The common wisdom in the Rails community at this time is that scaling Rails is a matter of cost: just throw more CPUs at it. The problem is that more instances of Rails (running as part of a Mongrel cluster, in our case) means more requests to your database. At this point in time there's no facility in Rails to talk to more than one database at a time. The solutions to this are caching the hell out of everything and setting up multiple read-only slave databases, neither of which are quick fixes to implement. So it's not just cost, it's time, and time is that much more precious when people can['t] reach your site.

Full interview here

• Advertisements - The big blocky crazy distracting ones in the middle of lists? On comments and story pages? Yeah, those are gone. Replaced with clean, crisp, google adsense. Hopefully this will be a nice compromise. Andy is our designer, and is the man, by the way. He's been tackling most of the ad issues.
  
• Frame Navigator - We weren't lying when we said it was something we planned to make optional. All the outrage from the technical crowd quickly made it a priority though. Shout out to Tom for getting it in over the weekend!
• Votes - Lots of little issues with votes, so allow me to get geeky for a second. We have servers. Lots and lots of servers. They're in multiple cities. Users are using different servers, talking to different databases, and pulling from different banks of cache (for now at least, each region has it's own distributed cache shared between servers). Each page load you make, things get cached and stored in memory -- this technique drastically reduces our database calls. Here's where it gets tricky. In some circumstances, we're caching the entire page, in others, only portions of the page. How do you keep the votes outside of what's cached, cached seperately so your not making a ton of unnecessary database calls, and distributed in a way where they won't fluctuate between servers? Now toss in ajax calls every 15 seconds that refresh the vote totals (sit on the page, and you'll see the votes change). I think we have this issue nixed, but we've only rolled it on in one specific place on the page -- the main list of stories. We can use this to monitor what's going on and test. If it appears to be working correctly now, we'll roll it out to the rest of the site. The other places should be accurate to roughly a minute or so. I can assure you though, all votes are being counted. It's just a display problem that we're working on.
• Formula changes - Part of the beta process is to figure out the right speed for the front page. We're not digg, we're a portal. We operate on similar principles, but we have a different user base and different traffic requirements. First, the formula was too slow -- the AOL copies Digg story stayed on the front page for days, which, if you're a fellow fan of irony, is simply rad. On the other hand, it's still "old news" in terms of the life of the story. Then we went to fast, and everything on the front page only had a few votes. Part of this is low traffic, part of it is the speed of the page changes. We're getting close though, lots of little tweaks -- Trey Long, just like Bow Wow, is a man amongst men. Even if he does prefer to be called "Baby Trey."
  
• Lots of little bugs: We had a crazy little bug on the comments page that would remove focus inappropriately as you were typing, we had a js error on vote handling that would show up from time to time, we had some advertisement bleed through that andy nixed, Craig introduced elements of formatting into Comments which I'm sure people will be thankful for, and a bunch of other stuff that I can't immediately think of..
  

• User suggestions - Most of the "cool" visible stuff will probably be based on user feedback -- we get a ton of it, and we read it all. When you log in, in the top right corner, there should be a "feedback" button. Use it and let us know what your thoughts are :) I'd still like to make the "pop up new window" thing an option, but depends on how loud people are about it.
  
• Better, more scalable multi-city approach - Craig's got some cool stuff prototyped in dev that should help us distribute this thing all over the US. What this means is faster response times and the ability to handle the onslaught of traffic thats heading our way when we switch to the full domain. The difficulty isn't so much in the premise of it, but in coordinating all the different folks that need to be involved.
  
• Better vote display - As noted in the above.. I'm hoping to roll out my crazy ass vote stuff all over the place.
• Other possibilities - If we can get the fires under control, we might be able to get Anchor Chat finished and online. The prototype in dev is, quite simply, friggin' amazing. Mad props to Christoph, Andy, Tom and Trey on it. They've done well.

• Thomas Vander Wal is now participating on tagsonomy.com. It's tagtastic.
  
• Steve Pavlina's 10 stupid mistakes made by the newly self employed.
• Rich Ziade touches on the need for Web2.0 to reach the general population
• Cache wrapper for XMLHttpRequest calls
• Reminders on how to be a cheap ass and save every last penny
  
• MCache Documentation. All the cool kids are using it.