In the same vein: OpenOffice doesn't print on Tuesdays (including partial explanation at #28, but the whole thread is worth reading) Root cause of the above The one-page doc that jams your printer...

Windows Calculator has been updated (version 10.1803.711.0) to now correctly calculates square roots for perfect squares (integers that are squares of other integers). Because of the arbitrary precision arithmetic library used by the Calculator app, the square root calculation is an approximation calculated using the Exponential Identity function.

Previously, when you would calculate the square root of 4, the result would be 1.99999999999999999989317180305609 which would be rounded to 2 when displayed, because we calculated enough digits to do the rounding correctly. However, as soon as you subtract 2, you would see the remaining digits.

After this update, the square root calculation now recognizes perfect squares and correctly returns exactly 2 for the square root of 4.

Thanks. So I'd guess this would work with any perfect square? As in, sqrt(16) - 4. I wonder if cube roots were the same way. Like rt3(8) - 2. (I don't think it even had that functionality anyways...

Thanks. So I'd guess this would work with any perfect square?

As in, sqrt(16) - 4.

I wonder if cube roots were the same way. Like rt3(8) - 2. (I don't think it even had that functionality anyways though? At least not in the simplified view.)

What a classic piece of IT/internet lore. I remember reading this for the first time in a forum probably 12 years ago, but it is certainly much older. I just love these pieces that have been...

What a classic piece of IT/internet lore. I remember reading this for the first time in a forum probably 12 years ago, but it is certainly much older. I just love these pieces that have been floating around forever.

This is old but gold, and I'll always read it and enjoy it and vote for it. It reminds me of a classic ground speed check which is in the same sort of class; I've read it, I'll read it again, I'll...

This is old but gold, and I'll always read it and enjoy it and vote for it. It reminds me of a classic ground speed check which is in the same sort of class; I've read it, I'll read it again, I'll vote it and try to get more people to read it because it is delightful.

In the same vein:

Might I ask what this is about? I completely missed ever hearing about this.

https://devblogs.microsoft.com/oldnewthing/?p=93765

https://blogs.windows.com/windowsexperience/2018/04/04/announcing-windows-10-insider-preview-build-17639-for-skip-ahead/

Thanks. So I'd guess this would work with any perfect square?

As in,

`sqrt(16) - 4`

.I wonder if cube roots were the same way. Like

`rt3(8) - 2`

. (I don't think it even had that functionality anyways though? At least not in the simplified view.)I still had the vm lying around, so I verified for 625

^{1/4}- 5 and 27^{1/3}- 3, similar "almost zero" results.This story was the first thing that came to my mind when I saw the other day's XKCD.

What a classic piece of IT/internet lore. I remember reading this for the first time in a forum probably 12 years ago, but it is certainly much older. I just love these pieces that have been floating around forever.

Time to also spread the love that is The Bastard Operator From Hell: http://bofh.bjash.com/

Hope it brightens someones day :)

~~It's from 2002! I honestly expected such weird problems to be 90s or even 80s era internet growing pains. But it happened this millennium.~~It happened in the 90s, I didn't read it right!

He wrote it in 2002 but this happened in 90's (he has a FAQ)

Ah, just read the date of the email! 90s makes more sense!

If you enjoyed that, you might also enjoy the emoji that killed Chrome

This is old but gold, and I'll always read it and enjoy it and vote for it. It reminds me of a classic ground speed check which is in the same sort of class; I've read it, I'll read it again, I'll vote it and try to get more people to read it because it is delightful.

Your "in the same vein" posts are great too.