17
$\begingroup$

This is going to seem like the smallest of potatoes, but has anyone noticed a change in how displayed reputation is rounded? Up until today, if one had, say, 17,750 reputation, then the reputation would display as 17.8k. Now it displays as 17.7k, not moving until the reputation actually moves through the 17,800 threshold.

Yes, I know, my life may never be the same. But I was just curious if this is part of some larger change.

$\endgroup$
  • 11
    $\begingroup$ Talk about first-world problems! :-) $\endgroup$ – Asaf Karagila May 28 '15 at 11:10
  • 2
    $\begingroup$ Anything over 10k and most people will just assume you're some kind of math genius. If I'm having trouble deciding which answer to accept between yours and some guy's who has a hundred points more or less than you that's not going to decide it for me. $\endgroup$ – Robert Soupe May 31 '15 at 1:17
10
$\begingroup$

Yes, that changed with the fix for the 1000k bug:

we're now explicitly rounding down in the last digit for those style numbers. Previously we were relying on string.Format's rounding behavior... which obviously had some issues.

as Kevin Montrose wrote over there.

It will certainly take a couple of days to get used to the new behaviour.

However, as I just found on meta.se:

Yeaaaah, sort of by design... But I don't particularly like it. Will dig in yet more tomorrow, i18n is just a pain around numbers in general.

So maybe better not get too used to it.

Update 29.05.2015: Indeed, the displayed numbers are now rounded again.

$\endgroup$
  • $\begingroup$ OK, thanks Daniel! $\endgroup$ – Ron Gordon May 28 '15 at 10:35
  • $\begingroup$ But, @Ron, see update, maybe (hopefully, I preferred the old way) it is going to change again. $\endgroup$ – Daniel Fischer May 28 '15 at 10:37
  • $\begingroup$ Thanks again. I dunno, I only care so much as I was so used to the rounding - it had been there as long as I have been on SE. If the powers that be want to move it back, then fine. However, my two cents is, if they have a bug affecting one small part of a scoring algorithm (that is brand new), then maybe a global change like this is not the best approach. $\endgroup$ – Ron Gordon May 28 '15 at 10:41

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .