Ask HN: How much of your time was spent building something that didn't go live? | Hacker News Reader

Ask HN: How much of your time was spent building something that didn't go live?

17 pointsRaed66713 comments

Over my 45+ years programming I’d guess about 25% of my time got spent on code that never went live. Sometimes the project got scrapped during development and started over with new requirements or tools.

Of the remainder more than half of the code had a short lifespan once it went live. Less than five years, some of it just a few months or a year.

I’ve worked on a handful of projects that have lifespans of a decade or more, but they stand out as the exceptions. The long-lived code stayed in production mainly because of organizational inertia, not because of exceptional code quality.

I keep all that in mind when tempted to polish a likely turd in the name of “code quality” and “best practices.”

That is also my feeling, recently some old colleagues told me that the project we worked for +18months has been decommissioned since the company bought a competitor and absorbed their platform.

That means all the whiteboards, technical debates, stress about modularity and extensibility are all now in a digital trash bin.

In some ways it's better if it doesn't go live. You get to do the fun part but skip supporting it. You can mentally move on and any blame or ongoing issue disappears. Plus you can still reference it on your resume.

Best project I ever worked on in a job never shipped. Learnt a lot and never had to support it!

I work for my state government and I’d say at least 50% of what I’ve worked on over the past 20ish years hasn’t ever made it to production or at least was used sparsely in production. Main reasons for this include: leadership cancelled the project and decided to spend 5-10 years on a RFP and implementation from another company. Leadership had some vanity pet projects they wanted to make them look good and once they left it died on the vine. My manager had some “great idea” and assigned me what amounts to busywork. A common thread on many of these projects is that I ended up having to play a BA role and gather requirements or in some cases make up my own requirements.

The category of “released but unused” is interesting. Each incident has a story to learn from and I’d love to read about them.

My favorite is trying four times to deploy an address book component that people would use (not primary application function, but very important nonetheless). Only on the fourth attempt did we see usage that was meaningful, and it was the simplest version (search based UX, which we all now know is key, but didn’t back then).

At previous jobs, I've worked in features and integrations that were used by very few people for very unimportant things. The reason we built them was always so the sales team could market them, even if they were near useless. It was very unmotivated.

Hell, if it was never used, but the features were interesting to build, I wouldn't have a problem at all. I prefer to work on systems that have code released as frequently as possible in direct response to customer inquiry. You know that code is being used and someone is benefiting from it.

80%-90% in the first 10 years of experience, less than 10% within the recent 5 years. Both personal and professional projects.

Once you go live, it's far easier to release new things. You can spend 2 years working on things that never come out. There, 100% of your time is spent working on things that fail to launch. But after the launch, that 100% is converted to 0% for the whole two years. You work on something else for the week, and then that goes live the next week too.

Also the team has to hold together long enough for launches too. No layoffs, salaries good enough, tech debt low enough. This seems to happen more further up the career ladder.

At work or at home?

At work the answer is probably about 5-10% of my time. Most of the work I've done at companies is either building something for clients (who usually want a website or app at the end of it), or for the company itself (ditto). For better or worse, the sunk cost fallacy is ubiquitous in the corporate world too.

For personal projects on the other hand, it's probably 40-50% of my time. Likely in part because I start and end a lot of projects really quickly, and spend way longer on the remaining ones than I should do.

I spend 30% of my time at work in meetings listening to people talk in circles about features that will never make it into the product. Does that count?

Estimating over my entire ~40yr career, I'd say about 1/3 of that time was on projects that were never released. However, much of that work was reused/repurposed for other projects later. I'd estimate that about 1/5 of my code went entirely unused in the big picture.

I was ruminating in my head about a project I spent almost 3 years on not going live. A little relieved to see that this is pretty common.

At work at current job 60% of my time was in a dead project.

At home probably more like 80% of my time

Reply on news.ycombinator.com