Beta

Slashdot: News for Nerds

×

Welcome to the Slashdot Beta site -- learn more here. Use the link in the footer or click here to return to the Classic version of Slashdot.

Thank you!

Before you choose to head back to the Classic look of the site, we'd appreciate it if you share your thoughts on the Beta; your feedback is what drives our ongoing development.

Beta is different and we value you taking the time to try it out. Please take a look at the changes we've made in Beta and  learn more about it. Thanks for reading, and for making the site better!

Excel 2007 Multiplication Bug

kdawson posted more than 6 years ago | from the be-fruitful-and-all-that dept.

Microsoft 806

tibbar66 writes with news of a serious multiplication bug in Excel 2007, which has been reported to the company. The example that first came to light is =850*77.1 — which gives a result of 100,000 instead of the correct 65,535. It seems that any formula that should evaluate to 65,535 will act strangely. One poster in the forum noted these behaviors: "Suppose the formula is in A1. =A1+1 returns 100,001, which appears to show the formula is in fact 100,000... =A1*2 returns 131,070, as if A1 had 65,535 (which it should have been). =A1*1 keeps it at 100,000. =A1-1 returns 65,534. =A1/1 is still 100,000. =A1/2 returns 32767.5."

cancel ×

806 comments

Microsoft just announced plans for their fix (5, Funny)

Anonymous Coward | more than 6 years ago | (#20737847)

They will be disabling multiplication in all future versions of Excel.

Re:Microsoft just announced plans for their fix (1)

hedkandee (1148031) | more than 6 years ago | (#20738123)

Maybe we can report a bug in windows version numbers, I never figured out what happend to windows 4-94, 96, 97 and 99-1999

In OOXML? (4, Funny)

gvc (167165) | more than 6 years ago | (#20737849)

Perhaps this is how multiplication is done in OOXML. They do leap years in dates wrong, too.

Re:In OOXML? (0)

Anonymous Coward | more than 6 years ago | (#20737869)

Thats why alternatives such as Matlab, SigmaPlot and OpenOffice are much preferred to Excel where I work.

Re:In OOXML? (5, Insightful)

Anonymous Coward | more than 6 years ago | (#20737997)

I use Matlab and Octave constantly for things everyone around me uses Excel for (I do structural engineering). I am no amazing hacker or anything, but I simply find it scads easier to use that sort of paradigm over the spreadsheet analogy for almost any application. That aside, Excel in particular seems constantly to try to outthink me and consistently to have these sorts of strange calculation errors.

Re:In OOXML? (2, Insightful)

XenoPhage (242134) | more than 6 years ago | (#20738253)

Huh.. Excel works fine where I work. They have all sorts of network drawings in Excel...

What? Hey, stop laughing. I'm serious! You should see these things.. Massive diagrams all built using the draw toolbar in Excel...

Personally, I've found Visio to work out nicely for network drawings, but apparently the other engineers like Excel...

Speaking of which.. Anyone know of a decent OSS replacement for Visio? And no, dia doesn't count. It's nowhere near what Visio does. Nor does it save a Visio compatible file.

Pentiums (5, Funny)

MyLongNickName (822545) | more than 6 years ago | (#20737853)

What happens if you use this on an older Intel chip? Do the issues cancel out?

Re:Pentiums (5, Funny)

644bd346996 (1012333) | more than 6 years ago | (#20737927)

Perhaps. I tried to test it out, but the hard drive with the swapfile caught fire before Excel finished loading.

Re:Pentiums (5, Funny)

Fuji Kitakyusho (847520) | more than 6 years ago | (#20737933)

Not quite. The cell in that case just reads 99,999.999998263, when it should be 65,534.99994721.

Is anyone using Excel 2007? (4, Insightful)

QuantumG (50515) | more than 6 years ago | (#20737861)

I bought a Dell this year, it came with Office 2003.

Yes. (3, Informative)

oatworm (969674) | more than 6 years ago | (#20737875)

Yep - my office switched to Vista and Office 2007. Then again, we're a networking firm, so it's in our best interests to use stuff while it's still "beta" so we know the bugs and quirks before our customers start playing with it.

As an aside, when I went to pick up a lease renewal form for my apartment complex, I noticed that the lady at the front counter was also running Office 2007, so I'd say it's out there - just not exceptionally widespread at the moment, compared to other versions of Office.

Re:Yes. (1)

PresidentEnder (849024) | more than 6 years ago | (#20738009)

The University where I work has a policy of encouraging users to save in the older .doc format, even though we have Office 2007 licenses for everyone, because we want the people who haven't upgraded to be able to open all the files they need. But there's high turnover with student employees, and the administrators don't tend to save things properly, so we end up having to upgrade everyone as soon as one person in a department has 2007. At least I have job security.

Re:Is anyone using Excel 2007? (2, Funny)

WK2 (1072560) | more than 6 years ago | (#20737905)

Is anyone using Excel 2007?

65535 is a popular number. It is 2^16-1. This bug was discovered as soon as QA started (the day after release.)

Re:Is anyone using Excel 2007? (2, Funny)

Anonymous Coward | more than 6 years ago | (#20738121)

65535 is a popular number. It is 2^16-1
No wonder I'm not a popular number. I'm 2^10+313, and I always figured it was the ear wax or poor hygiene.

Re:Is anyone using Excel 2007? (1)

The MAZZTer (911996) | more than 6 years ago | (#20738173)

I got Office Pro 2007 through a Microsoft promotion for free. There was also a recent promotion for college students to get Office Ultimate 2007 for $60.

Re:Is anyone using Excel 2007? (4, Funny)

QuantumG (50515) | more than 6 years ago | (#20738281)

Of course you did. You don't think they're going to pay you to do bug testing for them do you?

Yes, 2007 was dumped on Universities. (1)

twitter (104583) | more than 6 years ago | (#20738207)

Office 2007 was put into the "environment of abundance" funded by student fees at LSU about a year ago. Many people have been tempted to update their Office because they can do it to XP. Most of them hate it but it works and many are afraid it's still some kind of standard or that it will help them get a job and so on and so forth.

I've heard stories of the same kind of "deal" dumping at local high schools and well flattered administrators ate it up.

Hopefully, business have more sense. The rejection of Vista [slashdot.org] serves as a temporary relief because it's keeping people from buying new computers. Even later, business will want to avoid 2007 for bugs like this, the radical interface change and all the usual file format upgrade incompatibilities. All of these have shown in larger amounts because M$ is hungry and demands sacrifice.

Who are you going to call? (5, Funny)

renegadesx (977007) | more than 6 years ago | (#20737873)

1 2 3 4 5 5 6 7 8 9 10 11 12... ...65,533 65,534 100,000

Give em a break, even the Count from Sesame Street cant count that high.

Re:Who are you going to call? (5, Funny)

dvonhand (1136711) | more than 6 years ago | (#20737929)

Who are you going to call?


Ghostbusters?

Re:Who are you going to call? (1)

VGPowerlord (621254) | more than 6 years ago | (#20738069)

Who are you going to call?

Ghostbusters?

I think Buffy would be better in this case, since we're dealing with Counts.

Re:Who are you going to call? (5, Funny)

AJWM (19027) | more than 6 years ago | (#20738241)

1 2 3 4 5 5 6 7 8 9

But I think the Count can manage to count to six without stuttering.

Oh no! (0, Troll)

kwabbles (259554) | more than 6 years ago | (#20737887)

This is the most devastating bug in years. Heaven have mercy on us all.

Re:Oh no! (4, Insightful)

timmarhy (659436) | more than 6 years ago | (#20737931)

well, it could very well lead to massive fuckups if anyone happens to make a decision based on the flawed result.

Re:Oh no! (5, Funny)

renegadesx (977007) | more than 6 years ago | (#20737991)

This just in: Florida plans to do use Microsoft Excel to calculate the 08 election results.

News at 11

Re:Oh no! (1, Funny)

Anonymous Coward | more than 6 years ago | (#20738227)

That was 6 minutes after your post, my time. It's been 20 minutes already! Where's my news!?

Oh, hell (1)

$RANDOMLUSER (804576) | more than 6 years ago | (#20737899)

It's close enough for non-technical users, what's the fuss [wikipedia.org] ?

Dunno... (0, Troll)

Frosty Piss (770223) | more than 6 years ago | (#20737911)

Dunno, works just fine in my copy of 2007...

Re:Dunno... (0)

Anonymous Coward | more than 6 years ago | (#20738023)

It doesn't, however, work in mine. (excel 2007 on Vista, I don't think at the most recent patchlevel).

Re:Dunno... (1)

Duhavid (677874) | more than 6 years ago | (#20738109)

I have excel 2007 on XP, and I get the 100,000
result. My copy of excel 200 on Win2k gave the
correct result.

Re:Dunno... (5, Funny)

Ford Prefect (8777) | more than 6 years ago | (#20738201)

My copy of excel 200 on Win2k gave the correct result.

What, lxvDXXXV?

(And yes, what have the Romans ever done for us, apart from apparently producing correctly functioning spreadsheet software?)

Re:Dunno... (-1, Redundant)

AbRASiON (589899) | more than 6 years ago | (#20738119)

Mod this guy up, I'm in Australia using Excel 2003 SP2 (11.8117.8107) and it's working fine here.

Re:Dunno... (1)

JackieBrown (987087) | more than 6 years ago | (#20738181)

Did you even read the title of the summary?

Re:Dunno... (2, Informative)

chrisb33 (964639) | more than 6 years ago | (#20738213)

The bug shows up for me. I'm running on Vista, and my About box gives the Excel version as 12.0.6024.5000 (MSO 12.0.6017.5000).

obviously malicious (-1, Troll)

Anonymous Coward | more than 6 years ago | (#20737913)

I don't care how Microsoft tries to explain this one, because it's obviously malicious. I'm truly shocked that black box testing missed it, but there's no possible way white box/code inspection could have missed it. Since Microsoft does both kinds of testing, this was obviously a deliberate error injected by some disgruntled employee. Root cause: only a limited number of eyes are allowed to see the code.

Kids, this is why we need all software to be open source, even if it's not free as in beer.

Re:obviously malicious (4, Insightful)

Cassius Corodes (1084513) | more than 6 years ago | (#20737973)

Never attribute to malice what can be adequately explained by incompetence.

Re:obviously malicious (1)

aichpvee (631243) | more than 6 years ago | (#20738149)

Never attribute to incompetence what can be adequately explained by malice.

Can we move on now? It's completely possible to point out that microsoft is capable of extreme incompetence without repeating this nonsense.

Re:obviously malicious (5, Insightful)

Belacgod (1103921) | more than 6 years ago | (#20738189)

Sufficiently advanced incompetence is indistinguishable from malice.

Re:obviously malicious (0)

Anonymous Coward | more than 6 years ago | (#20738007)

I hear you brother, I specifically look for these types of errors when I am reviewing the code on open source spreadsheet programs. I am about 90% of the way through Gnumeric 1.0.1 and I should ready to put it into production in about another six months.

Re:obviously malicious (2, Insightful)

megaditto (982598) | more than 6 years ago | (#20738267)

Not the point. The fix would probably be available within minutes if the source code were open. Plus, a more rigorous review would have been triggered to make sure there aren't more errors of the same type...

With Microsoft, we will be waiting for days before they even notice it...

As such, their product stands unusable, possibly for weeks. And there is nothing we can do about it, not even sue for lost profits or damages.

I am pretty sure those that bought Office at an academic/company discount would also not be able to return it.

Re:obviously malicious (0)

Anonymous Coward | more than 6 years ago | (#20738271)

Gnumeric 1.0.1

Hurr hurr.

But seriously, if you want to know your code is secure, you take the code and you audit it. If you want other people to tell you the code is secure, promise, cross their heart, hope to die, etc etc (but the EULA says you can't hold them to it) then by all means, take their word for it.

If you audit the code and it takes you 12 months to do the most recent major version, it should only take you a few more months to audit the changes to that original code (you did take notes, right?), and a short while more to audit the changes during the audit of the patches... then you're all caught up, and you can audit patches as they come out.

Who knows, if you get good at it, people will start taking your word for it.

Good Luck! (1)

feranick (858651) | more than 6 years ago | (#20737917)

To all people in science and engineering (I know lots of them) using Excel to analyze their data...

Re:Good Luck! (1)

MightyMartian (840721) | more than 6 years ago | (#20737957)

I'll wager most are still using some Office2k variant, so I doubt this issue will affect them all that much.

Re:Good Luck! (3, Insightful)

bockelboy (824282) | more than 6 years ago | (#20737969)

If they are using Excel to analyze their data, I don't want to drive over a bridge they are designing...

None of the math or physics folks I work with would think about using Excel for their data.

Re:Good Luck! (1)

Shados (741919) | more than 6 years ago | (#20738131)

Funny enough, Excel 2007 is a decent OLAP client... so I guess its not Excel itself doing the maths, but its still analysing data =P

Re:Good Luck! (-1, Flamebait)

Jarjarthejedi (996957) | more than 6 years ago | (#20737983)

Yeah...cause the chances of having a number evaluate to 65,535 via multiplication is extremely common.

Yes this is a bug. Yeah Microsoft should fix it. No it doesn't really affect anything. The odds of this affecting your calculations are infinitesimal compared to the odds that one of the other 1000 problems you can have will affect your equations. I'm no statistician (perhaps I should use Excel to calculate it?) but I would image you have a better chance of winning the lottery than encountering this problem in a way that actually affects you, much better.

Re:Good Luck! (4, Insightful)

AJWM (19027) | more than 6 years ago | (#20738125)

Yeah...cause the chances of having a number evaluate to 65,535 via multiplication is extremely common.

What are the odds of that bug only affecting that number? It's a symptom that the underlying routine is totally fscked, in ways that could cause nasty surprises.

If it gave the wrong answer all the time you'd know it was crap and would just ignore the whole thing. It's when it only gives wrong answers some of the time that can lead you into a false sense of confidence.

You've also got to wonder, if it worked fine in previous versions of Excel, what the frack they were messing with to hose it up. It's not like somebody changed the rules of arithmetic recently, did they?

Re:Good Luck! (1)

the_fat_kid (1094399) | more than 6 years ago | (#20738161)

this rare and uncommon number is 931,422 less that you slashdot ID (yes, 1,028,864 lower than mine)
it is also the last five digits of my work phone number, the zip code for Leasburg, MO, and, according to WIKIpedia, "65535 is a frequently occurring number in the field of computing because it is the highest number which can be represented by an unsigned 16 bit binary number"

but I'm sure it will never come up

Re:Good Luck! (0)

Anonymous Coward | more than 6 years ago | (#20738219)

"it is also the last five digits of my work phone number, the zip code for Leasburg, MO, and, according to WIKIpedia"

Which are both horrible examples because you are rarely going to try to determine a phone number or zip code by multiplying numbers.

There are lots of us who do work with numbers in that range all that time and it's a real concern then, but your examples don't really apply to the situation at all.

Re:Good Luck! (2, Insightful)

SnowZero (92219) | more than 6 years ago | (#20738165)

Let's see, on one hand, we have a 16-bit integer. On the other hand, we have computers that can do almost a billion multiplies per second. Nope, shouldn't be a problem... oh wait. The Pentium bug was much more difficult to hit, yet with millions of them out there, people managed to hit it while doing important computations (money, safety, etc). Another post mentions structural engineers using Excel, and I doubt they are doing just simple sums and averages; Would you really want go through an earthquake in a building with safety margins calculated in Excel?

Re:Good Luck! (4, Insightful)

bmo (77928) | more than 6 years ago | (#20738175)

"cause the chances of having a number evaluate to 65,535 via multiplication is extremely common."

"No it doesn't really affect anything. "

Oh, it's only the largest unsigned 16 bit number and comes up in a crapload of places. Enough times that you should have _memorized_ it by now.

Jeez. Please turn in your geek card NOW.

--
BMO

Numerical Analysis of MS Excel (0, Informative)

Anonymous Coward | more than 6 years ago | (#20737923)

Years ago I wrote a parser and interpreter for MS Excel. At that time, the numerical representation was not IEEE floating point, but a truncated, poor cousin (8 bit) of a decent float implemetation called "RK numbers". I recall that I needed to interpolate and provide semi-randomized numbers for the gaps in the number line that 8 bits (mantissa + ordinate) would incur.
I wonder if MS still uses this implementation - which goes back to the early 1990s. It's a propietary, gappy implementation that was based, I think, on Mac / Motorola implementations.
Just my cut on it - I may be off base.

Re:Numerical Analysis of MS Excel (2, Informative)

Mr2001 (90979) | more than 6 years ago | (#20738117)

RK numbers in Excel [softhelp.ru] . They're not quite how you described them, though. It's a 32-bit value that can hold a 30-bit floating point number or a 30-bit integer (the last 2 bits are for the type), and if a number can't be represented either way, Excel will save it as a 64-bit IEEE float.

Google Spreadsheet bug (4, Interesting)

DJ_Perl (648258) | more than 6 years ago | (#20737925)

Open a Google Docs spreadsheet. Type =COMBIN(55,27)

So? (1)

Chouonsoku (1009817) | more than 6 years ago | (#20737935)

Who really needs the rest of those numbers between 65,535 and 100,000 anyway. If I can't count something on my 65,534 fingers, I just don't bother.

It only gets worse. (5, Funny)

The Earl of Sandwich (1160741) | more than 6 years ago | (#20737943)

Try =6*9.

Re:It only gets worse. (5, Funny)

$RANDOMLUSER (804576) | more than 6 years ago | (#20737987)

Try =6*9.
6 * 9 = 42
What's your point?

Re:It only gets worse. (1)

The One and Only (691315) | more than 6 years ago | (#20738053)

Wow, this is the first time I've seen the explanation to a joke modded Funny and the joke itself not modded up at all.

Re:It only gets worse. (1)

$RANDOMLUSER (804576) | more than 6 years ago | (#20738229)

I must totally agree. But then again, there's something essentially wrong about the universe...

Vista Sales (4, Funny)

xs650 (741277) | more than 6 years ago | (#20737945)

It's used in the algorithm that MS uses to report Vista sales.

Explains a lot! (1)

jcarkeys (925469) | more than 6 years ago | (#20737949)

Hmm... I guess MIT has a valid excuse for screwing up their SAT Math scores...

Re:Explains a lot! (1)

Tablizer (95088) | more than 6 years ago | (#20738071)

Hmm... I guess MIT has a valid excuse for screwing up their SAT Math scores...

They used Open Office, shhhhh....
     

Feature, not a bug (1)

wooden pickle (1006975) | more than 6 years ago | (#20737959)

Sounds like something intentionally put in for The DoD budget.

Another thing to try (0)

Anonymous Coward | more than 6 years ago | (#20737993)

Try this

Alt-F11 (to get into VBA mode, or Menu-Tools-Macro-Visual basic editor)
Ctrl-G (immediate window)

msgbox (20123 + 10)
You get 20133 as you would expect

try:
msgbox (20123 + 20123)
d'oh - can't handle this, must wait for advanced version of the software to come out :^@

hmm (-1, Troll)

smoondog (85133) | more than 6 years ago | (#20737999)

Working just fine on my OS-X mac...

Re:hmm (1)

doombringerltx (1109389) | more than 6 years ago | (#20738287)

They don't have excel 2007 for the mac

Microsoft Excel is the Worst Part of Office 2007 (2, Interesting)

ChaoticCoyote (195677) | more than 6 years ago | (#20738017)

I use the Office 2007 versions of Word, Excel, and OneNote on a regular basis for my Windows business work; the *only* one I've had *any* trouble with is Excel. I can blue screen my machines with Excel on a regular basis; unfortunately, one of my paying customers requires Excel 2007 because they invested heavily in it (long before I was contracted), so I haven't got much choice, since .xlsx documents won't open in any FOSS spreadsheet apps.

For my personal work, I use Gnumeric on Linux.

Re:Microsoft Excel is the Worst Part of Office 200 (1, Insightful)

QuantumG (50515) | more than 6 years ago | (#20738057)

Here's a tip. If an application can cause a kernel fault, it's not the application that is broken.

Re:Microsoft Excel is the Worst Part of Office 200 (0, Troll)

Anonymous Coward | more than 6 years ago | (#20738127)

Here's a tip. The same fuckers wrote the application and the kernel, and we hate them, and no one wants your snarky effort to show you took a high school intro to computers class.

Re:Microsoft Excel is the Worst Part of Office 200 (2, Insightful)

AJWM (19027) | more than 6 years ago | (#20738177)

If an application can cause a kernel fault, it's not only the application that is broken.

There, fixed that for you.

Re:Microsoft Excel is the Worst Part of Office 200 (1)

QuantumG (50515) | more than 6 years ago | (#20738243)

Uhhh no.

The application simply should not be able to cause a kernel fault.

Period.

I'm sure this happens at Microsoft. A bug gets reported that Excel 2007 is causing blue screens. The bug gets given to the Excel team, they look at the crash dump and find out that they are doing something that trips a bug in the kernel of Vista, and they change their code to work around the issue. Bug closed. Next time someone runs across the exact same issue, they do a workaround too. The bug in the kernel never gets fixed.

Re:Microsoft Excel is the Worst Part of Office 200 (1)

Torodung (31985) | more than 6 years ago | (#20738141)

Maybe someone figured out a way to use Excel 2003 as a platform for audio/video capture and it too was "fixed" to please the content industry? ;^)

--
Toro

So What? (1)

Ritchie70 (860516) | more than 6 years ago | (#20738021)

I literally don't know ANYONE who does any math, whatsoever, in Excel.

It's all tables and primitive databases. The guy in the next cube does some pretty graphs. That's as close as it gets.

Re:So What? (1)

MT628496 (959515) | more than 6 years ago | (#20738067)

Are you serious? People do it all the time, whether working on a budget or inventory etc.

Re:So What? (1)

Ritchie70 (860516) | more than 6 years ago | (#20738087)

I'm joking in my statement, but serious that I've never seen it at my work.

Re:So What? (1)

RogueCode (517348) | more than 6 years ago | (#20738137)

literally don't know ANYONE who does any math, whatsoever, in Excel.
Have you ever been in a investment bank, private equity, derivative/"high yield" funds ? ALL they use is Excel. I think that should explain one thing or two about subprime crisis...

Re:So What? (1)

Titoxd (1116095) | more than 6 years ago | (#20738167)

I literally don't know ANYONE who does any math, whatsoever, in Excel.
Then you are not a student.

Simply put, Excel does come with some nifty statistics functions, and the ability to put a chart or table in a Word document is priceless when doing written reports for science classes. True, it is not the first thing that comes to mind; the best program for such type of analysis is something similar to MATLAB, but Excel's matrix functions are not that horrible, putting these senseless blunders aside. Unless you have a lot of money to shell out for a MATLAB license (if you are a student, you can get a semi-reasonable price, but if you're a company, be prepared to pay several grand for a commercial license), Excel is the tool of choice for most of the work we do at an undergraduate level. It is widely available, so you know you can work on your data *cough*homework*cough* in many places.

While other FOSS software may come with this ability, it is a lie to simply say that no one uses Excel for mathematical analysis.

~~~~

hopefully openoffice.org (0)

Anonymous Coward | more than 6 years ago | (#20738033)

...did not copy this.

The number of the beast.... (1)

RealGrouchy (943109) | more than 6 years ago | (#20738047)

Nobody at Microsoft is actually allowed to use the number 65,535, lest Steve Ballmer come and rain fiery chairs down upon them.

(Should I have gone with the "MacBeth/Scottish Play" reference instead?)

- RG>

Check a C multiply on that machine... (2, Insightful)

tjstork (137384) | more than 6 years ago | (#20738079)

Just for the heck of it. There's always an offbeat chance that the multiply bug is in the CPU, not Excel....

Re:Check a C multiply on that machine... (1)

Vombatus (777631) | more than 6 years ago | (#20738265)

There's always an offbeat chance that the multiply bug is in the CPU, not Excel....

But it works as expected in other versions of Excel - that would rule out a CPU problem

I can see the OOXML tag now... (5, Funny)

mad.frog (525085) | more than 6 years ago | (#20738081)

<MultiplyLikeExcel2007>
        =850*77.1
</MultiplyLikeExcel2007>

Re:I can see the OOXML tag now... (5, Insightful)

ameline (771895) | more than 6 years ago | (#20738235)

That profoundly unfunny post (but insightful) will probably get modded as funny.

Lets wait and see :-)

Unnecessary abstraction (5, Funny)

mikvo (587789) | more than 6 years ago | (#20738089)

Multiplication is an unnecessary abstraction anyway. This should really be represented by summing the value of 77.1 entered independently into 850 cells: =sum(a1:a850).

Re:Unnecessary abstraction (5, Funny)

Rick Genter (315800) | more than 6 years ago | (#20738105)

I'd prefer to put 850 in 77.1 cells and do =sum(a1:a77.1)

Can't even imagine how to screw this up (0)

Anonymous Coward | more than 6 years ago | (#20738095)

If I were given requirements as given by the author of the summary, I don't think I can figure out a way to screw it up so horribly, consistently, and intentionally. It boggles the mind how to implement this so incorrectly - treating 65536 as a hex (0x10000) and then a decimal (10,000) is off by a factor of 0x10, or 10 depending.

At least the pentium errors were explicable - incorrect lookup tables sure. Randomly interpreting hex numbers as decimal, incorrectly? Wow....

MS Math (1)

javacowboy (222023) | more than 6 years ago | (#20738097)

So....now that Microsoft can't get ISO approval for OOXML, they've gone straight for an amendment to basic mathematics specs? :D

Is anyone testing OO.org spreadsheets? (1)

xenocide2 (231786) | more than 6 years ago | (#20738115)

I doubt the same bug occurs, but does anyone evaluate how OO.org fares in math flaws? I suspect some people hold OO.org to a lower standard than one should otherwise expect.

Granted, the above bug looks like a rounding error gone horribly, horribly, incredibly horribly wrong.

some limited testing (5, Interesting)

indaba (32226) | more than 6 years ago | (#20738133)

original 850 77.1 100000
425 154.2 100000
212.5 308.4 100000
8500 7.71 100000
but this evaluates correctly..

25 2621.4 65535

so it's not every multiplication that evaluates to 65535

I'm using Excel 2007 12.0.6024.5000

Re:some limited testing (1)

ocelotbob (173602) | more than 6 years ago | (#20738273)

I don't have a copy of excel in front of me, but what happens when you try (8500 * 1.000) * 7.71?

Retro (3, Funny)

starfishsystems (834319) | more than 6 years ago | (#20738183)

I always said that Microsoft would never successfully migrate from 16 to 32 bits...

OOXML (1)

FunkyRider (1128099) | more than 6 years ago | (#20738195)

That explains why OOXML can't be an ISO standard, it evaluates 65525 = 100000

Kinda Ironic (2, Insightful)

pembo13 (770295) | more than 6 years ago | (#20738205)

I find this bug kinda ironic since one of the big pros attributed to Excel over OO Calc is that it is so useful for data manipulation. Yet, the number which results in the bug isn't ever a particularly big number - well within a practical use case.

But the important question is.. (1)

AJWM (19027) | more than 6 years ago | (#20738211)

Does the built-in flight simulator still work?

(Yeah, I know, they took it out after Excel 97.)

Found only now? (0, Troll)

The MAZZTer (911996) | more than 6 years ago | (#20738221)

If this had been a bug with, say, some iPhone spreadsheet program, it would have been found and posted launch day.

Why has it taken so long for someone to find this? Other than the plausible explanation that MS' push for people to upgrade isn't going so well...

Microsoft confirms it (1)

iminplaya (723125) | more than 6 years ago | (#20738237)

Math is hard.

Let's go shopping

Works fine for me! (1)

superash (1045796) | more than 6 years ago | (#20738245)

I think it is fixed in Office 2003 with the SP2 update.

64K integers... (0)

Anonymous Coward | more than 6 years ago | (#20738251)

Should be enough for anybody.

Imperiled by binary decimals? (5, Interesting)

Protoslo (752870) | more than 6 years ago | (#20738277)

It sounds like they are doing small-number math in one representation (perhaps they use short fixed-width decimal representations) and then switching to another method (arbitrary length decimal numbers?) at the binary-inspired boundary 2^16...but somehow they got it mixed up with a different decimal boundary in the edge case.

Clearly the error is weirdly subtle, if 5.1*12850 gives the bugged behavior, but 8.5*7710 works just fine. In fact, I verified that all permutations of a bugged combination =A*B of the form =A/2*B*2 are bugged. Further...all of the buggy decimal values have no perfect floating point binary representation. 77.1 has an infinite binary expansion using IEE 754, while 8.5 has an exact representation. It seems likely that they are only using their BCD format (or whatever) when binary floating (or fixed) point just won't cut it, but then their internal->decimal conversion code chokes on 2^16 for some reason, while the binary (whether it is floating or fixed point) conversion works just fine (possibly because it doesn't have a boundary at 2^16--maybe it has its own threshold bugs ;p).

Don't worry its a feature.... (1)

Chineseyes (691744) | more than 6 years ago | (#20738289)

This is just the new proprietary numeric system. While the rest of the use use base2, 8, 16, or 10 Microsoft operate on baseMS. The 100,000 baseMS (65,535 base10) page white paper is forthcoming.
Load More Comments
Slashdot Account

Need an Account?

Forgot your password?

Don't worry, we never post anything without your permission.

Submission Text Formatting Tips

We support a small subset of HTML, namely these tags:

  • b
  • i
  • p
  • br
  • a
  • ol
  • ul
  • li
  • dl
  • dt
  • dd
  • em
  • strong
  • tt
  • blockquote
  • div
  • quote
  • ecode

"ecode" can be used for code snippets, for example:

<ecode>    while(1) { do_something(); } </ecode>
Create a Slashdot Account

Loading...