Ask HN: Side Projects Gone Deadpooled
As eps pointed out in the previous thread (http://news.ycombinator.com/item?id=1772224) about side projects gone big, there were also an enormous amount of side projects that just went dead. I am sure many of HNers are curious to hear and learn from them.
So feel free to list what the project is, how did you come up with the idea, and of course, when and how did it go deadpooled.
68 comments
[ 3.1 ms ] story [ 140 ms ] threadThe trouble is, side projects go unfinished and become dormant, but they never really die - they haven't been born yet.
Maybe one Ask/Tell/Show allowed per user every week?
I think that this is enough and we don't need any more limits. Moreover some of them are really good and insightful in my opinion, so I don't feel the need of a limit at all.
I have http://erldocs.com - nice ui to the erlang documentation http://arandomurl.com - blog / html5 games http://pastebin.me - pastebin that executes js http://erlangotp.com - information about erlang http://talk.io - chat client
some are virtually silent, some are reasonably popular, I dont worry about promoting them or getting visitors etc, 2 of them are still yet to be done and I am under no pressure to finish, I can just do them in my own time.
I've done a lot of side projects. True to what people have said in here, most of them still exist, but get no attention from me or anyone else. Some I have taken down. I mentioned a few successful ones in the other thread.
http://allyourwords.com, a site that made about $1000 selling word-to-website associations (kind of like million dollar homepage), but then I took it down due to maintenance.
http://ohbigdeal.com, was auto-updating by crawling forums for amazon deals and adding my affiliate code, but I could never get it to take off. Now I haven't touched it in months and the update code appears to have broken.
donotpress.me, donotpressme.com, etc., was a gimmicky attempt at a viral site that went nowhere.
Eigenshare - online document sharing for latex-using scientists. Basically, built in document sharing, version control, etc. Couldn't get anyone to use it.
Timeserieszen - scrape a lot of timeseries data from various government agencies, and make a nice display for it (answers questions: "how have annual working hours evolved since 1960"). Never really followed through on it.
Quite a few more.
i used to let people tweet messages to @going2rain and have them show up on the bottom of http://goingtorain.com/ in a random rotation, but the process of switching it over to oauth was just not worth the trouble so i disabled that functionality.
There's a way around it. IIRC, you just have to sign up for the developer account using the same credentials as your Twitter account, and it allows you to create a permanent key. It's been months since I did it, but the app I built still works.
If something did develop traction I would support it for years and years even if it ended up having very few members (livelog was one of those).
I find it difficult to kill off projects that have a few dedicated users, usually I contact them to find a way in which I can let the project go away without hurting their feelings.
Right now the biggest one that is on my 'deadpooled' list is the free drupal hosting site that I started a few years ago. At some point it had a few thousand domains on it, right now I think we're down to below 50, and every month one or two make the move to another hosting provider. Another three years and I can kill it.
Websites are like children, five minutes of work to launch them and you must support them the rest of your life ;)
(Some of these are offline because they are/were hosted on my personal computer which is currently on a moving truck)
Afk247 - http://afk247.com/ - an away message ranking site where people could vote on their favorite away messages submitted by users. This was back when AIM was my main mode of communication at school. Was going to make money with AdSense. Never really took off, but I did get an article in the school newspaper about it!
MadStatter - http://madstatter.com/ - Baseball stats site. I'm a huge baseball fan and love all those weird stats that go along with it. This site had a couple of (imho) cool interfaces to display interesting stats. Tried using AdSense but didn't make much.
SecureShirts - http://secureshirts.com/ - I wanted to learn how to screenprint my own t-shirts, so I bought all the equipment and learned how. I decided to print my own line of nerdy shirts and sell them online. I sold a few, but never got really big.
TweetShirts - http://tweetshirts.com/ - Twitter themed t-shirts. Made a tiny profit, but had a lot of leftover inventory when I decided to kill it.
Sideways Shirts - http://sidewaysshirts.com/ - In addition to my own nerdy shirts, I did custom screenprinting for events/groups/etc. I did it in my spare time, but became profitable and it was a lot of fun. This covered the cost of printing my inventory for SecureShirts and TweetShirts.
I had to deadpool all of the shirt printing businesses when we decided to move to CA after launching Notifo during YC. We left our townhouse in NC and relocated to a tiny apartment in SF which didn't have room for any of the equipment. I'm sad to let that go, but it was very time consuming and labor intensive. I'm glad for the learning experience though.
Flixpulse - http://flixpulse.com/ - This was the first Twitter Movie Review site that I know of. All of these new studies coming out about doing sentiment analysis in tweets all came along after Flixpulse had been out for a while. I did Bayesian analysis on a manually-trained corpus of movie-review tweets, and then it became pretty well self-learning after that. It was fun but eventually the old little computer that hosted it (900MHz gateway that was forever old) finally ran out of RAM to hold the Bayesian filter data. Oh well. Tried doing affiliate movie ticket sales through Fandango and affiliate movie poster sales, but nothing really came of it.
TheRentMap - http://therentmap.com/ - Apartment listings on a Google Map interface which led to affiliate apartment programs. Not quite dead yet, but it would take significant effort to get it to a point where it really makes any money.
There are a couple others on the verge of the deadpool, but some that I think might make fun side-projects for a while longer.
Mainly I do all these projects as learning exercises. Each one has something new to it that I had never done before, so after a while I have an arsenal of knowledge to throw at bigger, more complicated projects.
I think Notifo will be taking up most of my time for the near-future, and one I am desperately going to keep from the deadpool :)
http://findforme.net
I had a total absolute blast coding it the summer before last (python and django, interfacing with the Amazon affiliate program REST apis). But I never managed to build a user base for it (probably because it's just too complex to use), and eventually moved my focus elsewhere.
Tweet my pasteboard and Speed Tweet - two iPhone apps I wrote. I probably spent 50-100 hours on them, and $200 (to be a registered apple developer for 2 years). One was ad supported, one was 0.99. Total revenue about $20. Both are officially dead now that only oauth sign on is allowed for Twitter.
Tankbuddy.com - my first shot at a ruby on rails app. You could enter the details about your fish tank(s), and record readings (ph, temperature etc). I had big plans. Sadly, after a few weeks I wasn't even using the site myself, and was still paying $15 a month for hosting. Traffic was essentially 0 and the project was killed when I cancelled the hosting account.
I have a new project (http://localbeer.me - helping you find locally breewed beer) that I am very hopeful for. I paid for a year of hosting in advance, so even if I choose to ignore the project for a few months it's not costing me anything. This week was big, I had over 200 visitors from a link I posted to reddit.
As others have said, I have many more side projects, but they're not dead, just forgotten. Unlike others, to date not a single one of my side projects has turned a profit.
Thought it'd be a cool way to build a location based city guide -- but clearly this had to be mobile and we didn't have the right dev team to build the rest...we were manually entering information (we figured the best content would be curated, rather than aggregated) Deadpooled in feb 10..
2004 - http://openpodcast.org - At the dawn of podcasting, I thought a podcast feed anyone could fill with audio content would be cool. The system was email-based: you sent an attached mp3 to submit@opencast.org and it would show up in the feed (there was also a phone number you could call to leave a recording). It was pretty popular as podcasting was getting off the ground. Adam Curry helped promote it.
2008 - http://quotably.com - Got the idea that a threaded view of twitter would be useful. Built it in a weekend, posted about it here on HN Sunday night, Arrington wrote a TC post, Scoble tweeted it, and next thing I knew there were upwards of 250K uniques Monday. I ended up killing it 6 months later because twitter turned off the firehose.
http://twitpic.com/2vedlp http://twitpic.com/2vef3n
You get a bookmarklet to save a listing from any site, keep track of the relevant bits of data (company, square feet, salary, bedrooms, mileage, etc), and move them through a little workflow (interested, applied, contacted, lost, etc). All the fields and workflows are completely customizable.
I built it after a simultaneous job and apartment search, having been incredibly frustrated that craigslist didn't have a "my craigslist" like "my ebay" or "my monster", and the fact that you're never looking at just one site when you're searching. I was always scared that I was applying to the same job or apartment twice.
It's been on the backburner for a while, since my day job has gotten busy, so there's no help available and signups are invite-only right now. But you guys seem like the kind of people who can figure it out pretty easily ;)
Here are some invite codes:
http://pastie.org/1208657
I'll try to bang some out next week...
MIDDLE SCHOOL
HelpersSeekers.com - would have been yahoo answers! like site. Somehow never launched. Spent months developing.
HIGH SCHOOL
CricketFreak.com - a cricket news portal. Somehow never launched. Spent months developing.
trackmail(sp?).com - send email through it and get a notification when it's read(a personal omg moment)
COLLEGE
ChupChaap.com Classifieds site for India.
TextbookDaddy.com - textbook exchange site. launched. shut down. within few days
way too cool domain I need to re-register.com woot for young women. developed for weeks. shut down in couple days after launch
Hollr.org - CL-like site. Never launched. Good call!
ClassHunt.com - let students at my uni get a txt soon as a seat opened up in a class
became a huge hit on campus, over 50% of freshman signed up, had to shut down after uni's course registration moved to peoplesoft which made it impossible to parse latest class data
Good news: I pivoted to another product in same market which is now my start-up
http://websitesthatdontworkwithoutwww.com - 'naming and shaming' websites that have a shoddy sysasmin, and a massive petpeeve of mine, but it hasn't been getting any traction so I plan on taking it offline.
Youtubetopten.com (no longer online) - daily top ten of youtube videos voted for by facebook users. Never got any traction either.
(2001) gbafan.com -- A Gameboy Advance review site. It did pretty well. Had a staff of three writers and good amounts of traffic. I wrote the review submission system in PHP and it was my first exposure to web development.
(2006) saidsecrets.com -- Anonymous polling site. I can't for the life of me think of the name of that polling site that runs on the reddit engine, but it was just like that. I wrote it 100% from scratch using XMLHTTPRequest as a means to learn AJAX/JS before moving onto an AJAX framework.
(2006) fitaculous.com -- a Social networking site with a fitness theme. The idea was people would motivate their friends to keep working out/eating healthy. It tracked calories (or weight watcher points) as well as your workouts, BMI, weigh-ins, etc. At its peak it had about 2500 members. I did it with Rails.
(2008) -- Turrets iPhone game. I only had it in the app store for a 3 days and it made me $200, so I was actually pretty happy about that. I pulled it as my employer at the time did not allow that kind of moonlighting and I feared losing my job.