Hacker News Reader

Comment by simonjgreen | Hacker News Reader

Yet another battle won in the 0 indexing vs. 1 indexing war

It just shows you how reality-denying the 1 indexers can be, going as far as calling a newborn 1 even though (if normal term) it was conceived roughly 0.75 years ago.

There is no indexing standard that is more fundamental than another. For example zero-indexing is particularly unnatural when working with mathematical series (https://math.stackexchange.com/a/1203057 ). It made a ton of sense for computing because every arbitrary sequence of bits represents a valid index, but it falls short in some other domains.

Don't consider it an age then, think of it as "the count of calendar years within which a person has existed."

A newborn has only seen 1 year, 2023. An infant born any time in 2022 has lived across 2 years. This simplifies things when you only need, or only have, someone's year of birth.

Why restrict to years? I have existed in two calendar millennia.

The more reality denying part is when you ignore that age is calculated on distance and not counting.

Lua is the “last standing man” now.

Every language for scientific computation uses 1-based indexes, to follow the tradition of using them in formulas. FORTRAN, APL, Julia, R, MATLAB, GNU Octane, etc.

I remember that (some versions of) BASIC back in the day had an option to switch between 1-based and 0-based indexing, to appeal to both Fortran and C folks.

In physics, relativists usually count from 0. Furthermore on the numerical side, C is at least as popular as Fortran.

Julia lets you switch the index base too.

Finally a special mention for Mathematica, which appears to be 1-based, but in fact is a thinly disguised 0-based lisp, {a,b} is just List[a,b] is really just (List a b) so of course if you want b you ask for element 2, element 0 is the head and is just List.

Numpy

Hold your horses, APL[1] is not dead yet. However, at some point in its history, APL did gain the ability to assign the Index Origin[2] to 0 if the problem called for it via:

    ⎕IO←0

0: intentionally started the indexes from 1 ;)

1: https://en.m.wikipedia.org/wiki/APL_(programming_language)

2: https://aplwiki.com/wiki/Index_origin

R still uses 1 indexing, Julia as well.

And FORTRAN

even weirder than that, as far as I can tell, as the whole year cohort ages together on the same day: https://en.wikipedia.org/wiki/East_Asian_age_reckoning

It's not that weird - I'm sure most folks spend their school-years referring to other kids as "the same year (grade) as me" or "the year (grade) above/below me". Even to this day, I intuitively group friends based on the academic year group they were in.

> Even to this day, I intuitively group friends based on the academic year group they were in

How far out of school are you?

I left university over 10 years ago, but I'm still aware of those friends who were in my year, vs others who were a year or two ahead or behind.

People from my year are the ones I lived with, did group projects with, worked until midnight trying to solve a difficult task that had been set, celebrated the last exam with, etc.

I certainly don't know the year for everyone. Students I mostly saw at parties, who studied different subjects (or went to different universities, for it was London) could easily be my year ±3 or more.

I feel like the op was mentally translating people they’d never shared an academic context with into how many academic years above or below them they were — I’d noticed myself doing this for a couple of years after I left education

I prefer calling it "floor-index" vs "ceil-index" rather than 0/1 or offset/count based

Both 0 and 1 indexing aim to answer the question: "what is the index of the 1st element?"

floor(1) = 1 and ceil(1) = 1 so I'm not sure how your names apply.

floor and ceil ideas apply to age and calendar years nicely because the underlying value, time, is on the real number line.

Reply on news.ycombinator.com