BUG in Windows Calc [modified]
-
For this answer i will give you +5 :laugh:
NMehta83 “We cannot fail until we fail to try”
-
It takes half and hour to reach to rating link. :laugh:
Rating is Always appreciated.
Regards,
Hiren Solanki.
Indian Forum | My Articles | My Profile
"You will always find me near 127.0.0.1"
-
For this answer i will give you +5 :laugh:
NMehta83 “We cannot fail until we fail to try”
NMehta83 wrote:
i will give you +5
It's 10 minute already, still not deposited :laugh:
Rating is Always appreciated.
Regards,
Hiren Solanki.
Indian Forum | My Articles | My Profile
"You will always find me near 127.0.0.1"
-
Follow the procedure click '4' click 'sqrt' click '-'(Minus sign) click '2' ('-8.1648465955514287168521180122928e-39') click '+' click '2' Answer should be 2 and It shows 2 only :~
NMehta83 “We cannot fail until we fail to try”
Looks like windows support person. take a 5.
Rating is Always appreciated.
Regards,
Hiren Solanki.
Indian Forum | My Articles | My Profile
"You will always find me near 127.0.0.1"
-
-
Follow the procedure click '4' click 'sqrt' click '-'(Minus sign) click '2' ('-8.1648465955514287168521180122928e-39') click '+' click '2' Answer should be 2 and It shows 2 only :~
NMehta83 “We cannot fail until we fail to try”
I already said that... (but -2, same thing however).
-
Looks like windows support person. take a 5.
Rating is Always appreciated.
Regards,
Hiren Solanki.
Indian Forum | My Articles | My Profile
"You will always find me near 127.0.0.1"
-
-
Follow the procedure click '4' click 'sqrt' click '-'(Minus sign) click '2' Answer should be 0 But It comes back with '-8.1648465955514287168521180122928e-39' answer ;P [Edit] Tested with XP and Seven _[/More Edit] Windows Server 2008 R2 comes same with yet another answer '-1.068281969439142e-19' _[/More Edit] [/Edit]
Rating is Always appreciated.
Regards,
Hiren Solanki.
Indian Forum | My Articles | My Profile
"You will always find me near 127.0.0.1"
modified on Monday, November 1, 2010 7:00 AM
-
Dalek Dave wrote:
Or -4!
why DD? can you explain?
Life's Like a mirror. Smile at it & it smiles back at you.- P Pilgrim So Smile Please
-
Another fails the knowledge about floating point.... When will they ever learn????? Edit: It seems to happen with any rounding around 0. If your result is not 0, the result is correctly rounded, but with values very near 0, it simply 'forgets'.... Edit 2: Press = again, the answer will be -2 as expected confirming the previous value is not rounded correctly, but still good enough due to its 'smallness' (e-39 is very small).
((λ (x) `(,x ',x)) '(λ (x) `(,x ',x)))
modified on Monday, November 1, 2010 7:07 AM
Actually, Calc internals have been replaced (around XP) with a long numbers library, so it is somehow noteworthy that this isn't enough. OTOH, the OP failed to notice tha 8e-39 is zero enough when playing around with square roots of 2.
Agh! Reality! My Archnemesis![^]
| FoldWithUs! | sighist | WhoIncludes - Analyzing C++ include file hierarchy -
Actually, Calc internals have been replaced (around XP) with a long numbers library, so it is somehow noteworthy that this isn't enough. OTOH, the OP failed to notice tha 8e-39 is zero enough when playing around with square roots of 2.
Agh! Reality! My Archnemesis![^]
| FoldWithUs! | sighist | WhoIncludes - Analyzing C++ include file hierarchypeterchen wrote:
OTOH, the OP failed to notice tha 8e-39 is zero enough when playing around with square roots of 2.
And that is why I cleverly did not point out who the FAILER actually was ;P
-
Actually, Calc internals have been replaced (around XP) with a long numbers library, so it is somehow noteworthy that this isn't enough. OTOH, the OP failed to notice tha 8e-39 is zero enough when playing around with square roots of 2.
Agh! Reality! My Archnemesis![^]
| FoldWithUs! | sighist | WhoIncludes - Analyzing C++ include file hierarchypeterchen wrote:
Actually, Calc internals have been replaced (around XP) with a long numbers library, so it is somehow noteworthy that this isn't enough.
I actually thought I heard it referred to as an "infinite precision" calculator. I've tested it before on values that would make the pre-XP calc choke, but apparently it's not infinite. Ah, googled it... apparently "arbitrary" precision is what it's called. I think MS changed it in the wake of that idiotic, much publicized and ultimately harmless Pentium "floating point" bug. People kept thinking they were finding more bugs, but they were just running into typical floating point inaccuracies.
He said, "Boy I'm just old and lonely, But thank you for your concern, Here's wishing you a Happy New Year." I wished him one back in return.
-
Another fails the knowledge about floating point.... When will they ever learn????? Edit: It seems to happen with any rounding around 0. If your result is not 0, the result is correctly rounded, but with values very near 0, it simply 'forgets'.... Edit 2: Press = again, the answer will be -2 as expected confirming the previous value is not rounded correctly, but still good enough due to its 'smallness' (e-39 is very small).
((λ (x) `(,x ',x)) '(λ (x) `(,x ',x)))
modified on Monday, November 1, 2010 7:07 AM
-
Follow the procedure click '4' click 'sqrt' click '-'(Minus sign) click '2' Answer should be 0 But It comes back with '-8.1648465955514287168521180122928e-39' answer ;P [Edit] Tested with XP and Seven _[/More Edit] Windows Server 2008 R2 comes same with yet another answer '-1.068281969439142e-19' _[/More Edit] [/Edit]
Rating is Always appreciated.
Regards,
Hiren Solanki.
Indian Forum | My Articles | My Profile
"You will always find me near 127.0.0.1"
modified on Monday, November 1, 2010 7:00 AM
It's called rounding error, and the problem has been around since computers ran on vacuum tubes. The error has become smaller over the years as bits have been added to operands, but it still occurs due the fact that the finite representation of values which are not, in fact, expressible as a mantissa and a fixed length exponent, can contain a small error. It happens to be very noticeable around zero, but it's present in all FP calculations. It is for that reason programmers have been taught for decades never to use a comparison to zero when the control variable is calcuated using floating point math. It leads to expressions like:
if ((x - k =< 0)|(x + k =< 0))
{
DoSomeThingUsefulBecauseXisZero();
}The
if
argument tests for x = 0, +/- k, a constant value. It stumped me when I first saw it in missle guidance test code, until my boss reminded me of what I'd been taught in my only programming class a few years earlier - a bit embarrassing... :-O Have they finally stopped teaching this? :-DWill Rogers never met me.
-
Another fails the knowledge about floating point.... When will they ever learn????? Edit: It seems to happen with any rounding around 0. If your result is not 0, the result is correctly rounded, but with values very near 0, it simply 'forgets'.... Edit 2: Press = again, the answer will be -2 as expected confirming the previous value is not rounded correctly, but still good enough due to its 'smallness' (e-39 is very small).
((λ (x) `(,x ',x)) '(λ (x) `(,x ',x)))
modified on Monday, November 1, 2010 7:07 AM
It's still kinda ironic that cheap calculators from 20 years back will still display what the end-user wants to see, while a sophisticated OS comes with a calculator app that thinks it's okay to let its internal workings leak onto the UI.
Regards, Nish
My technology blog: voidnish.wordpress.com Code Project Forums : New Posts Monitor This application monitors for new posts in the Code Project forums.
-
It's still kinda ironic that cheap calculators from 20 years back will still display what the end-user wants to see, while a sophisticated OS comes with a calculator app that thinks it's okay to let its internal workings leak onto the UI.
Regards, Nish
My technology blog: voidnish.wordpress.com Code Project Forums : New Posts Monitor This application monitors for new posts in the Code Project forums.
I still have my first calculator from around 30 years ago, still on its original battery. Still works but takes a good few seconds to respond to each key press. I also still have my alarm clock radio at the side of my bed that I got for my 8th birthday (28 years ago last month). Still keeps and displays perfect time although it is many years since I have tried the alarm or radio out.
Every man can tell how many goats or sheep he possesses, but not how many friends.
-
It's called rounding error, and the problem has been around since computers ran on vacuum tubes. The error has become smaller over the years as bits have been added to operands, but it still occurs due the fact that the finite representation of values which are not, in fact, expressible as a mantissa and a fixed length exponent, can contain a small error. It happens to be very noticeable around zero, but it's present in all FP calculations. It is for that reason programmers have been taught for decades never to use a comparison to zero when the control variable is calcuated using floating point math. It leads to expressions like:
if ((x - k =< 0)|(x + k =< 0))
{
DoSomeThingUsefulBecauseXisZero();
}The
if
argument tests for x = 0, +/- k, a constant value. It stumped me when I first saw it in missle guidance test code, until my boss reminded me of what I'd been taught in my only programming class a few years earlier - a bit embarrassing... :-O Have they finally stopped teaching this? :-DWill Rogers never met me.
Take a 5 for deep description.
Rating is Always appreciated.
Regards,
Hiren Solanki.
Indian Forum | My Articles | My Profile
"You will always find me near 127.0.0.1"
-
Follow the procedure click '4' click 'sqrt' click '-'(Minus sign) click '2' Answer should be 0 But It comes back with '-8.1648465955514287168521180122928e-39' answer ;P [Edit] Tested with XP and Seven _[/More Edit] Windows Server 2008 R2 comes same with yet another answer '-1.068281969439142e-19' _[/More Edit] [/Edit]
Rating is Always appreciated.
Regards,
Hiren Solanki.
Indian Forum | My Articles | My Profile
"You will always find me near 127.0.0.1"
modified on Monday, November 1, 2010 7:00 AM
Not a bug.
Need custom software developed? I do custom programming based primarily on MS tools with an emphasis on C# development and consulting. I also do Android Programming as I find it a refreshing break from the MS. "And they, since they Were not the one dead, turned to their affairs" -- Robert Frost
-
Follow the procedure click '4' click 'sqrt' click '-'(Minus sign) click '2' Answer should be 0 But It comes back with '-8.1648465955514287168521180122928e-39' answer ;P [Edit] Tested with XP and Seven _[/More Edit] Windows Server 2008 R2 comes same with yet another answer '-1.068281969439142e-19' _[/More Edit] [/Edit]
Rating is Always appreciated.
Regards,
Hiren Solanki.
Indian Forum | My Articles | My Profile
"You will always find me near 127.0.0.1"
modified on Monday, November 1, 2010 7:00 AM