Hacker News Reader

This is one of the things I did years ago that, with hindsight, could have been so much more.

I started up a website on the urging of some people from one of my Usenet hangouts nearly 10 years ago. It was specifically about demystifying university entrance interviews (and exams where applicable) at Oxbridge, something nobody else did at the time. It was as successful as I could hope, at the time; plenty of people visited and contributed their experiences, we got written up in the national press, we even had takedown requests from certain colleges.

My failure/regret is dropping it instead of growing it. It was partly because the tech running the site was terribly outdated (initially it ran in ASP on an Access database. Really) and although I'd converted it to PHP it was still the ugliest code I've ever written. (In my defence it was my first 'big' webapp type project.) As a result of this crazy spaghetti monster of code and associated bad practices, at one point I lost the live database and just... never got around to restoring an old backup.

Looking back, the site could have grown massively. There was a clear demand and we were ticking nice boxes such as university access for all, state school equality, etc. We could have expanded to other universities, personal statements, traditionally difficult subjects, internship advice, graduate careers interviews, etc. We could have ridden the social wave. We could have charged. Etc.

I'm pleased with the first year or two of the site but the fact I just kind of dropped it and got on with other things still annoys me. I was incredibly strapped for time but looking back, I could have done it. Really, I hadn't figured out what I wanted from the whole thing. I just thought it was cool that I'd made a website and got people I'd never met to contribute to it. Once I'd sort of achieved the 'have a website with some information on it' goal, I lost interest, and never really thought about how it could really become something big.

Someone else thought so; they bought the name (not the domain, funnily enough) when I let it expire, and copied the format - and the content, including my lame humour - word for word. They're not growing it either, but they haven't been graceful to me (basically yelling at me for letting the site die, thus irritating my own annoyance at the same) so putting my hat in with theirs to take the site places isn't really an option for me any more.

My main failure(s) have always been, and continue to be, not sticking with ANY idea long enough to see it through. What I consider good ideas (for myself) are few and far between, so when I have one I tend to over analyze it to the point that I add so much to the idea that it becomes too much to take on or I eventually convince myself it really wasn't that great of an idea and I move on. I also hesitate to get real feedback on my ideas from my peers, because of some silly fear that one time it will really be goldmine and someone will execute it faster and better, which is ironic considering that I eventually write the idea off anyway. Would be interested to hear if anyone else goes through something similar.

A more tangible failure, which was also a mild success, was a tracking/analytics site I did for Xanga, a not so well known blogging/social community site. They allowed you to inject JavaScript into your themes, which made it ripe for a wide array of scripting features/widgets/etc. I was one of many tracking sites that would show you who was visiting your blog, name, IP, etc. Xanga eventually took my domain name, "XangaSpy" through legal pressure. The change in name had a huge impact on my traffic and while struggling to rebuild, my host corrupted my database and its backup. That was all she wrote. Xanga eventually added similar features themselves.

At peak I had 15,000+ registered users and was making ~$200 a month from Adsense. The tech wasn't terribly advanced and I certainly wasn't making a living, but it was still a mild success for myself as it was long before I really understood web development and well before I took my first entry level job as a developer.

Lessons learned? Building an app that piggy-backs on someone else's service could find your work obsolete should the service implement similar functionality. Also, backup, backup, backup.

In decreasing order of painfulness (which is a good proxy for emotional investment):

Diffle/GameClay. That one's public; you can read about it on the web.

The MUD I started when I was 14. This was the first major programming project I worked on; I taught myself C and C++ and Java and UNIX networking for it. But eventually I found it was too big a problem for my programming experience, and gave it up.

RuBeans. Right after I'd finished a Netbeans plugin for my job, I figured that Ruby was becoming hot (this was early 2006) and really could use a decent IDE. I worked on it for a couple months (in my spare time) then gave it up when I found that another group was a lot further along than I was. I think they ended up getting bought by Sun, and Ruby support folded into the default Netbeans installation.

inAsphere.com, a dot-com for teenagers by teenagers. Back when I graduated high school in 2000, a bunch of friends and I were employed by this venture-backed startup to create a "teen content" site run exclusively by teenagers. We had our funding pulled after 3 months.

There are also a bunch that I don't really consider worthy of a blurb - getting turned down from jobs (ITA, Twitter, and FriendFeed are the notable ones), jobs I passed up that ended up being successful (turning down the employee #2 position at DropBox is the big one), programs I wrote at various jobs that never got used, etc. And a few that could easily have been failures, but that I turned into successes at the last moment - basically everything related to graduating from high school and college falls into that category.

"The MUD I started when I was 14 ... I taught myself C and C++ and Java and UNIX networking for it ... eventually ... gave it up."

sounds more like a good learning experience than a failure.

That applies to nearly all of my failures. :-)

And most likely not just yours. In fact, I'd say that success is merely the re-application of lessons learned during previous failures. And that success without preceding failures is very rare and should be right next to 'winning the lottery'.

Nah, it is just him.

Many years ago I had the choice between staying with a successful company I helped start where I had little control over the overall technical direction and joing a very risky venture which gave me enormous freedom including the opportunity to immediately hire several friends from college, all of whom were very smart but without much work experience.

I choose the latter which turned out to be a big mistake. Not only did I walk away from a small fortune but nothing went right at the new company. Even though I did my best to secure everyone good offers, as a whole the team was unable to work together. Everything fell apart when a key partner we were counting on to raise funding instead used our ideas to help secure a management position at a competitor.

In the next company I start, I will be much more careful in who I hire and who I choose as a partner.

Probably not what most people here are looking for, but my epic failures as a rationalist are recorded in this LW sequence: http://wiki.lesswrong.com/wiki/Yudkowsky%27s_coming_of_age

In (what should have been) my second-last semester of school I had a very solid job lined up for post-graduation. Then I D'ed a course by 3 points on a 150-point exam, which meant I had to tack an extra year on to my schooling, and lost the job. What really kills me is that there was an assignment I didn't hand in for that class that would have pushed me over the edge - it was done and all, it was just cold that day and I couldn't be assed to bike to school to submit it.

Having said that, the job no longer exists (the company got bought out and everybody fired), I got a much cooler job that summer by chance, and now I'm in a whole new industry doing stuff I love - so it ended pretty well. But I still consider that course one of my great failures.

In 1998 my brother and I sat down and planned a web service which would allow travelers and writers to write journals and upload their photos whilst on the road. It would focus on simple image uploads, short chronological posts and then notifying your friends/family of new content through a subscription service. I wrote a functional prototype which included the ability to post/upload via email. We bought a domain and had plans to launch the service at some point. Looking back on this now, we had really started to develop a part of a rudimentary hosted blog service (chronologically arranged posts which facilitated discussion via comments) with elements of Flickr (image uploads and comments and social discovery of images uploaded near your current location) and the ease of use of Posterous (email all your content directly into the publishing system).

Beyond a working prototype we never pursued this in any serious way. Of course the execution is what matters and I think it's a failure that we never executed on this service to see if we could have created something that users adopted.

Wow, it soulds like you invented blogs almost 5 years before Evan Willams started Blogger.

Blogs were around for a long time before they had a name. I started mine in 1998. We just called them "personal websites".

As to Travel Blogging, I actually implemented the grandparent's idea back in 2005, right when the Google Maps API became available. We just passed the 10,000 blog mark last month. Check it out:

http://www.blogabond.com/

It's never made a dime of direct revenue, but the consulting work it generated has kept me traveling these last several years.

When you say blogs were around for a long time, I know there must have been people publishing journal type sites, but what I think was key with blogger was that anybody, not just webmasters, could post frequently and effortlessly.

"Blogs were around for a long time before they had a name."

Exactly. We simply called this concept 'Travel Journals'.

The main obstacle we faced in 1998 was that most people didn't have digital cameras yet and the scanning and uploading of photos was a serious timesuck.

I like your site, thanks for sharing.

Over-developing a product that consumers didn't want.

I learned it would have been much better to rush out an early prototype than to spend 7 months developing apolished product that no one would use.

So many people hear things similar to this, nod in agreement and then proceed to make the same mistake! I don't get it. Somehow they think what they are doing is not the same thing. I guess this is the reason most good advice have no positive impact.

It's hard to evaluate your own work objectively. There's a bunch of psychology around sunk costs, etc. that don't affect you when you're looking at someone else's project.

A friend of mine is doing this right now. The project involves a whole pile of electronics, a card system and a bunch of other things, he's already working out how to make millions of the things but he doesn't have a customer yet.

The scary thing is he is hugely successful at most of the things he does. It's going to be interesting to see how it comes out.

I sold myself short — I moved to a place I thought was a second rate. And what do you know? It was. Truth is my belief limited me. I was bound to confirm it.

I ignored intuition — I should've connected with people. Focus on what I want and doing it instead of learning more. Maybe that's harder for the self-taught to see.

Spending a decade in enterprise software.

Please explain. Were you not advancing your skills as much, or were you just plain unhappy?

"Please explain. Were you not advancing your skills as much, or were you just plain unhappy?"

Writing a coherent answer would take longer than I have today,but essentially I was bored out of my mind after writing the n-th insurance/banking/leasing system.

Some initial thoughts can be seen in a blog entry I wrote long ago.

http://ravimohan.blogspot.com/2006/07/but-martin-enterprise-...

As I said, this was written a while ago and I would make a more refined argument today, but that will have to wait till I have some free time.

Deciding to join the laboratory of a really shitty advisor. I managed to survive, graduate, and even thrive.

But just a little bit less laziness at the beginning during my advisor search would have helped a lot.

Moving just a little too slowly is my main problem. I tend to think in abstractions and try to implement those rather than implementing a particular solution and then abstracting it.

My personal biggest failure is not doing the wrong thing or doing it wrong, but simply getting the timing wrong.

That seems to be the story of my life so far.

A few years ago, a friend of mine and I were DJing a lot. Just when CDs were coming in and people were starting to accept that DJs might not use only vinyl forever, we considered starting a website to sell electronic music digitally. At the time the bandwidth costs were so high that we couldn't have made a profit, so we shelved the idea. We totally missed the boat on that one.

In my current company, which I once had a formative role in (I'm one of 3 founders), has made far too many mistakes to list here. The biggest ones IMO have been lack of quality control in hiring due to a desire to grow too big too quickly, and not realising how important it is to decide on the company culture that you want and defend it to the death. Now I feel like I work in a bank.

I made one of my sites nonprofit. Unfortunately this did two things: gave me less incentive to promote it, and caused problems when it came to creative strategies to market it.

EDIT: Also caused problems when it came to sustaining it fiscally.

I attempted to start a mobile software company(Symbian, well before the iPhone). The biggest mistake: attempting to start the thing without adequate funding. We attempted to bootstrap, and despite the fact that we built a great application we simply didn't have the money to drive traffic and build the exposure we needed. We had quite a bit of critical success, but that just didn't translate to sales.

I'm a strong believer in proper capitilization of your start-up (not a popular opinion on this board).

Here is mine: http://news.ycombinator.com/item?id=737192

Not if you don't let it fail it isn't! I still don't see any pricing plans listed on your site. Get on it!!

In 1999, I bought 10 domains of several large companies and trademarks, who at the time did not have any non .com domains. Think something like toyota.co.uk, but different country.

I sat on them for 2 years, then let them expire. Right now they are all owned by the rightful companies, but I don't know how that came about.

You wouldn't have been able to keep them and would probably/likely had to spend tons of money defending them. Only exception is nissan.com (not a car site!), but it is the exception that proves the rule.

In a way you dodged a major bullet.

Now, if you had bought some nice generic domain names (ie cars.co.uk) and let them expire that would be a different story...

My biggest failure: Overworking myself and avoiding put all my energy and focus in only one thing at once.

Here's a true story involving theft, pregnancy, heart-attacks and ruined friendships. Definitely my worst failure...

http://www.maximise.dk/blog/2009/06/worst-case-scenario.html

Not joining a fully sponsored two month summer studentship MRC Lab, Cambridge,, UK because I wanted to startup something during summers back home India. It is the only lab with most number of Nobel Laureates in the world (12 and the 13th one this year from the same lab won the prize for his Ribosomes research)

Looking back it seems I should have joined it because it would have been a tremendous experience (and more so because the start I was doing in the summers didn't work). But, hey, at that time starting up was so rosy that everything else looked secondary. Though it was my first real brush with entrepreneurship, I still feel I should have gone to that lab for 2 months.

I should have pushed Apache Rivet more back when Tcl was still a reasonably popular language.

Didn't do backups when I was about 16yrs old. One dark and stormy night, I was experimenting on Stacker or Doublespace or some such, and ended up losing all my files.

All my files in this case included every single song (tracker/module) I made since I was 12, which I think was about 50 pieces. Gone. I cannot remember if they were masterpieces or not, but I sure do miss them.

Talk to your kids about backups!

Giving up too early. Thrice.

Next time around I am going to ride it out for as long as possible.

Something to read in the general context of failures: 'Technology and Courage' [pdf] http://research.sun.com/techrep/Perspectives/smli_ps-1.pdf

-Sophomore year of college, I took the spring quarter off. What did I do with it? I spent much time and money outsourcing the creation of a social network based on iterative drawing. I was truly the "idea" entrepreneur. It failed, mostly because at the time I had no understanding of the concept of product/market fit. I also gave up on the idea prematurely, and lost steam.

-Freshman year of college, I outsourced the creation of oldefriends.com, a site to help you reconnect with your old e-friends (internet buddies) .. I shoulda written it myself and shoulda launched it on HN

-Senior year of college, I outsourced the creation of a multiplayer Flash clone of worms. This time, no concept of minimum viable product, and I ended up with a product nobody wanted to play.

All in all, I spent ~5k learning two valuable lessons.

1. Build something people want.

2. Figure out that they want it early on, before you invest a lot of time. If you can build something that you want for yourself, that's quick validation of #1 and #2.

Fortunately, I made all of that back and way more when I launched & sold a successful Facebook app in between sophomore/junior year.

My worst professional failure was adding Hibernate to my project at work. I was charmed by their prospects, everyone reviewing it positively and quality of their docs.

It turned out that I've spent more time making Hibernate do whatever I need without failing than doing everyting else. It followed the principle of most surprise: I could never guess what'd it go given a code snippet, provided all the complex machinery involved under the hood.

Now I fear ORMs. Wrote my own for that project, it's primitive but does its stuff and is completely predictable.

I use LLBLGen, I had to spend almost no time learning it, and nothing ever goes wrong, basically. My only complaint would be the syntax is a bit verbose, but I don't really know if there's any way around that in a statically typed language. But its nice to be able to put 0 thought or work into persistence code and have everything just magically work.

Figments of my imagination. Every single 'obscure object of affection' that became a thorn in the flesh was something that did not/would not be subjected to 'reality test' early in the game.

At this late point in the game my advice to the young ones is "Test the hypothesis early and often".

Ask HN: What are your failures?