The Daily WTF

Markus does QA, and this means writing automated tests which wrap around the code written by developers. Mostly this is a "black box" situation, where Markus doesn't look at the code, and instead goes by the interface and the requirements. Sometimes, though, he does look at the code, and wishes he hadn't.Today's snippet comes from a program which is meant to generate PDF files and then, optionally, email them. There are a few methods we're going to look at, because they invested a surprising amount of code into doing this the wrong way.

Today, we look at a simple bit of bad code. The badness is not that they're using Oracle, though that's always bad. But it's how they're writing this PL/SQL stored function:

The end of the quarter was approaching, and dark clouds were gathering in the C-suite. While they were trying to be tight lipped about it, the scuttlebutt was flowing freely. Initech had missed major sales targets, and not just by a few percentage points, but by an order of magnitude.Heads were going to roll.Except there was a problem: the master report that had kicked off this tizzy didn't seem to align with the department specific reports. For the C-suite, it was that report that was the document of record; they had been using it for years, and had great confidence in it. But something was wrong.Enter Jeff. Jeff had been hired to migrate their reports to a new system, and while this particular report had not yet been migrated, Jeff at least had familiarity, and was capable of answering the question: "what was going on?" Were the sales really that far off, and was everyone going to lose their jobs? Or could it possibly be that this ancient and well used report might be wrong?The core of the query was basically a series of subqueries. Each subquery followed this basic pattern:

There are things which are true. Regular expressions frequently perform badly. They're hard to read. Email addresses are not actually regular languages, and thus can't truly be validated (in all their many possible forms) by a pure regex.These are true. It's also true that a simple regex can get you most of the way there.Lucas found this in their codebase, for validating emails.

The title of this week's column is making me hungry.To start off our WTFreitag, Reinier B. complains "I did not specify my gender since it's completely irrelevant when ordering a skateboard for mydaughter. That does not mean it is correct to address meas Dear Not specified." I wonder (sincerely) if there is a common German-language personal letter greeting for "Dear somebody of unknown gender". I don't think there is onefor English. "To Whom It May Concern" is probably the best we can do.

Kirill writes:

The CEO of Delia's company retired. They were an old hand in the industry, the kind of person who worked their way up and had an engineering background, and while the staff loved them, the shareholders were less than pleased, because the company was profitable, but not obscenely so. So the next CEO was a McKinsey-approved MBA who had one mission: cut costs.Out went the senior devs, and much of the managers. Anyone who was product or customer focused followed quickly behind. What remained were a few managers handpicked by the new CEO and a slew of junior engineers- and Pierre.Pierre was a contractor who followed the new CEO around from company to company. Pierre was there to ensure that nobody wasted any time on engineering that didn't directly impact features. Tests? Wastes of time. Module boundaries? Just slow you down. Move fast and break things, and don't worry about fixing anything because that's your successors' problem.So let's take a look at how Pierre wrote code. This block of PHP code was simply copy/pasted everywhere it needed to be used, across multiple applications.

Python's "batteries included" approach means that a lot of common tasks have high-level convenience functions for them. For example, if you want to read all the lines from a file into an array (list, in Python), you could do something like:

Most of us, when generating a UUID, will reach for a library to do it. Even a UUIDv4, which is just a random number, presents challenges: doing randomness correctly is hard, and certain bits within the UUID are reserved for metadata about what kind of UUID we're generating.But Gretchen's co-worker didn't reach for a library. What they did reach for was... regular expressions?

We like trains here at Error'd, and you all seem to like trainstoo. That must be the main reason we get so many submissions about brokeninformation systems."Pass," saidJozsef. I think that train might have crashed already.

There are real advantages to taking a functional programming approach to expressing problems. Well, some problems, anyway.Kevin sends us this example of elegant, beautiful functional code in C#:

Greg was fighting with an academic CMS, and discovered that a file called write_helper.js was included on every page. It contained this single function:

Gretchen saw this line in the front-end code for their website and freaked out:

One thing I've learned by going through our reader submissions over the years is that WTFs never start with just one mistake. They're a compounding sequence of systemic failures. When we have a "bad boss" story, where an incompetent bully puts an equally incompetent sycophant in charge of a project, it's never just about the bad boss- it's about the system that put the bad boss in that position. For every "Brillant" programmer, there's a whole slew of checkpoints which should have stopped them before they went too far.With all that in mind, today we're doing a news roundup about the worst boss of them all, the avatar of Dunning-Kruger, Elon Musk. Because over the past month, a lot has happened, and there are enough software and IT related WTFs that I need to talk about them.For those who haven't been paying attention, President Trump assembled a new task force called the "Department of Government Efficiency", aka "DOGE". Like all terrible organizations, its mandate is unclear, its scope is unspecified, and its power to execute is unbounded.Now, before we get into it, we have to talk about the name. Like so much of Musk's persona, it's an unfunny joke. In this case, just a reference to Dogecoin, a meme currency based on a meme image that Musk has "invested" in. This is part of a pattern of unfunny jokes, like strolling around Twitter headquarters with a sink, or getting your product lines to spell S3XY. This has nothing to do with the news roundup, I just suspect that Musk's super-villain origin story was getting booed off the stage at a standup open-mic night and then he got roasted by the emcee. Everything else he's ever done has been an attempt to convince the world that he's cool and popular and funny.On of the core activities at DOGE is to be a "woodchipper", as Musk puts it. Agencies Musk doesn't like are just turned off, like USAID.The United States Agency for International Development handles all of the US foreign aid. Now, there's certainly room for debate over how, why, and how much aid the US provides abroad, and that's a great discussion that I wouldn't have here. But there's a very practical consideration beyond the "should/should not" debate: people currently depend on it.Farmers in the US depend on USAID purchasing excess crops to stabilize food prices. Abroad, people will die without the support they've been receiving.Even if you think aid should be ended entirely, simply turning off the machine while people are using it will cause massive harm. But none of this should come as a surprise, because Musk loves to promote his "algorithm".Calling it an "algorithm" is just a way to make it sound smarter than it is; what Musk's "algorithm" really is is a 5-step plan of bumper-sticker business speak that ranges from fatuous to incompetent, and not even the fawning coverage in the article I linked can truly disguise it.For example, step 1 is "question every requirement", which is obvious- of course, if you're trying to make this more efficient, you should question the requirements. As a sub-head on that, though, Musk says that requirements should be traceable directly to individuals, not departments. On one hand, this could be good for accountability, but on the other, any sufficiently complex system is going to have requirements that have to be built through collaboration, where any individual claiming the requirement is really just doing so to be a point of accountability.Step 2, also has a blindingly obvious label: "delete any part of the process you can". Oh, very good, why didn't I think of that! But Musk has a "unique" way of figuring out what parts of the process can be deleted: "You may have to add them back later. In fact, if you do not end up adding back at least 10 percent of them, then you didn't delete enough."Or, to put it less charitably: break things, and then unbreak them when you realize what you broke, if you do.We can see how this plays out in practice, because Musk played this game when he took over Twitter. And sure, it's revenue has collapsed, but we don't care about that here. What we care about are stupid IT stories, like the new owner renting a U-Haul and hiring a bunch of gig workers to manually decommission an expensive data center. Among the parts of the process Musk deleted were:

Because we still have the NWS, I learned that "A winter storm will continueto bring areas of heavy snow and ice from the Great Lakes through New Englandtoday into tonight." I'm staying put, and apparently so is Dave L.'sdelivery driver.Dave L. imagines the thoughts of this driver who clearly turned around and headed straight home."Oh, d'ya mean I've got to take these parcels somewhere!? in this weather!? I can't just bring them back?"

Abdoullah sends us this little blob of C#, which maybe isn't a full-on WTF, but certainly made me chuckle.

Eric writes:

As the saying goes, there are only two hard problems in computer science: naming things, cache invalidations, and off by one errors. Chris's predecessor decided to tackle the second one, mostly by accurately(?) naming a class:

Andrew worked with Stuart. Stuart was one of those developers who didn't talk to anyone except to complain about how stupid management was, or how stupid the other developers were. Stuart was also the kind of person who would suddenly go on a tear, write three thousand lines of code in an evening, and then submit an pull request. He wouldn't respond to PR comments, however, and just wait until management needed the feature merged badly enough that someone said, "just approve it so we can move on."

I have a feeling we're going to be seeing a lot of AI WTFerry at this site for a while, and fewer stupid online sales copy booboos. For today, here we go:Jet-setterStewart wants to sell a pound, but he's going to have to cover some ground first."Looks like Google are trying very hard to encourage me to stop using their search engine. Perhaps they want me to use chatGPT? I just can't fathom how it got this so wrong."

Sammy's company "jumped on the Ruby on Rails bandwagon since there was one on which to jump", and are still very much a Rails shop. The company has been around for thirty years, and in that time has seen plenty of ups and downs. During one of those "ups", management decided they needed to scale up, both in terms of staffing and in terms of client base- so they hired an offshore team to promote international business and add to their staffing.A "down" followed not long after, and the offshore team was disbanded. So Sammy inherited the code.I know I'm generally negative on ORM systems, and that includes Rails, but I want to stress: they're fine if you stay on the happy path. If your data access patterns are simple (which most applications are just basic CRUD!) there's nothing wrong with using an ORM. But if you're doing that, you need to use the ORM. Which is not what the offshore team did. For example:

Tim has been working on a large C++ project which has been around for many, many years. It's a tool built for, in Tim's words, "an esoteric field", and most of the developers over the past 30 years have been PhD students.This particular representative line is present with its original whitespace, and the original variable names. It has been in the code base since 2010.

There are a lot of cases where the submission is "this was server side generated JavaScript and they were loading constants". Which, honestly, is a WTF, but it isn't interesting code. Things like this:

Just because you get fired doesn't mean that your pull requests are automatically closed. Dallin was in the middle of reviewing a PR by Steve when the email came out announcing that Steve no longer worked at the company.Let's take a look at that PR, and maybe we can see why.

Decreasingly hungry thrillseekerWeaponized Fun has second thoughts about the risk to which they're willing to expose their palate."In addition to Budget Bytes mailing list not knowing who I am, I'm not sure they know what they're making. I'm having a hard time telling whether 'New Recipe 1' sounds more enticing than 'New Recipe 3.' I sure hope they remembered the ingredients."

I'm a JSON curmudgeon, in that I think that its type-system, inherited from JavaScript, is bad. It's a limited vocabulary of types, and it forces developers to play odd games of convention. For example, because it lacks any sort of date type, you either have to explode your date out as a sub-dictionary (arguably, the "right" approach) or do what most people do- use an ISO formatted string as your date. The latter version requires you to attempt to parse the sting to validate the data, but validating JSON is a whole thing anyway.But, enough about me being old and cranky. Do you know one type JSON supports? Boolean values.Which is why this specification from today's anonymous submitter annoys me so much:

Today's anonymous submitter spent a few weeks feeling pretty good about themselves. You see, they'd inherited a gigantic and complex pile of code, an application spread out across 15 backend servers, theoretically organized into "modules" and "microservices" but in reality was a big ball of mud. And after a long and arduous process, they'd dug through that ball of mud and managed to delete 190 files, totaling 30,000 lines of code. That was fully 2/3rds of the total codebase, gone- and yet the tests continued to pass, the application continued to run, and everyone was just much happier with it.Two weeks later, a new ticket comes in: users are getting a 403 error when trying to access the "User Update" screen. Our submitter has seen a lot of these tickets, and it almost always means that the user's permissions are misconfigured. It's an easy fix, and not a code problem.Just to be on the safe side, though, they pull up the screen with their account- guaranteed to have the right permissions- and get a 403.As you can imagine, the temptation to sneak a few fixes in alongside this massive refactoring was impossible to resist. One of the problems was that most of their routes were camelCase URLs, but userupdate was not. So they'd fixed it. It was a minor change, and it worked in testing. So what was happening?Well, there was a legacy authorization database. It was one of those 15 backend servers, and it ran no web code, and thus wasn't touched by our submitter's refactoring. Despite their web layer having copious authorization and authentication code, someone had decided back in the olden days, to implement that authorization and authentication in its own database.Not every request went through this database. It impacted new sessions, but only under specific conditions. But this database had a table in it, which listed off all the routes. And unlike the web code, which used regular expressions for checking routes, and were case insensitive, this database did a strict equality comparison.The fix was simple: update the table to allow userUpdate. But it also pointed towards a deeper, meaner target for future refactoring: dealing with this sometimes required (but often not!) authentication step lurking in a database that no one had thought about until our submitter's refactoring broke something. [Advertisement] ProGet's got you covered with security and access controls on your NuGet feeds. Learn more.

Paul's co-worker needed to manage some data in a tree. To do that, they wrote this Java function:

Denise's company formed a new team. They had a lot of low-quality legacy code, and it had gotten where it was, in terms of quality, because the company had no real policy or procedures which encouraged good code. "If it works, it ships," was basically the motto. They wanted to change that, and the first step was creating a new software team to kick of green-field projects with an eye towards software craftsmanship.Enter Jack. Jack was the technical lead, and Jack had a vision of good software. This started with banning ORM-generated database models. But it also didn't involve writing raw SQL either- Jack hand-forged their tables with the Visual Table Designer feature of SQL Server Management Studio."The advantage," he happily explained to Denise, "is that we can then just generate our ORM layer right off the database. And when the database changes, we just regenerate- it's way easier than trying to build migrations.""Right, but even if we're not using ORM migrations, we still want to write migration scripts for our changes to our database. We need to version control them and test them.""We test them by making the change and running the test suite," Jack said.And what a test suite it was. There was 100% test coverage. There was test coverage on simple getter/setter methods. There was test coverage on the data transfer objects, which had no methods but getters and setters. There were unit tests for functions that did nothing more than dispatch to built-in functions. Many of the tests just verified that a result was returned, but never checked what the result was. There were unit tests on the auto-generated ORM objects.The last one, of course, meant that any time they changed the database, there was a significant risk that the test suite would fail on code that they hadn't written. Not only did they need to update the code consuming the data, the tests on that code, they also had to update the tests on the autogenerated code.Jack's magnum opus, in the whole thing, was that he designed the software with a plugin architecture. Instead of tightly coupling different implementations of various modules together, there was a plugin loader which could fetch an assembly at runtime and use that. Unfortunately, while the whole thing could have plugins, all of the abstractions leaked across module boundaries so you couldn't reasonably swap out plugins without rewriting the entire application. Instead of making a modular architecture, Jack just made starting the application wildly inefficient.Denise and her team brought their concerns to management. Conversations were had, and it fell upon Jack to school them all. Cheerfully, he said: "Look, not everyone gets software craftsmanship, so I'm going to implement a new feature as sort of a reference implementation. If you follow the pattern I lay out, you'll have an easy time building good code!"The new feature was an identity verification system which called for end users to upload photographs of their IDs- drivers' licenses, passports, etc. It was not a feature which should have had one developer driving the whole thing, and Jack was not implementing the entire lifecycle of data management for this; instead he was just implementing the upload feature.Jack pushed it through, out and up into production. Somehow, he short-cut past any code reviews, feature reviews, or getting anyone else to test it. He went straight to a demo in production, where he uploaded his passport and license. "So, there you go, a reference implementation for you all."Denise went ahead and ran her own test, with a synthetic ID for a test user, which didn't contain any real humans' information. The file upload crashed. In fact, in an ultimate variation of "it works on my machine," the only person who ever successfully used the upload feature was Jack. Of course, since the upload never worked, none of the other features, like retention policies, ever got implemented either.Now, this didn't mean the company couldn't do identity verification- they had an existing system, so they just kept redirecting users to that, instead of the new version, which didn't work.Jack went on to other features, though, because he was a clever craftsman and needed to bring his wisdom to the rest of their project. So the file upload just languished, never getting fixed. Somehow, this wasn't Jack's fault, management didn't hold him responsible, and everyone was still expected to follow the patterns he used in designing the feature to guide their own work.Until, one day, the system was breached by hackers. This, surprisingly, had nothing to do with Jack's choices- one of the admins got phished. This meant that the company needed to send out an announcement, informing users that they were breached. "We deeply regret the breach in our identity verification system, but can confirm that no personal data for any of our customers was affected."Jack, of course, was not a customer, so he got a private disclosure that his passport and ID had been compromised. [Advertisement] Keep the plebs out of prod. Restrict NuGet feed privileges with ProGet. Learn more.

"Math is hard, especially timely math," explainsThe Beast in Black.

A Representative Line is a short snippet that makes you think, "wow, I'd hate to see the rest of the code." A CodeSOD is a longer snippet, which also frequently makes you think, "wow, I'd hate to see the rest of the code," but also is bad in ways that require you to look at the relationship between the lines in the code.I bring that up, because today's code sample is a long section, but really, it's just a collection of representative lines. Each line in this just makes me die a little on the inside.Belda found this:

Sawyer was talking with a co-worker about how their unique session IDs got created. The concern was that they were only five characters long, which meant there could easily be collisions.They started by looking at the random number generation function.

Knowing the kinds of readers we have here, I strongly suspect that if you drew a Venn diagram of "TDWTF Readers" and "TikTok Users" those circles wouldn't overlap at all. But TikTok is in the news, and because my partner uses TikTok, I'm getting second hand smoke of all of this, I think there's some interesting things to talk about here.If you've been avoiding this news, good for you. For a long recap, Ars can bring up up to date.. But as a quick recap: TikTok is owned by Bytedance, which is based in China, and subject to Chinese laws. TikTok, like every other social media company, is basically spyware, tracking your behavior to sell your eyeballs to advertisers. Over the past few years, all three branches of the US government have decided that the "Chinese ownership" is the problem here (not so much the spying), and passed a law to ban it unless a US company buys it. The whole thing has turned into an idiotic political football, with Biden saying that his waning days of the Presidency wouldn't enforce the ban anyway, and then the whole thing turns into a Trumpist political football as the incoming President is playing Calvinball and making decrees that he did not (at the time) have any authority to make in the first place.Because of this ban, TikTok ceased operating in the US on Saturday night, displaying banners discussing the ban, and appeals directly to Trump to undo it. On Sunday, TikTok came back up, now with banners thanking Trump for being ready to work with them.Now, I'm mostly not interested in commenting on the political aspects of this, and you're mostly not interested in hearing it. But for the record: this whole thing is stupid. The root cause of the problem is that the US has no real consumer privacy law, but fixing that problem would be bad for companies like Google and Meta. So, instead, we play Whac-a-Mole with apps, dressing up Sinophobia as a national security threat, and we dance around the 1st Amendment issues. And then the whole thing of a President just deciding to rewrite a law at his whim is terrifying if you like the world to operate according to predictable rules, and presages a rather awful next four years.What I really want to talk about is conspiracy theories. Because when TikTok came back up, it was suddenly flooded with "IT professionals" who were positing a dark conspiracy: during the downtime, Meta purchased TikTok and migrated all of TikTok's services into Meta's infrastructure. That 12-15 hours of downtime was just the right amount of time to do that switcheroo.Now, I'm not going to link to any of these videos, because a) as stated, I don't use TikTok, b) TikTok requires you to use the app to watch videos, so screw that, and c) these people don't deserve more views. So there's an element of "take my word for it that this is happening," but also bear with me- because this isn't really what this article is about.Now, I am not a Site Reliability Engineer, and have no interest in being one. But I've worked with large retailers building global backends for point-of-sale systems where they're handling every point-of-sale transaction in the world. So I have some thoughts about the idea that migrating billions of videos and hundreds of millions of user accounts over to Meta's cloud can be done in 12-15 hours of downtime.Which, for the record, TikTok mostly uses Oracle's cloud, so add that to the "I Hate Oracle Club" scorecard.Let's assume Meta purchased TikTok. Would it have needed to spend 12-15 hours down so that Meta could migrate TikTok to their datacenter? Of course not. What an absurd thing to say. As this (Instagram) video rightfully points out, the only people taking down a website for a migration are a dentist office in Ohio in 2007. TikTok is a social media service handling hundreds of millions of users and billions of requests- they're already distributed across multiple datacenters. While spinning up services on a new datacenter isn't a simple task, it's a task that they've certainly already built the tools for. As part of their demand management system, they simply have to have the ability to spin up new service instances quickly and easily- at the scale they operate, that's the only option.They're a massive distributed system. Adding new infrastructure nodes and mirroring your data elsewhere is a solved problem. All it really takes is time and the budget to run more infrastructure than you need to service requests during the migration.The real costs are that if you're running in a cloud, you're likely not just using it as a set of virtual private servers- you're likely using your host's infrastructure-as-a-service abstractions, and that means that you might be tightly coupled to their implementation of a variety of cloud services. The real costs are that you'd need to make code changes to actually support a new cloud provider. And that's definitely not happening in a 12-15 hour time frame.But this is a dumb conversation to have, because if we assume Meta bought TikTok: there's no need to migrate anywhere. In this scenario, Meta has the keys to TikTok's infrastructure. Whatever they want to do, they can just... do. Sure, it means paying Oracle for hosting, but TikTok is making money. It's a net win. Over the next months or even years, Meta could move TikTok services into their private cloud and perhaps save costs, but there's no need to migrate on a tight timeline.Okay, so with all that said, what an idiot I am, right? Here I am, arguing against people I don't know, who definitely aren't going to read this. I don't even like TikTok, and think every social media app is a step down from just plain old RSS feeds, because I am an old person. We're deep into "someone is wrong on the Internet" territory. So why did this drive me up onto the soapbox?Because hearing all this conspiracy mongering nonsense reminds me of an important truth: everything looks like a conspiracy when you don't know how anything works.If you don't know how cloud deployments work, TikTok's downtime can look like a conspiracy. If you don't know how election systems are designed, any electoral result you don't like can look a lot like a conspiracy. If you don't know how the immune system works, vaccines can look like a conspiracy. If you don't know how anything works, a flat Earth starts making sense.Realistically, no one of us can know how everything works. In an ideal world, we can defer to experts who do know how things work. But to defer to experts, we need to trust expertise.And as a society, trust in experts has been eroding. A lot of it comes from propagandists who want their ignorance to be valued at least as highly as expertise. Being loudly wrong about things is a great way to get attention and with that, money. Alex Jones made many millions being loudly wrong.But it's not just loudmouthed morons that are eroding the trust in experts. Experts can and have abused the public trust. The poster child for "Worst Person Ever" is Thomas Midgely, Jr., who lied to the public and created a gigantic public health disaster, then went on to create a gigantic environmental disaster (in his defense, CFCs destroying the ozone layer wasn't something he knew about, but he absolutely knew about the dangers of leaded gasoline).And even more than that, in a society where peoples' prospects look worse with each passing year, with entire generations deciding that buying a home and having children are just going to be out of reach, we have to ask: what good is it to listen to experts if it doesn't lead to good outcomes? When all the experts work for a big mega corporation and put their big brains to work figuring out how to turn my eyeballs into dollars, or are working for think tanks and government agencies captured by those corporations, what good are experts?All in all, it looks bleak. There's no easy fix for any of this. The systems which make expertise viable have eroded over the past decades, taken for granted. Public trust in just... everything has weakened. Fixing this requires broad social changes. A minor tech blog that focuses in the ways people screw things up is not going to drive broad social changes.But I think there's one thing I can drive from here, and it comes back to this one simple statement: everything looks like a conspiracy when you don't know how anything works.So, I'm going to put out this call: when you know how things work, share that. Share what you know! Share it on social media. Share it on your own personal blog. Share it in local meeting groups. Hell, share it on TikTok, because gods know, they need it.But also don't forget the flip side: when you don't know, be careful about finding conspiracies. When you don't know how something works, it might look like a conspiracy. But, frequently, it's not- you're just ignorant. And honestly, we should be as open about our ignorance as we are about our knowledge. We should know what we don't know, or at least know when we're stepping outside of our areas of confidence.So let me close with this: do you have a place you're sharing the things you know? Do you think it'd be of interest to our readers? Use our submission form, and use the subject/title "Reader Link". If I get enough interesting links, I may do a roundup of them.Tomorrow, we'll return to our regularly scheduled programming. [Advertisement] Utilize BuildMaster to release your software with confidence, at the pace your business demands. Download today!

Janet's company had a glut of work, and thus didn't have the staffing required to do it all. It didn't make sense to hire on any new full-time employees, so they went the route of bringing on a few highly paid consultants, specifically ones who specialized in one specific problem: talking to a piece of hardware purchased from a vendor.The hardware in question was a scientific which communicated over a serial line. This device provided a lot of data that represented decimal values, but that data was not encoded as an IEEE float. Instead, they used two integers- one for the data, and one representing the number of decimal places.So, for example, "555.55" would be represented as "55555 2".Now, in embedded devices, this isn't too unusual. It's entirely possible that the embedded CPU didn't even support true floating point operations, and this was just how they decided to work around that.When communicating over the serial line, the device didn't send the data encoded in binary, however- it did everything as text. This was arguably helpful as it meant a technician could communicate with the device directly over a terminal emulator, but it meant any software talking to the device had to parse data.Which brings us to the code written by the highly paid consultants. This code needs to take two 16-bit integers and turn them into a single decimal value. Let's see how they did it.

Casanova Matt swings for the fences."OKCupid (they don't capitalize the K, but I do, for propriety)must have migrated their match questions through Excel duringa recent site revamp. These answers should obviously be 1-2and 3-4, but maybe I could have 2 with Jan and 4 with Margaret(Mar to friends)."

Roger took on a contract to fix up a PHP website. During the negotiations, he asked some questions about the design, like, "Is it object-oriented or more procedural?" "No, it's PHP," said the developer.Which about sums it up, I suppose. Have some date handling code:

Keige inherited some code which seems to be part of a drawing application. It can load brush textures from image files- at least, sometimes it can.

David was integrating a new extension into their ecommerce solution, and found this un-representative line:

Grun works for a contracting company. It's always been a small shop, but a recent glut of contracts meant that they needed to staff up. Lars, the boss, wanted more staff, but didn't want to increase the amount paid in salaries any more than absolutely necessary, so he found a "clever" solution. He hired college students, part time, and then threw them in the deep end of Perl code, a language some of them had heard of, but none of them had used.It didn't go great.

Someone online said we run a Mickey Mouseoutfit. Angered beyond words, we consulted legal@disney.com and they threatened to findthat guy and sue him. So to anyone else who thinks this column is Goofy, you should know that the world's definitive authorities insistthat it absolutely is not.But these guys? This website actually iskind of goofy, according to resolutioner Adam R. who crowed "Someone forgot to localize some text for the new year!"

Antonio's team hired some very expensive contractors and consultants to help them build a Java based application. These contractors were very demure, very mindful, about how using ORMs could kill performance.So they implemented a tool that would let them know any time the Hibernate query generator attempted to perform a cross join.

Bejamin's team needed to generate a unique session ID value that can't easily be guessed. The traditional way of doing this would be to generate cryptographically secure random bytes. Most languages, including PHP, have a solution for doing that.But you could also do this:

Today's anonymous submission is a delightfully simple line of JavaScript which really is an archetype of a representative line.

Ted's company hired a contract team to build an application. The budget eventually ran out without a finished application, so the code the contract team had produced was handed off to Ted's team to finish.This is an example of the Ruby code Ted inherited:

Happy 2025 to all our readers. I can already tell thisyear's columns are going to be filled with my (least)favorite form of WTF, the impossible endless gauntlet offlaming password hurdles to jump over or crawl under. Please comment if you know why this week's column has this title and why it doesn't have the title Swordfish.Peter G. starts off our new year of password maladies with a complaint that is almost poetic.

Ash's company outsourced to an offshore vendor.This is an example of what they got back:

Twenty five years ago today, the world breathed a collective sight of relief when nothing particularly interesting happened. Many days begin with not much interesting happening, but January 1st, 2000 was notable for not being the end of the world.I'm of course discussing the infamous Y2K bug. We all know the story: many legacy systems were storing dates with two digits- 80 not 1980, and thus were going to fail dramatically when handling 00- is that 1900 or 2000?Over the past few weeks, various news outlets have been releasing their "25 years later" commentary, and the consensus leans towards this was no big deal, and totally fine. Nothing bad happened, and we all overreacted. There may have been some minor issues, but we all overreacted back then.So I want to take a moment to go back to the past, and talk about the end of the 90s. Let's go for it.via GIPHYIt's the End of the World as We Know It25 years on, it's really hard to capture the vibe at the close of the 90s. We'll focus on the US, because that's the only region I can speak to first hand. The decade had a "it was the best of times, it was the worst of times," aspect to it. The economy was up, lifted in part by a tech bubble which had yet to pop. The AIDS epidemic was still raging (thanks, in part, to the disastrous policies of the Reagan administration). Crime was down. The Columbine Shooting was hitting the national consciousness, but was only a vague hint of the future of mass shootings (and the past, as mass shootings in the US have never actually been rare). The Soviet Union was at this point long dead and buried, and an eternal hegemony of the US seemed to be the "end of history". On the flip side, Eastern Europe was falling apart and there was war in Kosovo. Napster launches, and anti-globalization protests disrupt cities across the country.Honestly, I feel like Woodstock '99 sorta sums up the last year of the decade. A music festival with a tradition of love and peace is wildly unmanaged and held in a hostile environment and devolves into chaos, violence, and sexual assaults.With the millennium looming, people were feeling weird. There was a very real sense that the world was coming to an end. Not literally, but the sense of a looming apocalypse of some king was inescapable. It's easy to be the rational one and say, "this is just an arbitrary mark on an arbitrary calendar, it doesn't mean anything", but the mass public and the zeitgeist at the time wasn't feeling rational.When you add the Y2K bug into the mix, people lost their goddamn minds.The Vibe of Y2KWe'll talk about the technical challenges of the Y2K bug, but honestly, I think it's less interesting than the vibe.What people knew was this: computers ran the world, and at midnight on December 31st, 1999, every computer was going to freak out and try and kill us. Don't take my word for it.Honestly, would anyone have cared if the Backstreet Boys climbed into a bunker in 1999? That feels like the end of their reign as the boy band of the moment. Even Dr. Dre, who is clearly trying to be reasonable, doesn't want to be on a plane that night. Christina Aquilera's mom told her not to use elevators.Or check this guy, who's less afraid of the technical problem and more "the social" one:It wasn't all panic, like this long segment from the Cupertino City Council:And certainly, hero of the site Peter, knew it was boring and dry:But NYE 1999, people were unplugging all their appliances so the computers in them wouldn't freak out. In the run up, people were selling books to help you prep. Survival guides abounded. Some of them took this as a dire warning about the dangers of global warming and sexuality in media.The public poorly understood what Y2K meant, but were primed to expect (and prepare for) the worst. From this distance of hindsight, we can see echoes of the panic in the response to the COVID pandemic- a very real problem that people wildly misunderstood and reacted to in all sorts of insane ways.The ProblemLet's get back to this idea of "some programs represented years with two digits". From the perspective of a modern programmer, this seems weird. It sounds like we were storing dates as stringly typed data, which would be a really silly thing to do.So I want to discuss the kinds of systems that were impacted and why. Because in the 90s, people thought their PCs might blow up at the changeover, but your desktop computer was never really at any risk. It was legacy mainframe systems- the big iron that ran half the world- that was at risk.To understand the bug, and why it was hard to fix, we need to spend some time talking about how these systems worked. Well, work, because there are certainly a few still in use.We're going to focus on COBOL, because I've had the misfortune to work with COBOL systems. Take my examples here as illustrative and not "authoritative*, because there are a lot of different kinds of systems and a lot of different ways these bugs cropped up.Now, as a modern programmer, when we think about representing numbers, we think about how many bits we dedicate to it. An 8-bit integer holds 256 distinct values.Mainframe systems used "flat file databases". As the name implied, data was stored in a file- just dumped into that file with minimal organization. A single application may interact with many "flat files"- one holding customers, one holding invoices, and so on. There were no built-in relationships or foreign key constraints here, applications needed to enforce that themselves. On a single mainframe, many programs might interact with the same set of files- the accounts receivable program might interact with invoices, the shipping supervisor would also look at them to plan shipping, an inventory management program would update inventory counts based on that, and so on.These interactions could get complex on any given system. And those interactions could get more complicated because multiple systems needed to talk to each other- so they'd need data interchange formats (like EDI or ASN.1).In COBOL, you'd describe your flat files with a "data division" in your program. That data division might look something like this:

The Hatter was framed! He didn't even do it! Nil Corpus Delecti, et cetera.Yet Yitz O., up to some kind of skullduggery, observed a spacetime oddity."When trying to compare some results from a GetOrders call via the ebayapi, I noticed something weird was happening with the DateTimes in the response.The attached is 3 calls to get the same order, made in quick succession.The millisecond part of all the DateTimes matched the millisecondpart of the *current* time (which you can see in the TimeStamp field.I assume it's because they rolled their own DateTime functionalityand are Getting a UTC time by subtracting the difference betweenthe local time and the UTC time, and one of those values doesn't havethe millisecond value in it, but it's the ebay api so who knows."Undoubtedly a bug that nobody ever noticed because they probablyjust ignore the millis altogether.