hayley 365


thoughts on the upcoming Rails Rumble 2013

16 Oct 2013

2013-10-16 23:10:10 -0500

Today's lovely topic is the Rails Rumble coming up this weekend.

After experiencing heartache and/or general disappointment in previous Rails Rumbles, my expectations are quite lower this year: I'm basically in it for the free hosting.

See, I know from experience that I'll probably get a month or two of a free VPS out of the Rumble. And I've got a potentially expensive-to-run project that I'm building where I'd like to give myself a bit of breathing room to make it profitable before I'm actually paying out of my own pocket (the hope is to never pay out of pocket at all).

See, if there weren't previous code, I'd just make that project my Rumble project because the potential initial traffic boost from the Rumble might be an awesome way to figure out exactly what sort of VPS this projects needs. But I digress…

my actual hopes for the Rails Rumble

What I'd love to do is figure out a project that meets the following guidelines:

buildable in a weekend

This one goes without saying right? I should have something complete-feeling because this thing only lasts 48 hours.

MVP in a day, but has obvious layers

The goal here is that, if things go poorly, then I'll at least have my MVP. If they go awesome, I won't have run out of ideas of things to add. The "layers" part is because I don't want to end up with my MVP, but then I keep going and end up with something half-assed because I wasn't able to get the project to its next logical step.

something I've wanted to build anyway

I've got a list of things I'd love to build. Frankly, a lot of them are out of my league technically (or feel that way anyway), but still a weekend of do-or-die might be perfect for those things that I think are just out of reach.

is capable of being profitable

Frankly, the trick here would be for it to not be too profitable because, otherwise, someone's just going to steal my idea and do it better.

I'm not that great of a programmer. Nor am I that great of a designer. It's just that, over the years, I've become more and more of a doer. Basically, that fact that I'm actually going after stuff and getting it done is my only advantage.

But here's a competition where I presume that most of the people viewing the entries are tech people. Tech people quite capable of outdoing my efforts.

And here is also a competition where your viewing audience specifically knows you pulled this off in 48 hours. As an audience member, this is incredibly empowering. 48 hours. Even I could do this.

This of course increases the odds that someone will go and do your project better than you.

appeals to the rumble community

Looking back at previous winners, I see a clear theme. Either they're projects with general appeal or they're projects designed to help tech people get their work done better/faster/smarter.

My entry last year went nowhere. Frankly, this is mostly to blame on the fact that it wasn't all that impressive given that I picked something way out of my league technically and wasn't able to finish all that much of it.

But second to that, it was something that wouldn't have appealed to many in the rumble viewing audience.

I know that I personally would only vote for a project that I either found hella useful or exciting.

is something that could really benefit from a boost of initial traffic

I've never figured out how to properly promote a site to get it off the ground.

Well, actually, since I have such a fear of negative feedback and tend not to directly engage in online discussions period, I seldom ever promote a site of mine at all.

To date, my most successful site traffic-wise, only got where it did because I happened to hit upon an SEO trick that I can't apply to any of my other sites.

So it would be nice to come up with a project idea where maybe the rails rumble traffic would be enough to get it off the ground. Instead of how most of my sites just hang out in obscurity for ages.

actual ideas

new earthquake site

Confession: I already have like 4 or 5 of them (lost count??). Unfortunately, the one that was actually starting to get traction ended up being kicked down by Google at some point (never did find out the cause) and lost like 98% of its traffic overnight.

And unfortunately, it was also an EMD (exact match domain) so when it stopped ranking for the title of the site (the EMD), it basically became unfindable.

But ever since then, I've wanted to do a brandable name earthquake site.

But funnily enough, the idea that I might work on for the rumble is something kind of niche and is, again, an EMD. So basically, it's something that could end up suffering the same exact fate where it stops ranking for its own name and then is virtually unfindable.

reddit movie finder

Did you know that Reddit has a whole bunch of random subreddits (okay, I know of like 3) that are dedicated to posting links to full movies online?

The concept behind those things is that they're full movies on sites like youtube where they're going to get yanked for copyright infringement pretty quickly, but meanwhile, you'll be able to watch them.

I feel like this is the idea with the most appeal.

Frankly though, the copyright thing worries me. I would just be posting links and/or embedding movies rather than hosting anything myself, but it's just murky enough to make me paranoid.

reddit mturk scraper

Though I have found a more passive way to make my beer money, I'd still love to design a better interface to reddit's hits worth turking for subreddit.

For one, just being able to easily filter out the dead HITs would be a huge improvement.

make it karaoke

This is something I'd love to build but what frankly worries me about this is the associated costs.

The idea is simple: * user uploads mp3 * server process it and makes it karaoke * server gives download link to user

There's potential for a whole lot of bandwidth and CPU usage here. And without any clear profit path, it would likely be a project-in-the-red every month. I mean, running ads probably ain't gonna pay the bills and I tend to doubt that this could ever really be turned into a paid-for service.

One thought for the future would be to see if client-side javascript would be capable of doing the necessary mangling such that know audio files ever touch the server. The user's system handles all of it.

perfect pitch training site

Personally, I think it's possible to learn perfect pitch. Specifically, I think it's possible to memorize the sounds of everything.

Here's why: I fell in love with Tori Amos a little late, so by the time I became a fan, there was a bucket-ton of a back catalog to listen to.

How this relates to perfect pitch is that, it was infrequent, but there were cases where I'd play a piano note out of context, and I would go "that's how that Tori Amos song starts". And I was right in 100% of all cases.

It's far from proof that memorizing pitches will actually work, but I'd like to try it.

Here's what I'd do: I'd experiment with things like starting either small, like maybe 2 notes, or maybe starting only with Cs, where you'd only have to memorize octave gaps. And then I'd drill myself on that limited set of notes, slowly expanding to more and more.

But less than that, I've been underwhelmed with the online options for perfect pitch training. I think I can build something better. Of course, then I have to figure out how to get the word out.

the future

Sadly, none of these ideas excite me. I mean, that's the problem in some ways because I'm deliberately going after things I think I can actually pull off in a weekend.

Add to that the fact that right this very second, I'm more compelled to learn how to build something with Meteor.

So as things stand now, in some ways, I'm more dissapointed than when I began this because now I realize that none of these ideas feel very compelling.

And at this very moment, I also can't even think of the ideas that scared me technologically. Or at least, nothing that would rely on a Ruby backend. All of my technologically scary ideas are based on front-end JavaScript.

Frankly, I think I'm gonna have to take a look back at all of my random "someday maybe" project lists and find something on it to get really excited about. Or find something in one of the above ideas that makes it compelling to me.

Until then…