Bullshit Software Projects
earthly.dev
earthly.dev
Unfortunately, "bullshit project" and "highly visible project" are not mutually exclusive opposites.
Some of my biggest bullshit projects have been projects that had several prominent eyes watching them closely, demanding frequent status updates and demos (the subject of which was not always articulated).
Nothing worse than working on a project that you feel is doomed to fail or at least fail to meet overly lofty expectations, yet still feel entirely responsible to try to pull something, anything off.
Also, never doubt the ability of managers and even less skilled developer-leaders to concoct improperly defined software projects to solve poorly defined problems.
I've seen, multiple times, management say "I have problem {X}, and I think a project doing {Thing A} will solve it", only for the same management to have "problem drift" such that "problem {X}" morphs into loosely-related "problem {Y}", which "{Thing A}" won't actually solve, but too late -- management is now convinced that problems {X} and {Y} are virtually the same thing, and {Thing A} which was poorly conceived to begin with, can now definitely solve {X} AND {Y} if only you'd read their minds and connect whatever vague dots they have floating in their heads for them and built the proper {Thing} for it".
Wow... this comment turned out to be a rant.
tldr; I somewhat relate to this article.
There's never enough time to actually design anything, you have to make sure you have something to show by the end of the day.
It would be like an architect and builder assembling a shoddy movie set facade of a house without a proper foundation or framing, just to please a client who expects to see visible, street-facing progress on a steady and consistent basis.
Perhaps it falls to engineering to educate management on setting proper expectations, and creating a proper timeline with milestones.
But much like an employee at Tim Horton's isn't often in a position to tell the boss that time spent physically recovering will be far more useful than cleaning an already clean floor, we're often put in a position of enabling improper client and management expectations, with an implied "or else".
This has been a consistent recipe for burnout for me. Don’t do it. Nothing will change because “you’re not paid to think about these things”. You’ll spend your energy talking to a wall and become a target because you’re threatening the status quo.
Better spot this kind of environment early and leave as soon as you get a whiff. It’s not worth your health and they’re not interested in improving anything.
Things usually get that way for a reason and that reason is that no one who matters is interested in fixing things. And those who suffer these inefficiencies usually have learnt helplessness and made themselves ok with the situation (or left after the polite amount of time, 6–24 months depending on the country)
Even then, the chances of the org being open to change are very slim as we both seem to have experienced. Do you really want to put your health on the line for a small chance at fixing a company that’s not even yours and isn’t very likely to reward you for the waste reduction anyway? I know I don’t and quit rather than try to fix.
This project also was planned to have a Phase I and Phase II, and they ended up eating all the budget and time in Phase I and ... I guess Phase II wasn't really all that important?
Usually it's planned to build on whatever foundations were built in Phase I. But by the time you actually get to Phase II, it's morphed into something radically different and incompatible with Phase I. So you have to tear apart and rebuild all the previous work as well as you can while simultaneously trying to build the new stuff - on a tighter deadline because you're already over budget and time.
And if you don't burnout, the only thing you have to look forward to is a new Phase III, which is somehow even more unfeasible.
I've found it worth building a literal picture or map of what the hell is motivating these managers and decision makers to persist with the bs. So often it's schedules or money or hobby horses or vested interests, egos, sunk cost fallacy. Time and again bs proliferates because bs begets bs.
Fight bs by separating out the concerns into standalone problems, and season with a healthy dose of yagni.
My first day on the job, I worked pretty fast. I collected all the trash I could find, and when I was done, I sat down for probably 10 minutes trying to figure out what to do next. We were few people, and the worksite was pretty big.
During lunch, I get a call from my supervisor at the temp agency - she sounds worried and asks what is happening? Apparently one of the construction managers had briefly observed me sitting down, and immediately called my supervisor to inform her that I was not working, and that this arrangement was not working out. They wanted to booted off, and replaced with another temp worker ASAP. We agreed that I'd work "extra hard, and walk the extra mile" rest of that shift.
One hour into next part of the shift, I find out that there's not much more to do. All the trash has been collected - and I have 3 hours left. I kind of just say fuck it to myself, and start walking from one side of the project to the other, carrying the same bags of trash. I'm doing this non-stop for 3 hours.
I continue doing this for the next days, whenever I'm out of actual things to do.
My supervisor calls me, excited, and says I'm now getting glowing reviews from the site managers - and asks me if I want to extend the contract by 4 weeks.
Years later I end up working for a shitty firm that monitored keyboard activity, your desk presence (via movement sensors on the desks!), and a range of other frightening KPIs. Once again, I end up "refactoring" code all day long - essentially just moving blocks of code up and down, doing no real change to the code. I got "exceeds expectations", but soon jump ship.
Hoooooly toxicity. Good to hear you jumped ship, that sort of environment can't make for a good working environment.
Let me guess, your Skype/Teams icon still changes color when you stop moving your mouse for sufficiently long time, and you are technically prevented from changing this setting...
Managers who do not really know how to manage people.
Workers who lack the work ethic to be left unsupervised feeding back into the first item.
At least the construction site manager had a point - morale is contagious, and you being seen sitting around by other, busy workers would not have been good for their morale.
another factor i've seen is fear, people won't change because everyday someone see someone else slacking a bit, so why work if they don't.. causes a race to the bottom. a strange form of distributed equilibrium. and people implicitely know this but nothing changes.
and this accumulates at nation scale where politicians will say that to improve economy or job quality, it would require so much resources, money or else.. when in fact it's mostly a mental switch that has to be flipped.
A few common warning signs:
0. No mission project summary, or a moving finish-line fiasco in progress
1. Top-heavy management budget allocation, or 4 levels of managers involved with zero stakeholder input into the project design/testing stages
2. No hallway testing, or your interface requires explanation and has already failed before coding starts
3. Entrenched incompetence, or people doubling down on failed project inertia rather than recognizing the issues
4. Relative skill value perception in HR, and the trivialization of workmanship standards. Translation: Letting the family Dog drive.
5. Failure to translate business ops needs into actual deliverables
6. Failure to quickly fire people engaged in counterproductive behavior.
7. Perfectionism: nothing is perfect in all use-cases
8. Novelty bias... Translation: Letting the puppy drive.
When I’ve been on greenfield projects that have 10-20 people starting work on it all at the same time, it’s almost guaranteed to fail. Decisions will be slow/tedious, bureaucracy will arrive instantly, individual accountability will be nowhere to be found.
With that being said, projects that get boatloads of ppl on day 1 are rarely bullshit projects, per-se. They tend to be good ideas, an important new project, just executed in a way that’s bound to fail.
But I do think in the early stages of a project, the number of people ultimately making the decisions (and writing code) needs to be small. More people can be added later, if necessary, but in the early stages of a greenfield project the team must be small.
I’ve been on projects where 10-20 people are thrown onto it before a line of code exists and it’s an absolute shit show: decisions take ages, devs are constantly stepping on each other’s toes and extremely unproductive, there’s no personal accountability, etc.
Great comment so, because it confirms that I am not the crazy one!
One thing I still go to Tim's for occasionally is their bagels. Only because they're cheap, and recently I noticed that while there were no other customers in line, the employees—who are largely there as the result of abusing the temporary foreign worker program and don't at all get paid enough for the work they do—were relentlessly cleaning anything they could, such as dusting behind the display signs. Meanwhile, I'm staring at the repulsive mess of Tim Biebs and soggy "Dream Doughnuts", wondering how the hell we got here.
I'm 30, so only discovered it on its way out, but I make a point of stopping in if I find one.
Also, the lawyer would send something and ask us to sign the same day. Try sending a lawyer something and asking them to sign the same day. It’s not gonna happen. After reading the document myself, I found several typos, including a few serious ones (incorrect dollar amounts of coverage), and the lawyer acted like it was no big deal and almost annoyed that she needed to correct the mistakes, almost as if they weren’t even needed to be corrected. But the lawyer would constantly point to something as “you signed it! There’s nothing we can do.”.
It was a strange dichotomy that the lawyers applied to the world. For them, they were loosey goosey towards us, never being clear in emails, but anytime you needed something from them, you got pointed in every direction but towards them. Even getting the final number to pay was an exercise in frustration.
Lawyers cache everything in language that doesn’t pin them down in any way. They are like spiders weaving these webs throughout society.
Silicon Valley really nailed how lawyers behave, especially the interaction between Hooli and Pied Piper’s lawyers. The lawyers create all these binding relationships but absolve themselves of anything and everything.
> The lawyers create all these binding relationships but absolve themselves of anything and everything.
Lawyers are professional social engineers...they achieve their goals through manipulation of the social fabric. There's a reason so many go on to become politicians.
And there's a reason programmers like writing programming tools as well.
And a reason some gamers seem to think game dev is cool.
I suspect that all three are somewhat the same reason.
And burnt out lawyers? Sow mayhem or become philanthropists?
That video is full of clips of a lawyer speaking candidly and clearly, too, and offering to shred a document that was just signed. :)
In the UK the only thing that will get you struck off is embezzlement of client funds. In other words, a crime. You may get struck off for deliberately lying to a client, but that depends somewhat on the niche and on plausible deniability. The likelihood goes up the more prestigious and corporate the clients are.
Otherwise, for basic everyday work like property conveyancing, you can minimum-effort paper push and no one will call you on it.
Doctors, accountants and some parts of software dev can get by with a similar phone-it-in culture. You can be mediocre/bad and in some situations no one will notice and/or care.
Perhaps you're making the same mistake that, e.g. whoever made up the term "Smart Contract" did?
Contracts -- in REALITY -- do not "define" things about the world; that's sort of a programmy way to look at things.
They are, by definition, not the "thing" itself, they are a bunch of words that you can hopefully refer to in the usually unlikely possibility that one of the two parties involved decides, justifiably or not, that something screwy happened and they want to fight about it. "The contract" is in no way necessary for real life things to get done.
Although you are required to do your job, you secretly believe it is pointless and should not need to be performed.
( dupe of https://news.ycombinator.com/item?id=33709599 ? )
i believe in addition to bullshit software projects, there exist projects where the business case is good, but the implementation is very dysfunctional and has poor communication, bad or non-existent coding standards, code architecture and team boundaries that don't make sense, etc. you do talk briefly about this in the quote
> That job also taught me that tech and the business case for a business case are mostly independent. You can have a strong business case and be successful with garbage tech, and the reverse is also true. You should prefer a strong business case because customers are so important.
but i think it is perhaps worthy of it's own article. what to do when your company has thrown so many developers on a project that should have one? what to do when your "design" (term used extremely liberally) has no conceptual integrity? i wish i knew what fred brooks would do.
These are first-derivative bullshit jobs. My job provides an useful service in that it enables other people to do their bullshit jobs. Those jobs do not need to exist, and if they didn't, my job would not need to exist, either.
Just like advertisement tricks you into buying things you don't need, the advertisement-funded internet tricks you into paying attention to things you don't care about. We're better off without.
My perspective is that advertising sets a perverse incentive to create content that is indexable and clickable (machine-translated Stack Overflow answers, recipes with twelve paragraphs of backstory, fake download links for the latest Marvel movie) instead of content that is valuable.
While I would agree that advertising can result in distorted priorities it isn't inherently evil. We used to have an internet with both ads and rational incentives. Even then the value of the internet service could not cover itself without advertising.
The main point though was that the cost of the internet is a mix of direct and indirect cost. I seriously doubt there will be enough people willing into incur the direct cost for modern services at a level that will co er those indirect cost incurred by the network and content providers.
While an ad free internet sounds great it isn't very practical or necessary. The answer is on the user end rather than the provider end though. Stop rewarding the perverse incentives instead of pretending we can wish them away.
Now pick that up, your mother doesn't work here.
That is absolutely a job that shouldn’t have to be done.
The best job I’ve had was making educational tools for children, but it was also the worst paid. There seems to be a linear relation between pointlessness and pay.
Therefore, all meaningful jobs trend towards minimal/no pay.
Meanwhile everything that _makes money_, no matter how helpful to society at large, requires workers and will pay for them.
So the farmers, transporters, teachers, cleaners, garbage haulers, ... are paid badly. If they stop for more than a few weeks, society is doomed. Nurses are also a badly paid good example, doctors are an interesting edge case.
Top politicians, bankers, CEOs, a lot of management, ... has low or even negative value. Their incompetence and egoism is a huge cost to society, not only monetarily, but more so in health and well-being. They might disappear for months or even years before we really start to feel the impact. Hence, they make a fortune.
IT jobs are somewhere in the middle. IT is a godsend for large scale organization and prediction, but that's maybe 10% of our jobs, and taking away the other 90% probably won't hurt too much. Twitter lost 80% of its staff and it seems to take a while before the hurt starts setting in over there. So while IT is not paid as much as politicians et al, they are paid pretty well.
It might fit with your view, too: A society tries to grow moral compasses that make people do important work for free, as one way to build redundance.
Also, side note, ty for the CoRecursive podcast, lots of excellent episodes :) I especially like the more “story time” ones, e.g. ppl talking about the early days of Apple/Unix/SQLite, being a sort of pioneering remote developer, etc. My main piece of constructive criticism would be that the episodes on functional programming don’t translate too well to a podcast - interesting to read about, but too much detail to ingest in podcast format, I’ve started to skip those (even though I am a fan of FP). IMO the podcast would be more consistently great if there were more stories about programming, less describing of programming paradigms/techniques.
It sounds like you are saying you don't like the old episodes, which I totally get.
The functional programming episodes are where I started and it took me a while to figure things out.
I've thought about taking them down before or putting them in another feed but I unless I'm misunderstanding I haven't made in a year or more.
Would it be helpful to give them a different title or put them in a different season perhaps?
I could see two different feeds, but maybe unnecessary - e.g. maybe most listeners tend to listen to the new stuff and ignore the old stuff anyways.
Then at another company there were the bonus projects. Our annual bonuses depended on completing a project that was not part of our normal work - almost always the pet project of a manager in another department. Inevitably it'd be over halfway through the year before we got the first info about it, and by the time we got started, it had already been canceled or that manager had left the company and no one else knew anything about it. I never heard of anyone at that company getting a bonus project that wasn't DOA.
And this didn't happen to me, but one place I worked got a government contract. They had to hire someone officially approved, and a couple of us worked with him early on to discuss what he was going to be working on and how his project would interface with ours. Then, still pretty early on, the government agency paused his project. But it had already been funded and there was the contract, which said he couldn't work on any other projects. So he spent the rest of the year reading, playing games, and asking every now and then whether the project had been greenlit yet (it never was).
And another job, there was one guy I worked with, who despite being a fairly good developer and decent designer as well, routinely got tasked with typing in content that another department printed out and passed to our manager. That was one full day's work per week. We asked if they could give us access to their system so he could just copy and paste it over, but they said no. However, they let slip what their system was and when we looked at it, there was a public RSS feed of the content that they were printing out and handing over for him to retype into our system. It took less than one hour to setup something to process that feed in the background, import it to our system, and free up 20% of his time. That's hundreds of hours per year.
This stuff is very common everywhere.
If 8/10 new companies fail, it should be expected that most new product initiatives fail as well, at least at some phase of the cycle. What's hard to fathom is that for a massive company, the risks can entail quite a bit of spending.
Some things have strategic importance even if few customers are using the service/feature such as having a comparable feature set to competitors (because the customers always ask about that even if they never use the feature, it 'needs' to exist. Sometimes to keep things cooking against an alternative substitute as perennial leverage (like building out natural gas infrastructure in case the Russians decide to cut you off!).
Often things are done because just one, specific 'big' customer asks for it, or there is a regulatory concern. I've worked on projects that existed so there would be no appearance of malfeasance.
Often projects are basically 'know-how R&D' that may or may not pan out. Apple did a couple of iterations with cars and shut them down. Now apparently it's back up ... but could get shut down again.
And finally, though some might argue otherwise, sometimes people are 'doing stuff' because a company will not justify a paycheck for sitting around, and management doesn't want to layoff because it's just socially difficult and / or they want to keep resources. Often we feel that corporations are evil about layoffs and productivity, I feel it's a bit the other way - if we truly wanted to we could cut pretty hard without affecting productivity. Most people, even execs, don't really want to do layoffs, they will avoid it if they can.
But yes, a lot of work is straight rubbish.
The army may make you tired to make you sleep well and stay fit. Also a bs job makes you prepared to think less of bs commands later. The part when you discovered it could be a part of the plan, not an accidental realization.
If the intention was to maintain the team's physical fitness, surely there's a less injury prone exercise that could have been arranged. My guess is that the purpose was to condition you and your team to accept and perform a task without understanding the reason why the task was being performed.
Oh, you replaced the entirely server-side system with the UI of Craigslist with a hot mess of React that takes 5Mb of download to deliver 5kb of content? RDD.
Built an ML-powered recommendation engine when you only have three products and could put them all in a single grid for the customer to view? RDD.
That said, I have seen a company try, mostly unsuccessfully, to switch to a service-oriented architecture when it didn't make any sense to do so.
I also saw a greenfield project that the lead developer used to try out some hot new tech. So it wasn't something that already worked fine, but it was resume-driven development or at least chosen because the tech was new and interesting.
Rebuilding it using HotNewTech is a win/win solution.
The phrase is “If you have time to LEAN, you have time to clean”
At least he could’ve tried to add a rhyme at the end: “if you have time to sit, you have time to clean around it”
At an interview (doubtless for a BS job) I was asked "What is a good day?"
I replied: "I walk out of the building smarter than I walked in."
The job can be as stuck in foopid as it wants; my brain comes with me.
Yet every fresh manager at some point comes up again with the idea.
So.... that means 40-80% of the industry is on "bullshit" projects. It probably doesn't count maintenance programming or other IT work, it's probably new software jobs, so I'd drop that by ... half? 20-50% of the industry is in bullshit jobs?
Consider consulting projects, which are likely worse numbers because those are almost always "new" projects.
But don't forget, there's deck chains like astronaut architect jobs, prototypes/demos, migrations, etc that can fail right and left.
The issue is that the big successes are like winning the lottery for a company. Decent successes are often very visible, they MAKE themselves known. But those are the rare ones, most IT projects that succeed are modest goals, modest outcomes, and probably don't even make any exec's bullet lists.
The other issue is that management can be easily swayed by fast talking IT jargon spewers, and they can earn four or five years of sweet paychecks until the bullshit collapses.
The fundamental issue is that business management still doesn't think it's in the IT business. That IT isn't as (or more!) important than accounting and finance.
Several years later I worked at a company that had a successful product and then the bosses decided to add a chat app as a new product. The bosses thought they could get people to pay for chat.
Everyone else thought it was stupid. People don't pay for chat. I left that company but they did release the chat app after putting maybe 2 or 3 million dollars into it. Of course it was a spectacular failure. I hear that the chat app and it's servers still haven't been shut down, although it is no longer mentioned at all on the company's website.
Even the first example from the blog - the busy work for the time when he needed to wait instead of resting. From the perspective of the manager it is not that clear cut - if he allowed him to rest on the job it would be demotivating for the rest of the crew, who might get jealous. The manager would have to do something with that - maybe he would have to make sure that everybody was getting the same amount of that special job, or make there are also other jobs that are considered special and they could be traded or some other arrangements. In general it would add complexity to the management of the kitchen, maybe not much - just one additional thing to make sure that everything is fair, but there are countless of such cases.
Is this some sort of attempt at karma farming? Some of the comments are pretty identical too.
They take a few real issues, misdiagnose them and then swerve from fallacy to fallacy.
"And who knows more about a job than the person performing it?" It turns out usually many. A person who thinks their job is pointless isn't usually the best judge of a project. They may be right. They may just be ignorant of the bigger picture. Of course people feel this way but it doesn't make it true.
Starting from the first example of being told to work when it wasn't their break they come off more like a child asking why they have to clean their room or or brush their teeth because they lack rhe oerspective to make sense of it. They then conflate that as being equivalent to several completely different situations, again coming off as if they are a child who lacks perspective or understanding...
I would say someone is a bullshit person if they feel the need to express that their job is bullshit yet keep showing up every day.
You can go write software without bullshit for yourself any time you want. Turns out most people don't because they don't have the slightest idea how to run a business yet they are 100% sure the people who are running successful businesses are full of bullshit.
This doesn't feel in good faith. Many people out there enjoy having employment and finding new employment isn't always easy, or guaranteed to provide better circumstances in any way. For example, consider that many still live from salary to salary, or might live in countries where good practices don't have a significant mindshare.
Could people do better? Sure, but that just isn't the objective reality for everyone, therefore there's no need to be that mean towards them.
> Turns out most people don't because they don't have the slightest idea how to run a business yet they are 100% sure the people who are running successful businesses are full of bullshit.
Please remember that a system can generate business value and ensure that certain job positions even exist, and yet be badly made and problematic under the surface all at the same time. Profitable software and good software aren't the same thing from an engineering perspective, although they might be similar from a business perspective.
Alternatively, consider that many companies out there are simply burning lots of investor money on unprofitable systems anyways in hopes of them becoming good from a business perspective, regardless of the underlying engineering quality. Sometimes this means systems that are well engineered (though also overengineered), other times that means systems that still work but are a ticking time bomb.
Personally, I feel some empathy for the people in the article and don't want to dismiss their feelings.
But the whole thing conceptually makes no sense. If you hate your job then just hate your job. Quit if you can. Don't if you can't. But the "bullshitness" of it is irrelevant and the only person with skin in the game on how "bullshit" the job is, the employer who trades a resource to pay for the job to be done, says it isn't bullshit. Your perception of if your job is bullshit is irrelevant unless you are interested in doing something about it but by nature it's not the kind of thing that the person complaining about it is usually in a position to do anything about.
So if you want to make this your stand, then
a) quit
b) Lobby internally to get moved or whatever or unbullshit your job
c) accept that you are willing to do bullshit jobs for money
But just complaining that the jobs are bullshit is meaningless. No one is paying you to philosophize on the abstract value of your job. It counts for nothing. The world doesn't owe you to have the internal perception of value in your job. That's your job to achieve. Not the world's responsibility to provide.
Basically so what if it's bullshit? Let's say your job is literally to show up to an office and write "my job is not bullshit" on a whiteboard on repeat indefinitely. So your job is clearly bullshit. So what? You've accomplished absolutely nothing by pointing this out. Someone is willing to pay you to do it and you are willing to show up for the money. There's really no further interesting discussion to have.
But if you are going to make "my job is bullshit" the thing you are going to a loudly take a stand on, and then you wake up in the morning and show back up on your own free will, then I think that's indicative of a weak character of the person aimlessly complaining, not any other issue with the system.
I will concede that this is a fair point!
> But just complaining that the jobs are bullshit is meaningless. No one is paying you to philosophize on the abstract value of your job. It counts for nothing. The world doesn't owe you to have the internal perception of value in your job. That's your job to achieve. Not the world's responsibility to provide.
I will disagree with this, however, at least in cases where the premise of the project isn't hopelessly wrong in the first place (e.g. pet project with no future, but rather just a flawed project that could be successful with some action).
Identifying issues that make you miserable or make you feel like your job is meaningless can have multiple benefits:
- figure out how things got there, what is the root cause, which ways of thinking and practices are harmful
- figure out what other canaries are there for identifying such circumstances, whether it's attitude or culture
- think up actionable steps to either improve things and make it less meaningless, or identify why those couldn't be feasibly done
See JSON used in a DB column instead of a foreign key, because someone wanted polymorphic links but couldn't be bothered to use either multiple tables or the dreaded OTLT pattern, instead choosing perhaps one of the more problematic solutions instead? Take heed and call it out, if possible. See something like that happening in regards to how work is organized or what tools and practices you have to follow, e.g. having to fix every single SonarQube identified code issue when the quality profile is all wrong and many of those supposed issues are false positives? Take heed and call it out, if possible. See the whole team struggle because of the constant re-engineering that's also mentioned in the article? Take heed and call it out, if possible. If none of that is viable, look for the door at your convenience.
It's good that people frankly talk about their perceived issues in the industry, for example: https://devrant.com/
This is validating and matters for mental health, so they don't just suffer in silence, in more productive communities people might even have actionable solutions, e.g. for how to work with legacy code. That's also why student groups are popular when people are learning to code and decrease dropout rates - because people no longer feel completely alone, same idea.
Ergo, I don't think that complaining is useless, there is some benefit to even articles like this: http://www.stilldrinking.org/programming-sucks
And if your team has enough people who identify issues together, then you might just be able to do something about it together, even if there would otherwise be one or two other people who'd dictate their own rules and preferences that would make people miserable. Or don't, it's your choice. You can just bite the bullet and work in silence.