4 ms·

Time is a number. At least, it makes sense to subtract them (I guess you can say that time is a one dimensional affine space).

kzrdude 5y ago
just like pointers, it makes sense to subtract them and add offsets but to straight up add them makes no sense.
pjc50 5y ago
Time is a number and a location. If you normalize everything to a reference time such as the UNIX epoch, you can usually do subtraction. Even then you may have got the times from a non-monotonic clock ...
bsza 5y ago
Time is a number, but a date is not. 211231 + 1 != 211232.
- karpierz 5y ago
  To be specific, that representation of date does not implement integer addition. To be fair, neither do most "number" types (uint, int, float, etc). For uint, you actually have x + y = (x + y) mod MAX_UINT. For float, sometimes x + 1 = x. The issue is not that date is not a number, but that people poorly reason about common cases of date arithmetic. Similar stuff comes up for floats and ints but they're just easier to reason about.
  - gkop 5y ago
    > For float, sometimes x + 1 = x Would you spell this out for us?
    - eska 5y ago
      If x is a huge number, then elimination can happen. Floating point numbers store a mantissa and an exponent, m^e. When you have a huge number the exponent will be large. So the mantissa can not express the tiny number anymore.
    - cmeacham98 5y ago
      Floats effectively work by specifying a level of precision and a multiplier against that. For big numbers, you need a lower precision (larger gap) between the representable values. Eventually the precision gets so low that adding 1 doesn't reach the next possible value and rounds back down to your current value.
kaba0 5y ago
Bit of a nitpick, but due to relativity, time is not simply a number though.
- wyldfire 5y ago
  Relativity is not modeled in the vast, vast majority of computer systems that track time. So the nitpick is irrelevant.
  - kaba0 5y ago
    But there are rare software that model it (eg. GPS systems famously have to account for it) - so I think it is a textbook definition of a nitpick.

kzrdude 5y ago

just like pointers, it makes sense to subtract them and add offsets but to straight up add them makes no sense.

pjc50 5y ago

Time is a number and a location. If you normalize everything to a reference time such as the UNIX epoch, you can usually do subtraction. Even then you may have got the times from a non-monotonic clock ...

bsza 5y ago

Time is a number, but a date is not. 211231 + 1 != 211232.

karpierz 5y ago
To be specific, that representation of date does not implement integer addition. To be fair, neither do most "number" types (uint, int, float, etc). For uint, you actually have x + y = (x + y) mod MAX_UINT. For float, sometimes x + 1 = x. The issue is not that date is not a number, but that people poorly reason about common cases of date arithmetic. Similar stuff comes up for floats and ints but they're just easier to reason about.
- gkop 5y ago
  > For float, sometimes x + 1 = x Would you spell this out for us?
  - eska 5y ago
    If x is a huge number, then elimination can happen. Floating point numbers store a mantissa and an exponent, m^e. When you have a huge number the exponent will be large. So the mantissa can not express the tiny number anymore.
  - cmeacham98 5y ago
    Floats effectively work by specifying a level of precision and a multiplier against that. For big numbers, you need a lower precision (larger gap) between the representable values. Eventually the precision gets so low that adding 1 doesn't reach the next possible value and rounds back down to your current value.

kaba0 5y ago

Bit of a nitpick, but due to relativity, time is not simply a number though.

wyldfire 5y ago
Relativity is not modeled in the vast, vast majority of computer systems that track time. So the nitpick is irrelevant.
- kaba0 5y ago
  But there are rare software that model it (eg. GPS systems famously have to account for it) - so I think it is a textbook definition of a nitpick.