Live data from Hacker News

Ask HN: Failed project you spent 15 hours/week for 5 years on?

news.ycombinator.com

21–30 of 79 posts

Re: Ask HN: Failed project you spent 15 hours/week for 5 years on?

#21
Sure, I've had four projects that I spent years on and never went anywhere (at least in the way that was planned).

But that was certainly not wasted time and effort. Not only did it improve my skillset, but large amounts of the code written could be repurposed for other projects, giving a head start on them.

Re: Ask HN: Failed project you spent 15 hours/week for 5 years on?

#22
It is how I work usually for the past 30+ years: I spread my time between 3-4 projects always and most are my own. I work on them for 3-6 years per project until something hooks some success or does nothing at all. Most fail, I had a few successful ones that made enough to live a nice life with a family and a few turned into serious companies. I suspect I will keep doing this until the end of my life: it is a lot of fun, failed or not. But success is obviously a better feeling.

Re: Ask HN: Failed project you spent 15 hours/week for 5 years on?

#24
I took a year off and for fun, I spent 40 hours a week for 9 months on a website that used OCR and other packages for some games. It generated around $15/month in ad revenue, just enough to pay for my EC2 instance at the time. I kept it around but at some point the ad revenue I made dropped like a rock by 90% so I shut it down. It was fun nonetheless and I learned a lot, but it was not successful by any measure.

Re: Ask HN: Failed project you spent 15 hours/week for 5 years on?

#25
post #6

Multiple times. The perspective I've adopted is to look at what you learned in the process of each failure and how that improved your ability to do a better job on the next venture. The truth is that many don't have the stomach for "sticking with it," doing whatever is necessary to keep going in the down phases (I highly recommend the poem "If" by Rudyard Kipling and see if it resonates). You only truly "fail" when y…

I love the Kipling poem. A (lesser) companion for it is Wintle[1]

[1] https://www.goodreads.com/quotes/1033193-if-you-think-you-ar...

Re: Ask HN: Failed project you spent 15 hours/week for 5 years on?

#26

I've spent the last decade doing this. Building up my savings, draining them on an idea. I put in 40+ per week for a couple years on my last startup in the real estate space. Poached some top level VP's from the biggest firm in our country, but we ran through our capital too quickly due to too many pivots and licensing costs, I could only do no salary for so long, tough when your rent is $2.5k/mo. Now I am back at my…

> I was of the "first to market / build a unicorn" mindset

I'm not dissing that mindset at all, but it is essentially gambling. Over the decades, I've warmed to "the pioneers get all the arrows" mindset: that from a business perspective, it's better to be the second or third in a new area, because you can learn from the pioneer's mistakes. It's rare that the entity that invented a thing is the entity that makes the most profit from it.

Re: Ask HN: Failed project you spent 15 hours/week for 5 years on?

#27
Context: I'm the other co-founder of Typesense, mentioned in the article you linked.

Kishore and I have worked together on about 12 different side projects over the course of 13 years, and we've tried to adopt this mindset of consistency, persistence and long timeframes for each.

A few of these projects got good traction, but most of them didn't do well (at least revenue-wise). But here's the thing: working on all these projects consistently over the years, has also helped us learn about things like how to pick a market, how to validate our hypotheses, how to choose technologies when building products, how to maintain codebases over a decade, how to stay nimble, etc.

I would say that the sum total of our collective learnings through all these projects, have helped us significantly in our Typesense journey.

So I would say, showing up everyday is not a magic bullet to making a project successful. Instead, it's a magic bullet to continuous learning and building up a wealth of experience, that might just come in handy when you're working on your next project, which then increases your odds of success.

Re: Ask HN: Failed project you spent 15 hours/week for 5 years on?

#28

I worked nearly full-time for about 2 years using my personal savings on a project management software for labs ("Labengine"). The big hook was that it let you run common bioinformatics apps on your data and it would automatically convert the input file to the expected format (there were lots of competing bioinformatics file formats at the time). I had a few of my old professors use it and "love" it but nobody ended…

You could have made the best software in the world, but as someone who now works as an in-house developer for a medical research university, there's almost no way for individual researchers to have significant influence on IT budgets. Sure, there are always a few rock stars who pull in huge grants and thus have some control over institutional spend, but in general the operations side of academia has little influence on what software is bought/used. Because of this, I've found there's quite a bit of (understandable) apathy among researchers with regards to anything related to IT - you have no control over it, it's a miserable necessity for your real job.

Re: Ask HN: Failed project you spent 15 hours/week for 5 years on?

#29

I worked nearly full-time for about 2 years using my personal savings on a project management software for labs ("Labengine"). The big hook was that it let you run common bioinformatics apps on your data and it would automatically convert the input file to the expected format (there were lots of competing bioinformatics file formats at the time). I had a few of my old professors use it and "love" it but nobody ended…

I work in the medical research space and we have LIMS (lab info management systems for people outside this niche) and am consistently amazed by the number of companies trying to get into our IT systems with lab management packages.

It's a very tough space to break into it and my recent experience is that the big EMS/EMR vendors are increasingly trying to offer "modules" to do LIMS and bioinformatics analysis pipelines.

Every now and again I start thinking about unsupervised learning projects in this space but look around and see just how little spend on analytics software there is for bioinformatics tools and settle down to focus on helping folks with our cluster and general reporting needs . . .

Re: Ask HN: Failed project you spent 15 hours/week for 5 years on?

#30
post #6

Multiple times. The perspective I've adopted is to look at what you learned in the process of each failure and how that improved your ability to do a better job on the next venture. The truth is that many don't have the stomach for "sticking with it," doing whatever is necessary to keep going in the down phases (I highly recommend the poem "If" by Rudyard Kipling and see if it resonates). You only truly "fail" when y…

> You only truly "fail" when you quit.

Or run out of time?

You don't have infinite time to keep trying things.

Post reply on HN