How to Build a Web 2.0 Dating Site in 66.5 Man-Hours | Hacker News Reader

How to Build a Web 2.0 Dating Site in 66.5 Man-Hours

bakery.cakephp.org

12 pointsvlad17 comments

It doesn't seem as though he was actually going for a "parting the sea" effect. If that was the case, he probably would've spent quite a bit more time than 66.5 hrs on it. My first opinion was that it is actually one of the better designed websites I've seen recently, although everyone seems to be going for the green and orange scheme lately. While I agree with what omouse said about social networking sites not needing to be aesthetically appealing, there will have to be some features that people look for that differentiate them from one another. Simple and beautiful are two that come to mind, although that is a pretty subjective matter.

Good design seems to be working for Facebook. LinkedIn is also reasonably well-designed (if a bit busy and inconsistent). MySpace is the only social networking site with egregiously bad design that is worth speaking of, and it won't last forever.

It's a myspace clone. The nudge thing is the same as facebook's poke and OKCupid's woo. There are no dating-specific features and I really don't see anything novel about copying the Myspace browse option.

It looks better than those websites maybe, but obviously people don't care about the looks of a social networking site.

Other than that, it's awesome that he got a copy-cat site up and running 66.5 hours using PHP.

I think he was focusing on raw implementation time.

You could really spend thousands of hours in design of something with so many connected parts as a dating site.

The big thing about Myspace and Facebook is that they aren't overtly "dating sites." You can use them as dating sites, but don't have to tell anyone if you don't want to.

For most people, Free Online Dating is way too much like personals in the newspaper so they don't bother.

It'll be interesting to see if this works. My prediction is that it will wind up overrun by porn spammers.

"For most people, Free Online Dating is way too much like personals in the newspaper so they don't bother."

Yeah it's a bit...not creepy but not the norm.

Your prediction has a flaw: it assumes that it'll grow large enough to be attractive to porn spammers :P

Also, I still don't see what there is to have "work". It's just basic features that I'd expect any social networking site to have. There's nothing dating-specific and no hints seem to be dropped as to what's going to be added.

Description from the article:

Let this be a testament to Web 2.0 and the effectiveness of rapid development frameworks such as CakePHP: I built a full-featured dating website, from concept to launch, in 66.5 hours. In a typical 9-5 job this would amount to about a week and a half.

He should stick to his 9-5 job...

You're an idiot.

It's not the most original idea, but it's executed very well. Good looking design, responsive site. What's no to like? It's not something I'd use, but it's cool that he built something.

Thanks for the ad hominem.

But I'll concede that I am an idiot. I'm looking at how original this idea is and what features separate it from other websites instead of looking at how awesome the design is and how cool it is because he got it up and running in 66.5 hours!

I wasn't impressed when there were a few Lisp clones on Reddit made and I'm not impressed that there's yet another clone of MySpace made with a nice design.

But you're right, "it's cool that he built something."

I think he got a lots done for 66.5 Hours. I couldn't build something like this in that timeframe

Bootstrapacitor.com (now presumably dead, but of similar complexity to Mingle2) was done in about 66 hours. So was Scrutiny (http://www.amherst.edu/~scrutiny/).

The key phrase in his article was "This is the fifth site I've built using CakePHP so I know my way around." Bootstrapacitor was my second site using web.py, my 4th Python app overall, and reused a lot of code from an earlier Python webapp (which still hasn't launched yet, after 3 months and 200-300 man-hours in development). Scrutiny was my 3rd PHP application - the second (FictionAlley.org) took several hundred man-hours over 3 years to finish. (The first was a minor automate-this-task script and took about 30 hours over 4 days.)

You always become much more productive once you're familiar with your tools. The "writes code as fast as he can type" state that we all aspire too is nothing more than knowing your tools and knowing your problem domain so well that coding is just the act of connecting them. I've found that I can write about 100 lines/hour under these conditions (this is like 20 WPM, well under typing speed). 66 hours at 100 lines/hour is 6600 lines, enough to build a fairly full-featured app.

A sustained 100 lines/hour is insanely fast to me. Is that a common speed? I doub't i'd be able to cook up something on the same caliber of M2 in a mere 66 hours. Moreover, after 9-5 I usually feel completely beat (I fall asleep standing, in the train).

While M2 is arguably simple, in terms of coding, I'm still amazed at how quickly he had everything from brain dump to production. I had a seed idea which required me to think and rethink, read and reread, draw and redraw. And this whole process took a matter of months.

Is this speed within the norm though, considering you have a decent knowledge of the language you use? Does this time include rereading your code and rewriting for optimization?

Sustained 100 lines/hour is fast for me too. Burst speeds of 100 lines/hour over an hour or two are entirely reasonable. That's less than 2 lines/minute - if I'm coding as fast as I type I can usually do 7-8 lines/minute, so it's ample padding for debugging, optimization, design work, etc.

The conditions for burst productivity ("flow", in the psychological literature) are:

1. Well-defined, achievable goals

2. Automatic knowledge of your toolset - you don't have to look things up about your platform.

3. Fresh & well-rested

4. No interruptions

His article explicitly addresses 1, 2, & 4, and I suspect he arranged his time so 3 was satisfied too. He knew what he wanted from his site, and ruthlessly trimmed it so he only needed to implement features directly relating to the core goal. It was his 5th Cake PHP app, so he knew his toolset in and out. The 66.5 hours were not all spent in a row, and he had no interruptions or overhead while working on it.

These conditions also held in my Scrutiny example (done over Thanksgiving break my senior year of college, with some cleanup over winter break) and Bootstrapacitor (done over 2 weeks + 1 week of cleanup, finished about a week ago). They did not hold for FictionAlley (my first PHP webapp) or my startup (my first Python webapp, and we also don't really fully understand the problem yet), and those two projects take significantly longer.

I agree on his point that web 2.0 names will look dated pretty soon.

It is very pretty, I had almost the exact reaction that you wanted from people when I first saw it. What program did you use for your graphics?

I'd like to know what sort of day job he has that manages to squeeze 66.5 hours of project time into a week and a half.

Reply on news.ycombinator.com