travisglines

We just had a comments model (in mongoose) that had an array of subcomments (MongoDB arrays are awesome). Therefore the "trees" (not really trees as we implemented it) could only have one level of indentation ... but that's all we wanted anyways.

I think a hybrid system would actually best. Otherwise posts that don't get many votes wouldn't update their score.

The best would probably actively update on votes and do a catchall operation on ones that haven't been voted on or all on a certain periodic cycle.

Store a "score" to rank by in the database and then use some sort of time dependent function. AKA when you search the mongo for posts sort by "score."

Our score was calculated by something like this:

    var gravity = 1.5;
    var score = num_votes / Math.pow(t+2,gravity);

Gravity being how quickly a post would fall. Notice the t+2 ... it added 2 hours to all posts in order to eliminate things that were recently posted with like one vote sitting at the top.

Then we'd just run this function through the whole database periodically (every 5 minutes or something).

I actually did this for exipe.com in node.js/mongodb let me look it up.

I think you may be over generalizing here. Yes there are vast similarities between MongoDB and other NoSQL databases but many of the issues he pointed out here are specific to MongoDB (and certain versions of it at that).

You bet, send me a tweet (in my profile) ... also this is very relevant:

http://www.meetup.com/Los-Angeles-Hacker-News/

I'm moving there in about a week to work on a startup with a friend. I've been on the lookout for people to get to know in the area.

This talk on an intro to WebGL and how to optomize it is awesome: http://www.youtube.com/watch?v=rfQ8rKGTVlg

Also I found this talk on developing an HTML5 game interesting: http://www.youtube.com/watch?v=yEocRtn_j9s&feature=relmf...

Mostly because its eerily similar to a simple game I made years ago in flash:

http://www.travisglines.com/resume/planetdefender/PlanetDefe...

As someone who has spent the last six months coding in Node.js, I completely disagree with this notion.

One of my favorite advantages of Node is in fact that it is based on javascript. When building my application I can share code between client/sever side (string formatting methods for example) and don't need to do any mental context switching when it comes to languages. I've developed in javascript for years and all that I have learned doesn't go out the window here, all I need to do is learn how to use a few libraries.

The "faults" of javascript are primarily inexperience with event looped architectures that use callbacks (people complaining about having to use callbacks and the spaghetti code messes they make), people remembering the past horrible browser implementations of it (where something would work/be fast in one browser and not another ... Internet Explorer I'm looking at you) and confusion over how prototypical languages work vs. standard OOP.

Callbacks aren't that bad and a few extra nesting levels shouldn't create chaos in your code. Since its based entirely off of one javascript implementation (V8) the cross browser stuff is gone. How it handles OOP is different, just watch these videos first and you'll be fine: http://www.yuiblog.com/blog/2007/01/24/video-crockford-tjpl/

As far as speed goes it gets huge benefits by being javascript. Every time the Google coding machine feels like making V8 a little faster, Node gets faster too.

Yes its different, no its not worse ... its better but takes some getting used to.

Today feels like Google (attempt of) squashing YCombinator startups day.

First they're doing Youtube Live -> Justin.tv etc.

Then better flight search -> Hipmunk

It can be really hard to get your bearings sometimes if things are all the same shape/size, especially if there isn't many people in the app. It needed a central location or gathering point so I figured why not throw a giant obelisk thingy in there?

Its fun watching what people say about it. They want to know why its there, why its black and if its real.

Should now fall back to canvas rendering when webgl isn't supported.

I should make people spawn in random places instead off all at the same spot to change that (what you're seeing is like 100 overlapped blocks).

We actually did exactly what you're asking here a few weeks ago and applied to YC this round with it.

http://www.exipe.com

We built it in node.js and MongoDB. While node and mongo are young projects we have found them very stable and extremely fast. (we noticed a lot of problems reddit had with scaling and we were interested in the stack)

Our decision to build Exipe in node/mongo has been a bit controversial but we've found no problems thus far with it ... ~2 weeks out with slowly growing traffic. Check out a few of my latest blog posts here about the experience:

http://www.travisglines.com

Thanks for the comparison, I'm actually very impressed by the accuracy of Google trends.

I wonder given the lack of accuracy between quantcast/alexa/compete if there is room in the market for another traffic intelligence site that piggybacks off of Google trends somehow.

[Edit] Looks like its against their TOS: http://groups.google.com/group/google-labs-trends/browse_thr...

Hmm...

I agree totally, just was unaware of a better source of traffic data for the two sites. (not every site publicly shares its data as willingly as pg does)

I'll probably be using Google trends in the future, thanks for the heads up.

After looking at the chart (assuming as Swannie does, that Google trends is a significantly accurate source), it seems pretty apparent that HN doesn't have enough traffic to cannibalize significant traffic from them.

I wonder how much traffic HN is cannibalizing from more traditional tech news sources like venturebeat and techcrunch ... http://siteanalytics.compete.com/techcrunch.com+venturebeat....

The chicken and egg problem in localized services is extreme. Like others have said here, if someone is to replace craigslist they will need a very general destination type product.

The problem is is that not only do you need to reach a critical mass before the content is there, but you need to do that in each and every location. This is why it's such a challenge.

I think he might have April fooled you ...

I can certainly see the value of building up a following and leveraging that towards your startup. (plenty guilty of that myself)

It seems like all the extra time you would spend by going above and beyond to get a following/skills could be spent (with the same risk) doing a really cool startup on the side.

While I agree that leveraging intangible assets can be extremely powerful in creating success for web startups, very few people have such an asset to leverage. Those that do are probably looking to enjoy life a little bit more and may turn into angel investors or are looking to get out of the bipolar startup lifestyle.

How does one build an asset like this? By creating a web startup from scratch.

So yes if you happen to be one of the lucky few that have a nice asset like that, do the obvious thing you would have done anyways and leverage it. Otherwise join the rest of the crowd building one of the assets from the ground up with solid execution and a decent idea.

If (as I'm sure you do) you have the times along with the votes it would be cool to see if there was actually a bias when comparing the before response and after response datasets.

It seems like publicizing some preliminary results of the study before the study is complete would put the legitimacy of the final results very much so at risk.

constructive criticism ... for science

I like the idea a lot. Things like the photo, the importing from linkedin, the simplicity are nice. One thing that would be cool is to add collapsible sections where someone lists an experience (like welding) and if you open the section it tells much more about this experience (talking at length about the welding job for example). That way people could go over it quickly and if interested read more.

I'd make the link to the example page a little more obvious, I didn't intuitively think I could click on the image for whatever reason.

I think you can keep hitting the up arrow to find earlier matches too

When first stumbling my way around the shell, after learning about tab completion I discovered the magic wizardry that is command reverse/history lookup (cntrl-r start typing command ... see last command you entered that matches)

I wouldn't assume that your app goes viral by default. In fact the chances of this happening are very slim. The guys that made breakup notifier almost went off the new page when they went viral.

What I would do is put it out there and be ready to take the steps you need to, to keep it alive if it does go viral. Planning too far ahead for this is probably a waste of your time as you most likely wont be able to see the weak points before hand. Just move quickly if it does.

I wonder what the Y Combinator competition will look like in 3 years ...

Call me Captain Obvious, but I feel like quite often the time spent in obfuscating code in many cases might be a major component of why programmers that do this sort of thing get fired in the first place.

What goes on in someone's mind that says: "Maybe if I spend some time making my source code hard to understand I can better preserve my job." ?

In reality all it does is slow them down, waste their time and lead to the firing of the developer. (At which point we get to hear funny stories from people like jacquesmatteij)

If that time was spent improving/refactoring their code and working on new things, they'd actually have the job security they were so desperately seeking.

nek4life, where are you from St. Jay? I'm from Littleton myself and was pretty dumbfounded when I found the Caledonian Record as the #2 result on the front page of HN.

I actually just signed my grandmother up to their service here and all it is, is a paid subscription to pdf copies of their newspaper as they release it. Theres not exactly much to protect there and worst case scenario is that someone gets some free copies of the paper. I don't think its such a big deal for them to be a little lax with password security given what you get behind the security wall. (as long as people aren't using that same password everywhere ... which may be the real problem here)