The Daily WTF: Curious Perversions in Information Technology

chubertdev · 2014-02-12 Reply Admin

Steve The Cynic:
faoileag:
And we all are like that every once in a while. If I have a cold and need a painkiller, I'm happy to take a couple of aspirins. Do I have to know exactly how acetylsalicylacid works? No.
No, but it might be interesting to know how acetylsalicylic acid works...

My chemist fiancee explained it. It's not interesting.

chubertdev · 2014-02-12 Reply Admin

EuroGuy:
I wanted to check how python dealt with 0.57*100 so I typed:

print '%f' % 0.57*100

I must admit that the output surprised me... (I understand my error now, no need to point it out to me, thank you very much!)

I'm honestly surprised that SELECT 0.57*100 in T-SQL doesn't produce 100.

chubertdev · 2014-02-12 Reply Admin

DWalker:
anonymous:
DWalker:
Yep. If someone takes a screenshot and pastes it into Word, that's fine with me. When I take a screenshot, I paste it into Irfanview and crop the part I want to keep, then save the file as (usually) a JPG.
I don't fault non-technical users for pasting a screenshot into Word.
The only right way to take a screenshot is to paste it into Paint and save as a PNG.

That's the right way for users who know that they have Paint installed (not everyone knows about it), and who know what a PNG is. For "normal" users, pasting a picture into Word is just fine. There are bigger things to worry about.

I've been sent screen shots as giant bitmaps, PNGs, PDFs, Word docs, Powerpoint presentations, I think I even had one pasted into an Excel spreadsheet. Users are funny.

da Doctah · 2014-02-12 Reply Admin

chubertdev:
Steve The Cynic:
faoileag:
And we all are like that every once in a while. If I have a cold and need a painkiller, I'm happy to take a couple of aspirins. Do I have to know exactly how acetylsalicylacid works? No.
No, but it might be interesting to know how acetylsalicylic acid works...

My chemist fiancee explained it. It's not interesting.

Last time I took aspirin I ended up spending six days in the hospital with anemia from a bleeding stomach. I can only assume it works by giving you something so much worse that you forget about the headache.

2014-02-12 Reply Admin

foo AKA fooo:
0 comments. Does this mean I can post a frist comment, or that I can't?

It doesn't manner.

2014-02-12 Reply Admin

Regarding the bug report that reads as if it was written by a drunk schizophrenic - could it be George from The Chronicles of George is still alive and working in the IT industry? It must be him - how could more than one person like that exist on planet earth?

chubertdev · 2014-02-12 Reply Admin

I guess there are only a few of us that have written Red Bull-fueled emails at 4am after working for 37 consecutive hours of that nature.

2014-02-12 Reply Admin

Anon:
Erogenous amount? How about 80085?

819.80085

2014-02-12 Reply Admin

DWalker:
anonymous:
DWalker:
Yep. If someone takes a screenshot and pastes it into Word, that's fine with me. When I take a screenshot, I paste it into Irfanview and crop the part I want to keep, then save the file as (usually) a JPG.
I don't fault non-technical users for pasting a screenshot into Word.
The only right way to take a screenshot is to paste it into Paint and save as a PNG.

That's the right way for users who know that they have Paint installed (not everyone knows about it), and who know what a PNG is. For "normal" users, pasting a picture into Word is just fine. There are bigger things to worry about.

I was talking about the right way, not the way that's just fine.

2014-02-12 Reply Admin

I guess John didn't add them to his Buddy List, so he shouldn't complain when he gets the less effective version.

herby · 2014-02-12 Reply Admin

Bad English??

No, probably someone who is channeling Yoda, or does WAY too much FORTH.

As for Wells Fargo, they seem to want their emails confirmed, so they include "bugs" in them. Unfortunately, my email client doesn't do these things, and I keep seeing the "your email address is wrong" when I go about my online banking. Maybe they will wake up someday about this. Otherwise their web site is one of the best. Not loaded with silly graphics (that take time) or much Javascript (which takes even more time).

2014-02-12 Reply Admin

Please don't frisk the comments.

2014-02-12 Reply Admin

faoileag:
The wtf is that the sender claims this to be an error in java math, not understanding, like you do, the intricacies of working with floating point numbers on digital computers.

Not to mention this strange business of multiplying <1 values by 100, such that sending "57" instead of "0.57" should be considered an equivalent operation that circumvents the "bug". It's hard to follow just what they're trying to do there...

2014-02-12 Reply Admin

Mark Bowytz:
mikeTheLiar:
TDWTF:
Apparently JAVA Math is Hard!
Apparently Java is an acronym.

JAVA = Just Another Vague Acronym

No it's not, people just yell every time they mention JAVA....and for good reason (I can picture someone with tourettes yelling JAVA all the time).

DCRoss · 2014-02-12 Reply Admin

Bob:
Anon:
Erogenous amount? How about 80085?

819.80085

But why are they biq?

2014-02-12 Reply Admin

faoileag:
…:
Note that 0.57 can't be represented exactly as a binary floating point number, so .57*100 is not exactly 57. This means that it can actually be slightly less than 57, which can yield 56 if rounded with an unsuitable method.
This is quite a common error, so I don't see the WTF in the e-mail.
The wtf is that the sender claims this to be an error in java math, not understanding, like you do, the intricacies of working with floating point numbers on digital computers.

To be fair, he says it's a Java math error - and he's right. It doesn't matter that we all understand why it happens and why people should avoid floating point computation and all that rot, the bottom line (rightly or wrongly) is that Java (as most modern languages) is givin an incorrect result. Of course, because the error would always be small, you would think rounding the error away would be fairly trivial - provided you know you have to end up with an integer.

that is, it is not untrue to state that Java math is giving an incorrect result - and note also that he says "a java math error" not "a java.math error" which I think adds some ambiguity anyway)

2014-02-12 Reply Admin

TRWTF is that Rick is getting John's email.

chubertdev · 2014-02-12 Reply Admin

Orac:
Mark Bowytz:
mikeTheLiar:
TDWTF:
Apparently JAVA Math is Hard!
Apparently Java is an acronym.

JAVA = Just Another Vague Acronym
No it's not, people just yell every time they mention JAVA....and for good reason (I can picture someone with tourettes yelling JAVA all the time).

It's on par with Belg**m.

2014-02-12 Reply Admin

mihi:
<email>jason.weaverton&#3f;redacted-client.com</email>
Was it intentional that
a) hex entities are written like ? in XML? b) U+003f is actually a question mark and not an @ sign which is U+0040

?

And for that floating point Java problem: Java since version 1.0 has a BigDecimal datatype that will handle "human readable" decimal numbers just like you expect (in as much precision and rounding modes you like) - you won't be able to do trigonometry and similar stuff on them, but in case you are doing some financial calculations, it might be a good idea to use those instead of floating point data types.

yep, Groovy does this by default for example:

def num = 0.57 assert 0.57 * 100 == 57 assert num.class == java.math.BigDecimal

2014-02-12 Reply Admin

mick:
faoileag:
…:
Note that 0.57 can't be represented exactly as a binary floating point number, so .57*100 is not exactly 57. This means that it can actually be slightly less than 57, which can yield 56 if rounded with an unsuitable method.
This is quite a common error, so I don't see the WTF in the e-mail.
The wtf is that the sender claims this to be an error in java math, not understanding, like you do, the intricacies of working with floating point numbers on digital computers.
To be fair, he says it's a Java math error - and he's right. It doesn't matter that we all understand why it happens and why people should avoid floating point computation and all that rot, the bottom line (rightly or wrongly) is that Java (as most modern languages) is givin an incorrect result. Of course, because the error would always be small, you would think rounding the error away would be fairly trivial - provided you know you have to end up with an integer.

that is, it is not untrue to state that Java math is giving an incorrect result - and note also that he says "a java math error" not "a java.math error" which I think adds some ambiguity anyway)

I disagree. It's not a "java math" problem, because java math provides a way to prevent the problem. The problem only reared it's head because the coder chose to use a floating point.

If this was a financial application and that "bug" ended up costing the company millions of dollars, would you be OK with him blaming it on Java? I doubt it.

2014-02-12 Reply Admin

I actively encourage my clients to paste their screenshots into Word. Why? Because when you do that, Word preserves the scaling information and I can blow it up and see the thing they are trying to point out.

The alternative is that I get a screenshot of a full size window on a 24" monitor embedded into an email message using Outlook. Outlook doesn't preserve the scaling info and it's impossible to see the tiny message.

So apologies for creating a user base that does screenshots with Word, but it really is a useful tool. It's speedy to do and it uses tools the client is mostly familiar with. I do sometimes manage to teach them how to take a screenshot of the "active" window, but I'm happy enough to get a peek at the error message if it's legible.

chubertdev · 2014-02-12 Reply Admin

AN AWESOME CODER:
I disagree. It's not a "java math" problem, because java math provides a way to prevent the problem. The problem only reared it's head because the coder chose to use a floating point.
If this was a financial application and that "bug" ended up costing the company millions of dollars, would you be OK with him blaming it on Java? I doubt it.

Those are two separate issues. It's a Java math problem, but it's a known Java math problem. So the coder would be at fault for not being aware of something that just about every developer knows is broken.

chubertdev · 2014-02-12 Reply Admin

klc:
I actively encourage my clients to paste their screenshots into Word. Why? Because when you do that, Word preserves the scaling information and I can blow it up and see the thing they are trying to point out.
The alternative is that I get a screenshot of a full size window on a 24" monitor embedded into an email message using Outlook. Outlook doesn't preserve the scaling info and it's impossible to see the tiny message.

So apologies for creating a user base that does screenshots with Word, but it really is a useful tool. It's speedy to do and it uses tools the client is mostly familiar with. I do sometimes manage to teach them how to take a screenshot of the "active" window, but I'm happy enough to get a peek at the error message if it's legible.

Very good point. I wasn't trying to say all of those formats are wrong, just that there is a wide variety that they come in.

2014-02-12 Reply Admin

chubertdev:
AN AWESOME CODER:
I disagree. It's not a "java math" problem, because java math provides a way to prevent the problem. The problem only reared it's head because the coder chose to use a floating point.
If this was a financial application and that "bug" ended up costing the company millions of dollars, would you be OK with him blaming it on Java? I doubt it.

Those are two separate issues. It's a Java math problem, but it's a known Java math problem. So the coder would be at fault for not being aware of something that just about every developer knows is broken.

exactly. The error is within Java (and other languages) because of the implementation of floating point. Of course, just because the error is within Java, doesn't mean that the fault of any failure in an application that uses Java is Oracle's fault (in fact, I never mentioned fault in my original post, I just claimed that there is some justification to calling it a Java error). The liability will always be with the developer or at least the organisation developing and testing it (and so it should in most cases), but it's perfectly reasonable to refer to imprecisions in arithmetic as errors in the language that has them.

Let's look at it another way.... does 0.57 * 100 = 56.9999999? does 0.57 * 100.0 = 56.9999999?

If you claim that they don't then it's an error. If you claim that they do then you're the error.
Of course, it's a well known error so people should avoid it, but it's an error nonetheless (in fact the behaviour is even documented).

slightly aside, but I'd say this recommendation (by oracle) would absolve them from any issues caused by their error: float: The float data type is a single-precision 32-bit IEEE 754 floating point. Its range of values is beyond the scope of this discussion, but is specified in the Floating-Point Types, Formats, and Values section of the Java Language Specification. As with the recommendations for byte and short, use a float (instead of double) if you need to save memory in large arrays of floating point numbers. This data type should never be used for precise values, such as currency. For that, you will need to use the java.math.BigDecimal class instead. Numbers and Strings covers BigDecimal and other useful classes provided by the Java platform.

2014-02-12 Reply Admin

Very true.

Of course anything beats having a client email a phone photo to my very, very dumb phone which can barely handle SMS text messages. Killing that off without running down all my minutes is fun. Nonetheless I do appreciate the client's effort.

2014-02-12 Reply Admin

klc:
Very true.
Of course anything beats having a client email a phone photo to my very, very dumb phone which can barely handle SMS text messages. Killing that off without running down all my minutes is fun. Nonetheless I do appreciate the client's effort.

And I suppose I could learn to hit the "quote" button if I'm going to participate here. :P

2014-02-12 Reply Admin

swschrad:
I'm sorry, all comments must be in the form of a question...

Why?

2014-02-12 Reply Admin

swschrad:
I'm sorry, all comments must be in the form of a question...

Because.

2014-02-12 Reply Admin

mick:
chubertdev:
AN AWESOME CODER:
I disagree. It's not a "java math" problem, because java math provides a way to prevent the problem. The problem only reared it's head because the coder chose to use a floating point.
If this was a financial application and that "bug" ended up costing the company millions of dollars, would you be OK with him blaming it on Java? I doubt it.

Those are two separate issues. It's a Java math problem, but it's a known Java math problem. So the coder would be at fault for not being aware of something that just about every developer knows is broken.
exactly. The error is within Java (and other languages) because of the implementation of floating point. Of course, just because the error is within Java, doesn't mean that the fault of any failure in an application that uses Java is Oracle's fault (in fact, I never mentioned fault in my original post, I just claimed that there is some justification to calling it a Java error). The liability will always be with the developer or at least the organisation developing and testing it (and so it should in most cases), but it's perfectly reasonable to refer to imprecisions in arithmetic as errors in the language that has them.

Let's look at it another way.... does 0.57 * 100 = 56.9999999? does 0.57 * 100.0 = 56.9999999?

If you claim that they don't then it's an error. If you claim that they do then you're the error.
Of course, it's a well known error so people should avoid it, but it's an error nonetheless (in fact the behaviour is even documented).

slightly aside, but I'd say this recommendation (by oracle) would absolve them from any issues caused by their error: float: The float data type is a single-precision 32-bit IEEE 754 floating point. Its range of values is beyond the scope of this discussion, but is specified in the Floating-Point Types, Formats, and Values section of the Java Language Specification. As with the recommendations for byte and short, use a float (instead of double) if you need to save memory in large arrays of floating point numbers. This data type should never be used for precise values, such as currency. For that, you will need to use the java.math.BigDecimal class instead. Numbers and Strings covers BigDecimal and other useful classes provided by the Java platform.

The problem I have with saying it's a "java math problem" is that the problem isn't specific to Java at all.

It's like describing integer overflow as a "java numbers problem".

2014-02-12 Reply Admin

AN AWESOME CODER:
mick:
chubertdev:
AN AWESOME CODER:
I disagree. It's not a "java math" problem, because java math provides a way to prevent the problem. The problem only reared it's head because the coder chose to use a floating point.
If this was a financial application and that "bug" ended up costing the company millions of dollars, would you be OK with him blaming it on Java? I doubt it.

Those are two separate issues. It's a Java math problem, but it's a known Java math problem. So the coder would be at fault for not being aware of something that just about every developer knows is broken.
exactly. The error is within Java (and other languages) because of the implementation of floating point. Of course, just because the error is within Java, doesn't mean that the fault of any failure in an application that uses Java is Oracle's fault (in fact, I never mentioned fault in my original post, I just claimed that there is some justification to calling it a Java error). The liability will always be with the developer or at least the organisation developing and testing it (and so it should in most cases), but it's perfectly reasonable to refer to imprecisions in arithmetic as errors in the language that has them.

Let's look at it another way.... does 0.57 * 100 = 56.9999999? does 0.57 * 100.0 = 56.9999999?

If you claim that they don't then it's an error. If you claim that they do then you're the error.
Of course, it's a well known error so people should avoid it, but it's an error nonetheless (in fact the behaviour is even documented).

slightly aside, but I'd say this recommendation (by oracle) would absolve them from any issues caused by their error: float: The float data type is a single-precision 32-bit IEEE 754 floating point. Its range of values is beyond the scope of this discussion, but is specified in the Floating-Point Types, Formats, and Values section of the Java Language Specification. As with the recommendations for byte and short, use a float (instead of double) if you need to save memory in large arrays of floating point numbers. This data type should never be used for precise values, such as currency. For that, you will need to use the java.math.BigDecimal class instead. Numbers and Strings covers BigDecimal and other useful classes provided by the Java platform.

The problem I have with saying it's a "java math problem" is that the problem isn't specific to Java at all.

It's like describing integer overflow as a "java numbers problem".

the articale (and I) daid Java (math) error, not Java problem....

I agree that as an arbitrary statement it extends beyond Java and shouldn't be thought of as "the java problem" or even "the java error", but in the context of his email, where he's describing the problem to a client who (presumably) understands that their application is written in Java I'd say his statement is acceptable....

2014-02-12 Reply Admin

autark:
Regarding the bug report that reads as if it was written by a drunk schizophrenic - could it be George from The Chronicles of George is still alive and working in the IT industry? It must be him - how could more than one person like that exist on planet earth?

I sent some e-mail that looked like that bug report. For a short time I had Internet Explorer 11 installed on a Windows 7 PC, and I didn't know what it was doing to Outlook 2003.

When editing e-mails in Outlook 2003, sometimes I didn't notice that edits after the first autosave were lost. Later I did notice that some edits were getting lost, but when I repeated the edits and saved again, sometimes I didn't notice that the edits were lost again. And if that isn't enough, sometimes I did notice, but Outlook 2003 automatically sent the garbled messages before I had time to edit them again.

After gradually figuring out that Outlook 2003 was losing edits, I did several Google searches, and finally one Google search found the solution: Upgrade from Internet Explorer 11 to Internet Explorer 10.

Outlook Express had problems in its day too, but they weren't as bad as this.

2014-02-12 Reply Admin

For those that care or are interested (ie. no one), here's what C# does (Spoiler: it's the same as Java)

float f = .57f*100;
double d = .57*100;
int i = (int)(.57*100);

Console.WriteLine(f);
Console.WriteLine(d);
Console.WriteLine(i);

57
57
56

2014-02-12 Reply Admin

herby:
As for Wells Fargo, they seem to want their emails confirmed, so they include "bugs" in them. Unfortunately, my email client doesn't do these things, and I keep seeing the "your email address is wrong" when I go about my online banking. Maybe they will wake up someday about this. Otherwise their web site is one of the best. Not loaded with silly graphics (that take time) or much Javascript (which takes even more time).

Uh yeah, their web site is one of the best because they load their silly graphics in their e-mail instead of their web site? Did you submit a "bug" report?

2014-02-12 Reply Admin

mick:
I agree that as an arbitrary statement it extends beyond Java and shouldn't be thought of as "the java problem" or even "the java error", but in the context of his email, where he's describing the problem to a client who (presumably) understands that their application is written in Java I'd say his statement is acceptable....

Perhaps. But even the solution he described was a half-fix (rounding to fix rounding, instead of using a safer type), so I'm hesitant to give him credit of just being informal in his description.

I think the reason I reacted to that the way I did, though, is because statements like that are just fodder for those that want to say "that's why java is stupid and I use Brainfuck instead!".

I mean, it is stupid... but not because math is broken.

2014-02-12 Reply Admin

The scanned map reminds me of a customer I once dealt with. We had asked for a packet capture, and was sent a .BMP screen capture of a command window running tcpdump.

2014-02-12 Reply Admin

not Douglas R. Hofstadter:
Hi J.,
I have has an erogenous amount

Hi. Thanks for the reference.

2014-02-12 Reply Admin

AN AWESOME CODER:
mick:
I agree that as an arbitrary statement it extends beyond Java and shouldn't be thought of as "the java problem" or even "the java error", but in the context of his email, where he's describing the problem to a client who (presumably) understands that their application is written in Java I'd say his statement is acceptable....

Perhaps. But even the solution he described was a half-fix (rounding to fix rounding, instead of using a safer type), so I'm hesitant to give him credit of just being informal in his description.

I think the reason I reacted to that the way I did, though, is because statements like that are just fodder for those that want to say "that's why java is stupid and I use Brainfuck instead!".

I mean, it is stupid... but not because math is broken.

rounding to fix truncation of bad precision, not rounding of rounding.

It might be that he was dealing with an existing interface where a float was being passed and he was unable to modify the interface specification.....

...or maybe I'm clutching at straws.

2014-02-12 Reply Admin

mick:
Let's look at it another way.... does 0.57 * 100 = 56.9999999? does 0.57 * 100.0 = 56.9999999?
If you claim that they don't then it's an error. If you claim that they do then you're the error.

Huh? 56.9999999 is just as much an approximation as 0.57 is.

In fact it's every bit as much an approximation, when you count all the missing bits that would be needed to represent it exactly.

mick:
Of course, it's a well known error

But not well enough known (since so many people keep screwing it up despite all the tutorials that try to explain it) and not an error (though so many people don't understand that it's not an error despite all the tutorials that try to explain it).

Does 1.0/3.0 * 100.0 = 33.33333333? Is it an error that they're not equal? Is math broken?

Let alone the number of people who don't understand why 1 / 3 * 100 doesn't equal 33. Beyond some point, students do magically learn why this happens. One of the inventions in BASIC used to be that this was fixed for beginners, or floated for beginners, so that 1 / 3 * 100 did equal 33.33333333; and, um, 1 / 3 * 300 did equal 99.99999999, so I guess this industry is hopeless after all.

No wonder IBM was considering putting decimal floating point hardware back in some of their machines, after an absence of 50 years, despite its obvious inefficiencies. They know who their customers are. But 1.0 / 3.0 * 300.0 will still come out as 99.99999999 so it's still hopeless.

chubertdev · 2014-02-12 Reply Admin

Everyone knows that 100.0 / 3.0 = 33.3

2014-02-12 Reply Admin

chubertdev:
Everyone knows that 100.0 / 3.0 = 33.3

Yes, but an incredible number of people don't understand that they don't get that infinite precision from a computer, even one that stores representations in base ten.

2014-02-13 Reply Admin

Norman Diamond:
mick:
Let's look at it another way.... does 0.57 * 100 = 56.9999999? does 0.57 * 100.0 = 56.9999999?
If you claim that they don't then it's an error. If you claim that they do then you're the error.
Huh? 56.9999999 is just as much an approximation as 0.57 is.

In fact it's every bit as much an approximation, when you count all the missing bits that would be needed to represent it exactly.

mick:
Of course, it's a well known error
But not well enough known (since so many people keep screwing it up despite all the tutorials that try to explain it) and not an error (though so many people don't understand that it's not an error despite all the tutorials that try to explain it).
Does 1.0/3.0 * 100.0 = 33.33333333? Is it an error that they're not equal? Is math broken?

Let alone the number of people who don't understand why 1 / 3 * 100 doesn't equal 33. Beyond some point, students do magically learn why this happens. One of the inventions in BASIC used to be that this was fixed for beginners, or floated for beginners, so that 1 / 3 * 100 did equal 33.33333333; and, um, 1 / 3 * 300 did equal 99.99999999, so I guess this industry is hopeless after all.

No wonder IBM was considering putting decimal floating point hardware back in some of their machines, after an absence of 50 years, despite its obvious inefficiencies. They know who their customers are. But 1.0 / 3.0 * 300.0 will still come out as 99.99999999 so it's still hopeless.

Have you gone insane? There is no word of wisdom in this entire rambling.

1000.57=57 is not an approximation, it's an equality. 1000.57=56.999999999 is a falsehood. And 100/3 = 33.333333333 is also wrong, no matter how many 3's you write - barring you write infinitely many of them.

Finitely representing real numbers (no matter what your scheme) is going to introduce errors. The fact that this is unavoidable doesn't mean that the errors aren't errors. Moreover, in nearly every representation scheme of real numbers (certainly the floating point representation) you can represent some numbers exactly. In fact, you can represent 2^b numbers exactly, when using b bits. The number 57 is one of them.

2014-02-13 Reply Admin

anon:
Norman Diamond:
mick:
Let's look at it another way.... does 0.57 * 100 = 56.9999999? does 0.57 * 100.0 = 56.9999999? If you claim that they don't then it's an error. If you claim that they do then you're the error.
Huh? 56.9999999 is just as much an approximation as 0.57 is.
Have you gone insane? There is no word of wisdom in this entire rambling.
100*0.57=57 is not an approximation, it's an equality.

OK, I wasn't clear enough. The printed or displayed output from a computer, whether 0.57 or 56.9999999, is an approximation of the value stored in the computer's memory or register or disk drive.

anon:
100*0.57=56.999999999 is a falsehood. And 100/3 = 33.333333333 is also wrong, no matter how many 3's you write - barring you write infinitely many of them.

Yes, that's exactly[*] the reason why approximations are approximations.

[* Oops. Almost exactly ^_^ ]

Matt Westwood · 2014-02-13 Reply Admin

anonymous:
DWalker:
Yep. If someone takes a screenshot and pastes it into Word, that's fine with me. When I take a screenshot, I paste it into Irfanview and crop the part I want to keep, then save the file as (usually) a JPG.
I don't fault non-technical users for pasting a screenshot into Word.
The only right way to take a screenshot is to paste it into Paint and save as a PNG.

That's where I've been going wrong all this time. I've been saving it as a jpg, not a png.

It's easy to get elitist about what tools are being used. IN extreme cases it's "What, you mean you don't use Notepad++? What sort of neanderthal are you?"

Cue the "real programmers use" xkcd cartoon.

Matt Westwood · 2014-02-13 Reply Admin

DCRoss:
Bob:
Anon:
Erogenous amount? How about 80085?

819.80085

But why are they biq?

I think he probably meant to type: 816.80085

Matt Westwood · 2014-02-13 Reply Admin

chubertdev:
Orac:
Mark Bowytz:
mikeTheLiar:
TDWTF:
Apparently JAVA Math is Hard!
Apparently Java is an acronym.

JAVA = Just Another Vague Acronym
No it's not, people just yell every time they mention JAVA....and for good reason (I can picture someone with tourettes yelling JAVA all the time).

It's on par with Belg**m.

or Basingstoke.

tin · 2014-02-13 Reply Admin

DaveK:
Anon:
Erogenous amount? How about 80085?
What's so erogenous about "sboob"?

I take it you read thedailywtf on a calculator held upside-down? For the rest of us, it works up the right way.

2014-02-13 Reply Admin

Anon:
Erogenous amount?

True. Too many people would get a hardon from seeing such amounts in their bank accounts.

File under: Wolf of Wall Street.

As for emailing the map: Maybe they work with paper dossiers for every destination, so they'd normally photocopy (not print) the paper printout for "paper customers", and necessarily scan again the printout for "email customers".

I mean 'necessarily' not in the common broad sense, but in the narrow sense of '... given the employees capacities and limitations'.

Captcha: "Scan this again, Luke, it is your DISTINEO."

Zemm · 2014-02-13 Reply Admin

anonymous:
I was talking about the right way, not the way that's just fine.

I reckon the right way would be to use something like Jing that can save the PNG directly rather than go through the whole paste step.

Though just yesterday I got a screenshot in tiff format from a user of an ancient Mac! It was that file that made me realise I had Photoshop installed on my work computer, albeit an expired trial.

2014-02-13 Reply Admin

Steve The Cynic:
faoileag:
And we all are like that every once in a while. If I have a cold and need a painkiller, I'm happy to take a couple of aspirins. Do I have to know exactly how acetylsalicylacid works? No.
No, but it might be interesting to know how acetylsalicylic acid works...

Google is my friend, but when I have a flu I keep away so that he might not catch some virus.

2014-02-13 Reply Admin

da Doctah:
Last time I took aspirin I ended up spending six days in the hospital with anemia from a bleeding stomach. I can only assume it works by giving you something so much worse that you forget about the headache.

Aspirin has blood thinning/anti-clotting properties. Therefore, it can be helpful directly after a heart attack, but bad for you if you cut yourself or need to undergo surgery. Also, it is an acid. So if you have ulcers, it might worsen that condition while at the same time preventing a bleeding ulcer from stopping to bleed.

Best of Email: A Bad Opportunity, Bad English, Bad Math & More!

Leave a comment on “Best of Email: A Bad Opportunity, Bad English, Bad Math & More!”