Ask HN: What does your failure resume look like?

9 points | by escapist16 34 days ago


  • BJBBB 34 days ago

    1998 - automated troubleshooting program for use by people on the production line for power conversion hardware. Worked, but system abandoned by employer after I quit.

    2001 - 'real-time' monitoring and analysis of data coming from the ATE stacks on the factory floor with various trending graphs and other stuff for consumption by quality and manufacturing engineers. It worked well, but politics in the Mexico factory sunk it to create more jobs to do this stuff manually.

    2006 - automation of XRF process. The technician blew the power head, so most of this testing was eventually outsourced to a local lab.

    2007 - automation of "Type" tests and high-speed data acquisition for single fault conditions and thermals and other stuff. UL was too stupid to understand anything, and would not accept data unless the tests were done manually and data was hand recorded.

    2019 - 'universal' environmental recording and control systems for use in test chambers and agricultural. Reliable and accurate, but too expensive and required the customer to understand a bit of physics to implement the zone mixing feature.

    • escapist16 34 days ago

      That is impressive. Automated troubleshooting program in 1998 was cool. Do production lines today have that sort of ability?

      • BJBBB 34 days ago

        For that type of stuff, probably a stupid idea because the root cause was poor process control by the factory. Typically modern production lines mitigate troubleshooting and re-work by building it correctly, which was supposedly the original reason for all of the ISO 9k hoopla. And re-working stuff on the factory floor does bad things to long-term product reliability, so my solution to the problem at hand was doubly stupid.

    • DarrenDev 33 days ago

      These are the failures. There’s a common theme.

      2005 — Desktop app for creative writers. Made about $90k in total over its 10 year history

      2010 — Secret sharing website. Made $0 revenue.

      2011 — Hugely popular forum for poultry breeders in Ireland that ran for a few years. Made about $6k in advertising and sold it in 2015 for a few €k, can’t remember how much now.

      2012 — Desktop journaling app. Made about $15k over its 5 year lifespan

      2012 — Desktop password app. $3k over its short history.

      2014 — Editing app for creative writers. Made about $20k over the 4 years.

      2015 — Microsoft Word add-in for editing documents. Made about $50k over the past 5 years.

      2017 — Another Desktop app for creative writers. See 2005 above. Made about $5k in two years before pulling the commercial version.

      Takeaway: If you always do what you’ve always done, you’ll always get what you’ve always gotten. Building B2C apps for consumers who don’t have money or don’t like spending money is a model for continuous failure. It took me a long time to realise this.

      Current project is a B2B SaaS app for dev teams. Work in progress.

      • escapist16 33 days ago

        Wow. That is impressive.How as the Desktop app for creative writers that you created in 2017 different from the one you created in 2005?

        And true, building B2C apps is always a gamble unless you have a product that contributes to user's forward movement.(What i mean by forward movement is that there is no going back to life without that product). Almost all of the B2C products that have made it large satisfy that criteria.

        All the best for your B2B app. By the way, your username seemed familiar - just looked it up on Indie Hackers and Bam! there you were. Followed you :)

        • DarrenDev 32 days ago

          >How as the Desktop app for creative writers that you created in 2017 different from the one you created in 2005?

          That's actually a very good question. The follow on question might be "Why did I think it would succeed in 2017 when it didn't succeed in 2005?"

      • maps7 34 days ago

        Is your 2013 Twitter/Rotten Tomatoes still available to look at? Sounds interesting

        • escapist16 34 days ago

          Alas, it isnt. I got the upcoming, in theatres and opening soon movies information from Rotten Tomatoes using their API and then used the movie name to fetch all the tweets that had the movie name in them.