PLACEHOLDER ONLY Someone please write witty headline here

As the year rolls down to its inevitable conclusion, we're running out of Mondays. But have no fear, gentle reader – future Mondays will bring you further installments of Who, Me? in which Register readers share their tales of tech support gone wrong. On this particular Monday, meet a reader we'll Regomize as "John" who, it …

[Korev]

Until John's supervisor took a perturbed phone call: "Why does the website say 'Stinky Mac' at the top of the view?" the supervisor was asked.

While the boss tried to think of any good reason the website should proffer such a message to a user, the phone call got worse: "How does it even know everybody calls me Mac?"

So he wasn't the Apple of the boss' eye?

[Korev]

Yes I'm going I'll get my coat mac

[Detective Emil]

That would be the one with the factory-paired buttons, right?

[Paul Herber]

The buttons with round corners.

[The Dogs Meevonks]

Don't be silly, buttons cost extra, and they're 9.99 each. And for some reason you also need to purchase the button hole maker for a further 49.99 because 3rd party hole makers invalidate your warranty as does making holes in anything that we still own, because you only have a license to wear the mac on dry days.

[Wikster]

Only one button, sir.

[Korev]

A developer at a previous employer, once wrote some code that threw an exception saying "If this happens then speak to George" as he'd assumed that this would only ever be seen in debugging.

This was fine, until a customer with a sense of humour rang up asking to speak to George...

[Korev]

The self-same software once had a SELECT COUNT(*) statement missing the O out of count which has a much less amused customer complain...

[SVD_NL]

Protip: create a custom function that returns your boss's name instead of an error!

[Anonymous Coward]

I used to work for a large energy company that published a job advert to the internet for an "accocuntant".

Either HR didn't feel a need for proof reading or spell checkers, or perhaps they got exactly the candidate they were seeking, and through various promotions he's now Director of Strategy & External Affairs.

[Graham Newton]

On the subject of proof reading we had a technical author contracter write our user manual. She drafted the manual with questions embedded for the developers. The developers would answer them and she would amend the document accordingly. She then sent it to management for proof reading. They decided to forgo the proof reading and printed the manual which contained the comment. "Well you could do it that way, but anyone who thinks that is a good idea is a complete moron."

[heyrick]

"Well you could do it that way, but anyone who thinks that is a good idea is a complete moron."

Which may well be the most effective way to stop people doing it "that way". Neuter the idea before it's even begun.

[jj_0]

"contracter" - proofread pun intended?

[Bebu]

Pardon my sql :)

"SELECT COUNT(*) statement missing the O"

Probably a short for

select count(employee) from payroll

where payroll.payscale = 'C-Suite'

I can think of a few organisations where the count() aggregate function had better return at least a 32 bit unsigned to avoid an integer overflow for this query.

[jmch]

"SELECT COUNT(*) statement missing the O"

Very apt, I guess, that the "open bracket-asterix-close bracket" provided a visual representation of the "COUNT missing the O"

[bemusedHorseman]

...Mwah! Goodnight everybody!

yakkowarner.gif

[sebacoustic]

Re: Sudden Impact

You're going to need a need a bigger envelope for doing calculations on the back of.

Is little Timmy somehow fixed in space and manages to stop dead a 500kg bot doing a runner's pace while squishing just 25.4mm?

If Timmy _is_ fixed in space by having his back against a wall, there'd be little chance of survival.

In a more realistic "inelastic collision" crash, 15km/h is typically survivable, the deadly threshold being somewhere in between speeds being considered in heated welsh debates.

[mirachu]

*asterisk

;)

[John Robson]

I've had code which was heavily commented "Do not change this - it's more complicated than it looks. Talk to a, b, or c - or this line of management"

But the real question is - was george still there, or was there a new george, who worked in a different area (catering maybe)?

[Will Godfrey]

I've done a few "Warning: This is more complicated than it looks."

On every occasion someone has ignored this, screwed it up and then complained!

[ColinPa]

well change it back again

Someone wrote a lot of complex code,and wrote even more code for unit tests. This was handed on to another department. A while later she got an email saying the Unit tests were not working - and asked for help to get the unit tests to work.

She asked "have you changed the main code?" "Yes" "Well change it back again as you've broken it"

[Anonymous Coward]

Here be dragons code

I once spent a few weeks trying to understand a huge lump of poorly commented code left by a departed developer. I eventually put a comment at the top of it saying

// I am 99% sure this does nothing but I can't prove it so I am leaving it alone

I looked back at this code a year or two later and another line had been added

// I agree

It was linked in and there were code routes that would take you there but with the many layers of processing above and I was almost sure that no input could ever cause this code to run. Naturally this made it very tricky to verify.

[imanidiot]

Re: Here be dragons code

Comment it out in the next release and wait for the wails of anguish. If it remains silent, the code wasn't needed

[Kubla Cant]

Re: Here be dragons code

Quite apart from the morality of using a live system for testing, this doesn't prove anything.

The implication is that the circumstances that cause the dragon code to execute are hard to understand, and probably very rare. So it won't fail on the first release, while you're watching it, but it may fail several releases down the line, when your change is no longer evident.

[Killfalcon]

Re: Here be dragons code

Rare, or time-bound! I've seen code fail after forty years of continuous use, because that's how long it took for us to have three days in a row where a certain product didn't sell.

Imagine unpicking a 40-year old bug and finding that it was meant to be handled, but someone had commented it out because they didn't understand it.

[HelpfulJohn]

Re: Here be dragons code

The Y2000 experiment in coding was sort of this.

Lots of two-digit counts that were embedded in code that had worked for decades suddenly would have dumped when they found themselves not

incrementing but dropping back to zero.

Some four- and five-year projective codings had to be fixed well before the end of 1999.

It was all good, clean fun. Especially on 31/12/1999 when we waited for the I.C.B.M.'s to accidentally launch.

The entire event was down to us saving two digits when storing the year. In the 1980's no one thought that this could cause any problems.

By 1995, loads of us did.

Fortunately, we fixed most of it. No one launched. The Grid didn't collapse. We [obviously] survived.

Aren't I.T. guys great?

But has anyone thought of 9,999 A.D.? That deadline, too, is looming.

[Mike007]

Re: Here be dragons code

I'm waiting for the 19th January 2038, which I can guarantee there will be no major industry push to fix...

[John Brown (no body)]

Re: Here be dragons code

"If it remains silent, the code wasn't needed"

That's assuming anyone is left to deal with the wails of anguish when the "edge case" that needs it is triggered, eg only at year end, at on next Feb 29th which may 3-4 years away :-)

[F. Frederick Skitty]

Re: Here be dragons code

This brought back a suppressed memory. One company I worked at had a single C++ class that contained 30,000 lines of code. It was a horrific monster that displayed a table of data in a Qt table widget, but sadly the original authors has never understood the MVC pattern.

[Andy the ex-Brit]

I spent hours reducing some physical formulae (flow equations) in combination with some other processing to get a vastly simplified one that was just one line and took far less time to compute, but really wasn't recognizable. My comment was "trust me."

[Yes Me]

Irregular

IMNSHO that applies to all regular expressions, without exception.

[Killfalcon]

Re: Irregular

I always comment what a regex is meant to do. Usually before I even start writing the regex, in case I get confused partway...

[Korev]

But the real question is - was george still there, or was there a new george, who worked in a different area (catering maybe)?

<<Checks LinkedIn>>

Nope, but he was around for over a decade after that

[ColinPa]

Please speak to George

One of our developers also called George (perhaps the same one?) was a very careful developer. His code was

Select..

Case 0

Case 1.

Otherwise

/* There is no way we can get here. If you got here, phone George on...

*/

Write trace entry

Return unknown error

I think the problem was.. Originally the variable was a bit variable. This was upgraded to an int because there were now 3 valid values (yes, no, dont know) - not just 2

[Martin-R]

Re: Please speak to George

I once had to debug some misbehaving VBScript where something called "blnResult" turned out to be set, at different points, to True, False, -1, 0, 1 and (less understandaably) 2... I wonder if George had been near the same codebase?!

[Someone Else]

Re: Please speak to George

I've seen that code!

_{But there is nobody named George here. Someone here must have gotten it off of Stack Overflow...}

[Yes Me]

Re: Please speak to George

This is why certain theoreticians of computer science blather on about why strong typing is essential. It is a bit funny that -1 tests as True, but there you go. If you want to make it False, +=1.

[Will Godfrey]

Re: Please speak to George

I've learned (the hard way) to never test for true, always test for !false. That covers -1, 1 and ASCII 'Y' (yes)

Gets 'N' wrong of course - can't win them all.

[YetAnotherLocksmith]

Re: Please speak to George

I don't unstanded? My old boss wasn't called George!?

[BinkyTheMagicPaperclip]

Re: Please speak to George

On very specific occasions, your 'impossible error' can occur. Back in the mists of time a DOS antivirus package fell over with 'impossible error' on one specific system which had faulty memory and other components.

Fortunately in these days of protected mode operating systems, it will generally fail to boot or raise an exception before programs can be affected

[Doctor Syntax]

Re: Please speak to George

On very specific many occasions, your 'impossible error' can will occur.

FTFY

[Yet Another Anonymous coward]

Re: Please speak to George

One in a million errors occur 9 times out of 10

[katrinab]

Re: Please speak to George

Or, one in a million errors can occur several times per millisecond on a modern computer.

[heyrick]

Re: Please speak to George

Yup. Once had to deal with a bit of my code taking an impossible code path (that was, literally, not going to happen).

How? Something else somewhere else was off by one which screwed with the stack in just exactly the right sort of way that the program jumped from here to there like it fell into a wormhole, got confused, blew up.

That's when I learned that if something expects either TRUE or FALSE, be sure you can handle "hello" as an option (even if it's only ever called in a way that can give it only TRUE or FALSE).

Yes, it was loads of fun debugging that.

[spuck]

Re: Please speak to George

Just read this last week:

A software tester goes into a pub and orders 9,999 pints.

Another tester goes into the pub and orders -1 pints.

A third tester goes into the pub and orders jlakjwyd pints.

A user walks in and asks to use the men's room, at which point the pub explodes.

[Anonymous Coward]

Re: Please speak to George

I've seen that one. It made me piss my pants.

I immediately saved it to send to the head of QC at my client.

[Andy Landy]

Re: Please speak to George

there were now 3 valid values (yes, no, dont know)

ah, the old classic boolean... True, False and FileNotFound!

[imanidiot]

Re: Please speak to George

That's just fuzzy logic isn't it? Maybe hire the Muppets to help?

[Claptrap314]

Re: Please speak to George

You sure it wasn't FILENOTFOUND ?

[Killfalcon]

Re: Please speak to George

Early in my career, I ran into a problem with eComms (IBM mainframe terminal emulator) talking to VBA once. The gist, as I understood at the time, was that VBA thought "TRUE" is a byte value of 1 ('00000001'). eComms thought it was a bit with a value of 1 ('1'), and didn't fret too much about the rest of the byte, resulting in some odd intermitted problems where, sometimes, TRUE <> TRUE.

I fixed that by abusing casting, but if anyone happens upon my past life making terrible automation tools, I hope you're better paid for it than I was.

[Admiral Grace Hopper]

GEORGE (the OS) and George the man.

I had the pleasure of working on GEORGE 3 but I never met the man himself. I worked with his son, which was a good experience with him being a fine fellow and an excellent programmer.

[Don Bannister]

Some years ago, while using specialist development software written by a chap we knew, was rather surprised to encounter the error "Pointer f***ed" (and without asterisks !!)

[Yet Another Anonymous coward]

And it probably was

[Doctor Syntax]

Obviously not. You wouldn't want to raise f by ed

[YetAnotherLocksmith]

That should fail the testing process certainly!

[Arthur the cat]

At one place I worked there was a tricky bit of geometry code that was originally written in Fortran IV by someone with a grasp of mathematics but little programming style which would PAUSE 666 if it hit a case it couldn't handle. Some years later it got translated into C by someone who didn't understand the maths, and the PAUSE 666 statements got translated into an error message that was potentially user visible, although the guy doing the translation believed it would never actually occur. However, the code was being pushed a little more than it had been before, so not long after the C version of the software was released we got a complaint from a customer saying what does "one of Klas' hairy cases" mean?

[Anonymous Coward]

A media company I worked for had an idea for what we thought was the first use of a “Are you sure you want to leave this page” pop up/prompt. We rolled it out across our websites and it helped keep people listening to our streaming rather than closing the page by accident. Not long afterwards, under a month possibly we found a competitor was doing the same thing. Was quite chuffed that they’d thought the idea so good they’d pinched it. Less chuffed to discover that they’d also pinched our code off the website too. We presented our findings to the departmental head who asked if there was any chance this was a coincidence and someone had written the same thing. The lead developer said that there was no chance of that because they’d copied stuff that didn’t do anything much except identify the code as his.

H.O.D. said they knew the MD of the firm what had done it and would have a word. The conversation went along the lines of whilst we’re flattered that you liked our idea less flattered to find you’ve pinched our code to do it. By lunchtime the message didn’t come up on their website and was only replaced later with new code.

[Ken Shabby]

I’ve seen my own code appear with someone else’s name on it, sort of flattered and a little pissed off at the same time. Internal performance measuring software.

[Handel was a crank]

Tut

Should have used "rotten Apple" instead.

[trevorde]

For dev eyes only

Worked on a product where new functionality was put on a special menu so QA could test it early in the dev cycle. One of the devs implemented a feature & used some placeholder text until product definition decided what to call it. QA immediately contacted him asking: "Simon, what is a choad?"

[MiguelC]

Re: For dev eyes only

The problem with internal use errors is that they tend to get to the outside world... if there's a 1 in a million chance of it happening, it'll happen 9 times out of 10!

[sebacoustic]

Re: For dev eyes only

I added a "I'm feeling lucky" checkbox in a dialog for ourt internal software which can speed up your workflow a lot by performing checks giving only a pass/fail but not persisting any detail as to why. Works great because 90% of the time people already know it's going to pass.

This was supoposed to get a "proper" name before it went into production but of course did pop up before users' eyes. We were informed that is "unfprofessional", and the checkbox is now called "Quick mode" and nobody understand what it does.

[Anonymous Coward]

Re: For dev eyes only

I've been cautioned for using the names of 1970's porn star's names for input validation tests... not so good when a client is religious...

They are much better, IMHO, than using Rob, Robert, Bob, and Bobby Tables as test names.

[pfleming]

Re: For dev eyes only

If they are so religious how do they know the porn stars names?

[Evil Auditor]

"...never put anything on there that you don't want a user to see."

That's also true for, e.g., drafting a report that will ultimately be sent to a client. The draft, when sent to the client, contained our comments addressed to the client and also my snarky comments about the client not meant for the client to see. While I got sweaty palms when discovering this, it didn't have any negative consequences for me. I assume, the client simply thought that I'm socially challenged (aka a bit of a prick) and delivered what I needed.

[Anonymous Coward]

Re: "...never put anything on there that you don't want a user to see."

Some years ago we were given a presentation from a vendor trying to sell us the newest generation of one of their product platforms that we were using. We were told that this was an exclusive preview, and that we were their only customer worldwide who were seeing this documentation ahead of the official release. Strangely enough, there was another company name in our sector (from another country) in the header of every page of the design document.

(We bought the platform, we also bought the suddenly-needed upgrade and rebuilt our data center to support the additional capacity that the vendor didn't anticipate we'd (and needing an upgrade from 3 to 18 processing servers isn't exactly a rounding error...). Then another six months passed after which we tore up the contract and threw the platform out as it still didn't have the capacity or performance needed .)

[Anonymous Coward]

Re: "...never put anything on there that you don't want a user to see."

On a couple of occasions, I have send an 'internal consumption' email to a particular boss with questions or comments about a client. Nothing offensive, but definitely not what I would have written if it was intended for the clients eyes.

He of course just forwards it directly to the client.

while there was no major blowback, I am now much more careful about what I put in emails to him.

[Anonymous Coward]

Re: "...never put anything on there that you don't want a user to see."

I've worked for that bastard...

[Anonymous Custard]

Re: "...never put anything on there that you don't want a user to see."

Haven't we all...?

[wolfetone]

Could've been honest

John could've been brutally honest with his placeholder.

[Pete 2]

nothing here

> never put anything on there that you don't want a user to see

Such as the two rules of guaranteed success:

1.) Never teach others everything you know

2.) See rule #1

[Bebu]

Chance would be a fine thing

《Such as the two rules of guaranteed success:

1.) Never teach others everything you know

2.) See rule #1》

Exercises in futility:

A. trying to teach others anything.

B. encouraging others to learn or study something.

(TWoFT)

The only way of rescuing failure from the jaws of success is to ensure the negation of A and B.

[Anonymous Coward]

Re: nothing here

Happiness is being made redundant, from a major bank, with a very decent payout, because they assumed the email system I built didn't need looking after.

They assumed, incorrectly, that it didn't need to be managed because it just worked. Didn't need a peon like me looking after it.

Took a very generous redundancy payout and was in Thailand when it went completely titsup the first time - I so very much enjoyed the panicked phonecalls while sipping Mai Thais on the beach at Pattaya.

One of the first major problems I did actually fix, because I could still get into the system - they only deleted two of my admin accounts but missed the important ones ;-)

Then I had some worthless widget in IT Sec complaining I hadn't raised a change control!

Me: "I don't fcucking bas*rd work for the bank anymore and if you give me any more $hit, neither will you".

Got fewer phone calls after.

[A.P. Veening]

Re: nothing here

on the beach at Pattaya.

Knowing the beach there and how it has eroded since 2000, I highly doubt that unless it was way before 2000. There are some very nice beaches near Pattaya though, so where were you really?

[Doctor Syntax]

Re: nothing here

Maybe the payout had blinded you to them - either that or you hadn't discovered consultancy fees.

[trevorde]

Taking the hint

Worked on a product which was used by military, police, CIA, FBI, MI5, MI6, GCHQ et al. We added some "social network analysis" functions & I used some placeholder text for a hover over hint. Product manager thought that "Work out the most important person & who to kill next", whilst accurate, should probably be revised.

[wolfetone]

Re: Taking the hint

"Work out the easiest one to pin the blame on and kill them" instead?

[El blissett]

Title Optional

Type your comment here - basic HTML and hotlinks allowed.

Yes, I've been in this situation and there's few others that have made my blood run quite as cold. It's always a customer you don't have any problems with who ends up on the recieving end of it so you just have to leave it in your head (and certainly not in comments).

[b0llchit]

FIXME

FIXME: ...Add comment here...

[Caver_Dave]

Avoid

I've always in the past placed messages (using what ever method was appropriate to the language and situation) that said

"Unexpected path [nnnn]. Please report to your local IT or directly to software supplier."

where nnnn was a #defined number (or similar) carefully documented in the source code.

It doesn't have any user scary flags like the words error or warning, but does ask that it is reported.

A simple number can usually be conveyed without error, from whence the source code comment will explain what has happened and where the message was raised.

Now I'm writing certifiable code, there is obviously no need.

[G.Y.]

impossible errors Re: Avoid

The PL/M-51 compiler, when it hit a "can't happen" condition, would produce an error message consisting of the compiler pass and the code address

[stiine]

Re: impossible errors Avoid

Been there. Done that.

That's when you have to get hardware support in to check the physical machine(s) because something's probably loose.

[Red Eyes]

I n the late 80's I used a CAD/CAM program which I used to construct 2D drawings based on a customers paper drawing ready for generating gcode output for our CNC lathes. Contruction of the drawing was done through a command line interface. A line at and angle needed the prefix 'B' for bearing, so for a 45 degree line the command was 'B45' or 'B60' for sixty degrees, a horizonal line was 'B0', that is zero not an 'O', should you type 'BO', the screen cleared then filled up with the text ' You are stinky, have a wash'. You had to crash the program to stop it.

I was not best pleased on several occasions I did type BO

[imanidiot]

fun gags like that are fun the first time you encounter them, ONLY if you can then easily get on with your work. Anyone who puts stuff like that in production code so that it actually breaks your workflow... They deserve a keyboard with two letters that are always slightly sticky, no matter how much they clean it.

[Doctor Syntax]

Before I took over the job somebody playing with the then new, shiny RDBMS had added a test table for the engineers' cars. When it came time to migrate I decided that it wasn't a table that was going to survive but I took a look at what was in it. One car's colour was down as "Baby shit yellow".

[RockBurner]

My first car was a BMW 316 in that colour....

[Huw L-D]

The outside of my favourite bar is painted in what they lovingly refer to as "dog piss yellow".;

[UK Jim]

Beware of tag names

A friend related the story of a bank whose "high net worth" client group managed to send out an email to all of these important clients that started

"Dear $RICH_BASTARD",

...

(syntax may be wrong, but perhaps that was what happened!)

[Benegesserict Cumbersomberbatch]

Re: Beware of tag names

obligatory XKCD

[Duncan10101]

All those other exceptions

A colleague of mine was once writing a program in C++. It was a UI, so you can see this was a while ago. Anyway, unhandled exceptions tended to cause the whole app to exit, so he'd done everything possible to trap them, including the dreaded "catch (...)" just-in-case.

Now, once you're in the "..." block there's not a lot you can do, because the type information is gone, and if you'd been able to handle anything specific you would have done that above the "..." block anyway (*). Basically, he was giving the user an opportunity to not lose their work, despite his program having gone TITSUP. What did he do in that block? Put-up a message box, of course. There's noting useful you can really say in the message, because you don't know what's gone wrong, but he unfortunately chose to write "Bollocks". OK so it was a bad choice of message, but "Something went wrong" doesn't actually improve the user experience at all. Perhaps he should instead have chosen "Bollocking" ... because, of course, it appeared at a critical moment, and that's exactly what he got.

(*) yeah yeah, you can throw it again in some sort of handler-function that might be a bit more sophisticated, but let's not quibble. It wasn't my program.

[Yet Another Anonymous coward]

Re: All those other exceptions

Writes "round objects" on a report

Report comes back "who is round, and to what do they object?"

ObYesMinister

[CorwinX]

Re: All those other exceptions

Akin to A Round Tuit... As in I'll fix your problem when...

[Anonymous Coward]

Re: All those other exceptions

My mother-in-law had a wooden disk with TUIT written on it that she would offer to someone who used that phrase.

[Anonymous Coward]

Re: All those other exceptions

who is this General Waste and why does he keep leaving his inboxes around our office?

[Alister]

Not quite on topic, but we run backend systems for ticket purchases on a number of transport links, where customers can book a ticket and receive an email with a barcode or QR code which allows them to travel. Associated with the barcode is a randomly generated 12 character reference code used to store and refer to the transactions.

We do have some rules in place to sanitise the reference code, but on this occasion the system beat us.

We had a complaint, demanding that we refund the ticket price because of the rude word...

The ticket reference code was UrAWw4nK3rB8

I think the computer was right.

[Ochib]

Had a script that created users and a randomised password that was a Name of a Fruit, a Colour and an Animal

All worked fine until we had a complaint from HR and had to remove the following from the look-up table. Banana, Black and Monkey.

[Pete Sdev]

I sometimes swear in my comments. On one project many years ago I even had a small script that counted (grep and wc ) the quantity of swear words in the source code and record the daily value so I had a metric - "this sprint was tough, sweariness rose 19%").

One day I suspect my fondness of good old fashioned Anglo-Saxon will also bite me in the backside, though I'm more cautious with HTML and Javascript comments (even though JS is often a source of my swearing).

[Groo The Wanderer]

That would have been a guaranteed firing at any of the places I worked that were doing code reviews, and it would have gotten you blacklisted if it were discovered after you left, so you wouldn't get repeat business.

I've always tried to stick to somewhat more banal "eggs" in code, like having the system play a toilet flush when an uncaught exception is detected.

[Pete Sdev]

As it was closed sourced and my then boss, who had a similar sense of humour, knew about it and was amused, it wasn't a problem.

For exception/error messages I stick obviously to the banal informative.

[breakfast]

"Half the time when I'm writing Javascript I want to shout 'this is bullshit!' but then I realise I don't know what _this_ is."

[Philo T Farnsworth]

Laying my own "Easter Egg"

I have a similar story to tell on myself.

I don't use Macs on a regular basis -- I'm a Linux user and run my own mail server (that will be relevant later) -- but my partner does use a Mac and I need to deal with them on occasion to help out with system adminstration, etc.

For some reason that I don't recall, I needed to buy and install something from the online Apple Store whilst she was away and I had to sign up for an Apple ID in order to do so.

I recall the process being somewhat frustrating and when it came to the point where I needed to provide an email address, I used something on the order of '[e-mail]'. If you're not familiar with the 'postfix' mailer, everything after the '.' is ignored, so the email will be properly directed to my pseudonymous '[e-mail]' account.

Fast forward a couple of years and I'm in the bricks and mortar Apple Store, buying my partner an iPhone for Xmas. When I presented my credit card, the Apple ID popped up.

Raised eyebrows.

Fortunately, (a) the clerk had a good sense of humor and just gave me a good natured hard time about it and (b) I didn't use some of the more vulgar words I had been thinking when I signed up.

[Pete Sdev]

Re: Laying my own "Easter Egg"

Do live in the EU? Because I'd be curious of the justification of storing your credit card number (hopefully as a hash) linked to your email address for *years* unless you gave explicit consent.

[Philo T Farnsworth]

Re: Laying my own "Easter Egg"

I live in California.

And, now that you mention it, I did find the linkage a more than a tad creepy.

[OldSod]

Re: Laying my own "Easter Egg"

The "AppleID" is now Apple's "one ring to rule them all" for a user; it is how an Apple user's stuff is associated with that user no matter what platform they are on (iPhone, laptop, desktop, etc.) in space and time (i.e., switch from your laptop to a desktop, or go out and buy a new iPhone, then log in with your AppleID and all your stuff is there magically). But one of its original (if not *the* original) primary purposes of the AppleID was to identify the user for iTunes music sales/playback (DRM content), which was expanded to other sales purposes like the App store. As such, a "payment method" is part of the account record. If I recall correctly, there was a point in time where a valid payment method was a required element (it is now optional). I used an iTunes gift card rather than entering my credit card back then. I doubt the credit card number was hashed, because it was used every time a track, album, or app was purchased. Explicit consent seems to be met by the user entering it as part of creating the account.

[Anonymous Coward]

pr0n

Many years ago we were commissioned to provide a system that would manage and match a large number of digital facial images....

Suffice to say that one of the dev team had a multi gigabyte (terabytes were not a thing then) collection of images, which were used to test and train the data, and this was always part of the behind the scenes fun of the development.

We had a presentation to do. The presentation PC crapped out so we used the installed version on one of the devs laptops, which was connected to the dev database, and may have raised as many eyebrows as it did questions about our testing methodologies.

[Anonymous Coward]

Maybe not the best idea to use your full name as a place holder?

Or: how to use LinkedIn to troubleshoot

A customer called in and reported that he unable to change the system contact details in the web interface of his "Intelligent PDU" - whatever he entered, it always reverted back to the same incorrect setting. We suspected user error of course - but this man was convinced he'd seen the exact details in another installation for another client.

In comes LinkedIn: two usernames were given in the customer screenshot, one led nowhere, the other had a single hit: a recent graduate of a US university, who reported he was a "software intern", coincidentally in the months just before the firmware release.

With this information I was able to hand it over to engineering, with just an advice to search for the name in the source code, and pretty soon they came back with the cause - this forgotten routine was triggered by a timeout in internal network communication, to be fixed in next firmware release.

If the culprit is still reading this, and wondering if I am talking about him: are you still playing lacrosse?

[IceC0ld]

Re: Maybe not the best idea to use your full name as a place holder?

upvote as it amused me, I am SO not the culprit who plays Lacrosse though, posting that to make it clear, NOT that guy :o) I mean, Lacrosse, c'mon

[Paul Hovnanian]

Screen Animation

Years ago, I was responsible for supporting some shop floor automation for a small local aircraft manufacturer. Which included touch screen interfaces. Which some of the mechanics were in the habit of operating with whatever tool they had in hand.

As a hint, I developed (but never released into production) a UI that would randomly generate an animated crack originating from the touch point. Showed it to the boss on our development host. He chuckled.

The next day, a representative from the FAA stopped by to review the shop floor processes. I was asked to demonstrate the automation interface. So I went to the development terminal, brought up the factory interface and tapped an icon. A large crack appeared across the screen. The FAA rep chuckled.

[IceC0ld]

if you HAVE to leave a placeholder, at least make it cryptic :o)

TITSUP

they may wrinkle their brow, but YOU will know that :-

T - oday

I - t's

T - he

S - tinky

U - pgrade

P - rotocol

worth a shot LOL

[Boris the Cockroach]

We have data tables

for storing machine data (of course)

Now generally speaking although 10 pages of data are available , everyone generally just uses the first page.

At a previous job (hinted at last week) , the boss calls me over and starts on about "who's written this about me?" ... you see he'd scrolled down to the last data page (god knows why) and the entries read like

"<the boss> is a useless self pleasurer" and

"<the boss> has a small manhood" and

such and such for another 10 lines (although much much ruder and completely unprintable in a decent publication such as this one)

"Not me" I says, then remembers "Did'nt the last guy to do my job part with you on not so good terms?"

[CorwinX]

I didn't do that much coding...

More sysadmin than dev but always made sure there was a trapdoor in my code for any truly ridiculous event that happened.

Print [You are in a twisty maze, all the passages look the same, choose a direction]:

U D L R S

S for "Sideways" being the correct choice of course.

[tin 2]

"the horror that was nested tables inside nested tables inside nested tables"

Oi!

(returns to web app created in 2022 that does exactly that.... and is nice and responsive with it TYVM)

[Anonymous Coward]

but how well does it work with screen readers for the blind? (Those things were awful with tables back in the day, might have improved now I guess, I haven't checked)

[DS999]

XXX should not be here

Back in the late 90s one of the guys working for me was having problems with a Perl script that did some sort of user facing system function - I don't recall the details. I did not and still don't do Perl, so I wasn't able to contribute anything to solving the actual problem, but I suggested he do what I typically do with debugging stuff that seems to be doing something "impossible". Insert some diagnostic messages.

He pointed to a particular function filled with Perl gobblydegook and said the problem has to be here, but the logic shouldn't allow it to get to where it was printing garage data. I grabbed the keyboard and added a line to print "XXX should not be here" and told him to run it again. He tried it with various inputs, and was eventually able to get it to print that message. He started in on modifying the print command I'd added to print the value of a few variables to get more information, meanwhile my phone rings.

It was one of the math professors, who says "the computer is accusing me of having porn". I said there isn't anything we are doing to detect porn, it might be sort of web ad but he insisted it was not an advertisement it was "us" telling him that. He wasn't the best at providing useful diagnostic information, obviously, so I walked down the stairs to his office and I see "XXX should not be here" on his screen running the software we were debugging. Apparently my guy was looking at the public copy of the script and didn't tell me that when I modified it, then the math prof used it at exactly that moment.

I explained to him what that meant, everyone had a good laugh, and we all learned to be more careful to insure we aren't modifying the public copy of something during testing/debugging (and to make sure I told him to take the file ownership of all those Perl scripts out of our admin group, so that we'd have to become root to update the public copy)

[LessWileyCoyote]

I have posted it before, but it's worth another airing. Working on a large government system, a colleague eventually moved on to pastures new after complaining about the impossibility of getting a detailed spec from the systems analysts. One of his programs later caused some excitement by giving users the response "Error 4338 - some day they're going to specify this message".

[Kevin McMurtrie]

I worked on an online photo and document rendering codebase many years ago. The original photo rendering server was some really rough C. It wasn't thread safe, it didn't use memory efficiently, and its error handling was exit(-1). I started converting it into C++ for the bitmaps and Java for the rendering pipelines and storage abstractions. One thing that really needed improvement was diagnostics. I added a bit of Java code that would watch the JFIF data stream coming out of the C++ and inject a ton of diagnostics as a JPEG comment when the right location was reached. This was perfect. Customers would call about a bad image, we'd find it in the cache, extract the diagnostics, track down the bug, and purge everything that was corrupted. Even if the cache expired, they could send us the image.

Now it starts getting weird. When I Googled for the company's name, I see other photo web sites with my diagnostics dumps. People had been scraping the web site and posting their photos elsewhere. The diagnostics comment became the photo caption, and that became search results. This was happening a lot. 50% hilarious, 50% scary.

[aks]

Unintentional Internationalization

Back in the 1990's I was retrofitting Internationalisation to software that was already being sold in various parts of the world but needed to support the more exotic locations such as Japan.

A simple way for me to test my code was to have loadable language tables and the Welcome screen for the USA said "Howdy, partner".

Somebody (not me, honest) decided to add my new code tables into the next delivery.

Unsurprisingly, there were some faxes flying once the latest version got installed.

Happily, this was a quick fix and the final produce worked correctly for 29 languages, including Welsh.

[Anonymous Coward]

Plagiarism

We have a manager at work who plagiarises everyone elses documents. Because we have other incompetent idiots above him he continues to get away with it. Now AI is coming along, no doubt he'll be using that a hell of a lot.

One such document he took credit for he left the meta data in it, as he's an idiot. Came from his wife's place of work and had their company name on it. And as predicted, higher ups still never spotted this.