If you asked me what one thing I would like to change most about this site, it would undoubtedly be something so minuscule and tiny that the answer would be met with a Nelson Muntz-style laugh. A lot of the little things that have bugged me about this site over the last few months have been slowly ironed out and, as a result, it's working almost exactly the way I'd like it to. A rather exciting update to how search results are displayed will be released soon and, almost immediately thereafter, the next area that bugs me the most will be re-imagined: archives.
There are a lot of articles on this site. Perhaps too many. This is most clearly seen whenever someone visits the archives page and waits a full two seconds for all 273 kilobytes of HTML to download. It's far too big.
One way to resolve this little issue would be to delete the crap that "doesn't matter" anymore. This would sort of defeat the purpose of having an archive page as well as the core objectives behind the Noteworthy platform. The whole idea of Noteworthy is that people can have all of their content in a single location. Every blog post, Tweet, Facebook entry, Mixi post, IRC comment, photo, video, and meta object can fit inside of Noteworthy for future recollection.
So what's the solution?
I've been thinking about this for several months. Heck, I've been thinking about this longer than Noteworthy has existing. One of the many things that bugged me the most about WordPress was how slow the site would get whenever people pulled up the archives page. It took a solid 15 seconds despite having 8x the processing power and RAM that Noteworthy currently enjoys. However, despite the amount of time I've put into devising solutions, I've only devised a few that might be slightly workable.
The Bottomless Page
One of the options would be the "bottomless page" concept that is seen on many social sites like Twitter. As you get close to the bottom, another few months of archives would be added to the bottom and people can just keep on scrolling. Since Noteworthy is an API, this would be dead simple to implement. Robots would still be able to see every post in the sitemap XML file at the web root, and humans would get to see what they want to see ... hopefully.
I don't like this idea, though, as I don't like bottomless web pages. Of course it would be simple to make a button that people can push to load more archives, but this is still an inelegant solution.
Just like the main page, navigation links can be added to the bottom of the content allowing people to load results for a particular year or, baring that, the next X results regardless of its time period. It's dead simple and could be done with minimal changes to the Noteworthy API and front-end system.
I don't like this idea either. It's too simple. It's so simple that it's stupid.
Ditch the Archives and Use Interactive Search
This is exactly what it sounds like. How many people actually come to the archives pages to look for something? If somebody does happen to land on the page, they Ctrl+F and type in the stuff they want to find. Either that, or they use the search box to find the relevant posts and Tweets1. Why force people to load so much data for no reason when the next level of site search, something that will be damned hard to implement on other publishing platforms, is about to be released?
It's a valid question, but doesn't really solve the problem. I know there are many people who do land on the archives page and spend at least a minute scrolling through the titles. Not many, but enough to warrant putting the time into a decent re-imagining of how archives pages can be organized.
A Sliding Timeline?
The Sliding Timeline is something I've considered time and again, but I'm sure it's been done elsewhere. It would basically be little more than a slide bar with "1979" sitting on the left and "Now" on the right. People could slide their way through dynamically-created montage of existing content. Would it be easy to find stuff? I don't think so. Would it be cool? Maybe for about 15 seconds.
Not worth the development time2.
Your guess is as good as mine. The more data that goes on this site the more complex any solution might be. At the moment the Archives page shows just blog posts, but this isn't permanent. One thing is for certain, though; when I do determine a decent solution to the archives problem, it's going to be hard as heck to make but smooth as silk to use3.