Entries from December 2005:

Reddit Sux?

I’ve become a recent fan of Reddit, an online “social bookmarking” service that was a project of the Y Combinator Summer Founders Program during the summer of 2005. Anyone can post a link to Reddit, of a website or article they find interesting. Other Reddit members can read the article, and vote it up or down. Based on a super-secret magical formula, “hot” sites with a lot of recent points end up on the Reddit front page. The hot ranking mode works really well here, ensuring that the front page is usually full of fresh and interesting content.

The Reddit community is not a balanced representation of the whole Internet. It has an overabundance of hackers, specifically programmers, and more specifically people working with Lisp and other scripting languages like Ruby and Python. There are other social bookmarking sites with more well-rounded content, including Digg and del.icio.us. But Reddit’s niche makes it interesting to me.

Reddit recently added a comment feature, where users can write their own commentary on links posted to the site. This is a very useful feature, but its implementation is problematic. For example, here’s a link to the commentary about an article titled “Best Buy Sux.” There are three ways to view comments: the default view of hot (combination of most recent and popular displayed first); new (most recent shown first); and top (most popular shown first). This is unlike every other message board on the web, which all display messages chronologically. If you want to read through an entire thread, you can do so from top to bottom. But with this display, there’s no way to do that; you’d have to switch to new and then read from bottom-to-top.

More importantly, users appear to be moderating commentary based on whether they agree with it or not. As a result, important and articulate messages end up at the bottom if too many users disagree with them. For example, on the Best Buy Sux commentary: a user named jedberg posted a comment defending Best Buy’s actions in this situation. (Disclaimer: I met jedberg in person earlier this year, and he seemed pretty cool.) While it may not be the most articulate comment on record, he makes a decent argument, and it’s the precursor to a whole bunch of discussion later in the thread. But because people disagreed with him, or disagreed with the way he stated his opinion, his comment ends up at the very bottom. I couldn’t understand half the thread until I scrolled down, found his comment, and then scrolled back up.

I have two suggestions which I think will make the comments easier to understand. I recommend that Reddit add a traditional mode to view comments, where the oldest comments are listed first. I also think Reddit should change the default so that comments are viewed traditionally unless the user overrides it. If a lot of users do like the hot mode, the Reddit software could allow each user to set their own default view, like a user preference or through a cookie. But the hot mode doesn’t work well for commentary. New users shouldn’t be deliberately confused just because Reddit has a cool ranking algorithm.

MySQL Persistent Connections with WordPress

I received an inquiry from another WordPress blogger regarding a WordPress support inquiry I had opened in August. He wrote:

I was searching for answers to a very pesky “Cannot establish connection to database!” WordPress error, and read your thread about database persistent connections. I was wondering if you could give me a hand optimizing the database for WP, I am not a programmer by any extent of the imagination, and know nothing of servers, just got one by chance. (Edited slightly by Ryan.)

Continue reading →

The Art of the Start

I’ve been reading Guy Kawasaki’s book The Art of the Start for the last two weeks, and I’m quite impressed. Kawasaki’s book is billed as a tutorial for “anyone starting anything” — but really, it’s a guide for entrepreneurs starting businesses, specifically geared toward high-tech businesses looking for venture capital funding. Kawasaki, previously a product evangelist for Apple Computer, is now the CEO of Garage Technology Ventures. As a result, Kawasaki knows all about what’s needed to bring a product to market and obtaining financing (and customers). A lot of the tips in the book are phrased as “if only the companies pitching to me knew how to do it better…”, so it feels like you’re really getting inside information that most startups are unaware of.

For example, I read the chapter on partnerships today. (Joint ventures, not legal entities.) It was all about how to evaluate and negotiate a partnership opportunity so that you don’t get stuck with losers. These were ideas I never had even considered. If I was running a startup and Sun wanted to partner with me, I would have jumped at it because they’re a big company, no matter the strategic relevance. Kawasaki teaches you how to evaluate these kinds of opportunities to ensure they’re worth the time (and money). Plus it’s a fun read, which helps.

Just like Startup School, The Art of the Start is a very candid and a great resource from an accomplished insider. I highly recommend this to anyone looking to succeed in a technology startup.