What a strange series of articles. Somewhat amusing what people will spend their time ranting about.
IMO, if the data logically can’t be negative, use unsigned to document it, if it can, use signed.
Signed vs Unsigned isn’t really the issue here….
Silent, undefined and downright dangerous overflow behaviour vs Silent unuseful, but defined overflow behaviour is a major issue.
Complex comparison behaviour between signed and unsigned is an issue.
Is we just had signed and unsigned and had no weird and nasty overflow or comparison issues…. I agree with your 100%.
Alas we don’t in languages without a sane numeric tower, hence yes, it is an issue that has no “nice” resolution.
I like how they made a CSS3 emulation of the blink tag since most browsers don’t support it anymore. Seems like a lot of people are of the opinion that if they can, they should.
blink
Data point…..
https://labs.spotify.com/2015/08/27/underflow-bug/
Data point… http://googleresearch.blogspot.co.nz/2006/06/extra-extra-read-all-about-it-nearly.html
What a strange series of articles. Somewhat amusing what people will spend their time ranting about.
IMO, if the data logically can’t be negative, use unsigned to document it, if it can, use signed.
Signed vs Unsigned isn’t really the issue here….
Silent, undefined and downright dangerous overflow behaviour vs Silent unuseful, but defined overflow behaviour is a major issue.
Complex comparison behaviour between signed and unsigned is an issue.
Is we just had signed and unsigned and had no weird and nasty overflow or comparison issues…. I agree with your 100%.
Alas we don’t in languages without a sane numeric tower, hence yes, it is an issue that has no “nice” resolution.
I like how they made a CSS3 emulation of the
blinktag since most browsers don’t support it anymore. Seems like a lot of people are of the opinion that if they can, they should.Data point…..
https://labs.spotify.com/2015/08/27/underflow-bug/
Data point… http://googleresearch.blogspot.co.nz/2006/06/extra-extra-read-all-about-it-nearly.html