstrates that many cases, including overflow, where labeling behavior that has definite semantics in the real world as being undefined has resulted in security vulnerabilities that only occur due to the compiler being able to pretend overflow is impossible. Hence, it is a reasonable position to take, that the specification should match the reality of integer operations on a computer, rather than the theoretical idea of infinite precision.
This most recent comment from you, is to tell a person who is literally stating objective facts: testing for overflow is trivially easy if overflow is well defined with even “naive" mechanisms, and the specification labeling operations that have well defined semantics as being UB comes up in every discussion of the problems with C++ with C++ developers, security professionals, and government agencies. There is nothing in Marcin’s comment that was their opinion: they were reporting objective information.
You replied with “That's just nonsense. Seriously - there is no need to have wrapping integer overflow in order to detect problems."
The first part of that is directly claiming Marcin’s objectively true comment was false, in an a fairly condescending way, then continuing the condescension in addressing his comment about detection with a “there are other ways” (that is true, there are other ways, they are much more annoying, and are only needed due to overflow being UB), and then ignoring the rest of their - correct - comment.
I’ve put up with these replies from you, because for the most part you were arguing with my _opinion_, but this is your response to someone presenting a more or less entirely objective statement.
Treating other committee members as if their opinion is inherently wrong, or unreasonable, simply because it does not align with your viewpoint is uncalled for.
If you are finding this thread stressful, or frustrating, or whatever - it happens - just take a break, cool off, and come back to it then.
—Oliver
Received on 2025-09-04 21:45:40