Developing Retro Roundup

September 11th, 2017 12:11 PM
by
Filed under Musings;
1 comment.

Last week, I officially launched Retro Roundup, a curated RSS aggregator of retrocomputing news.
Retro Roundup banner
Or rather, re-launched: Retro Roundup was founded in 2005 by Kay Savetz, who approached me this past February about taking over the site. While keeping the same purpose and logo, I rebuilt the site in WordPress, adding taxonomies, email subscriptions, and more. After several months in development, it was finally ready to demo during the KansasFest 2017 lightning talks.

Those talks were two months ago, yet the press release announcing Retro Roundup was published just last week. What took so long?

The problem was that Retro Roundup didn’t have a defined end state. Unlike Juiced.GS, which has concrete deadlines resulting in a finished quarterly product, Retro Roundup will never stop growing. The more RSS feeds I add to it, the more content it will publish. How many feeds and how much content are enough to launch a website?

I was reminded of the development of Duke Nukem Forever, a video game that took 15 years to publish. The developers didn’t have a roadmap for what the game would look like when it was done; as a result, they kept adding new levels and features and scrapping old ones to be current with the latest technology, which was advancing apace with the game. But every product is outdated by the time it launches — at some point, you just have to declare that it’s met its goal and release it.

In my case, I thought I was done Retro Roundup in April — until I showed it to my librarian friend, Michele DeFilippo. She suggested I add "facet searches", which was not a term I’d ever heard, though I was familiar with the functionality: almost every e-commerce website offers parameters and filters to narrow search results. Adding this feature to Retro Roundup made the site infinitely more useful and usable.

Then I thought I was done — until a month later, when I attended WordCamp Portland, and met Scott Tirrell, a fan of the Retro Computing Roundtable podcast, on which I’m an occasional guest. I showed him the site, and he enthusiastically offered many more suggestions — from adding a search box to including YouTube channels among the site’s feeds.

With such great feedback, I could’ve kept working on Retro Roundup indefinitely. What pushed me to finally release the site was Kevin plugging it on episode #45 of the ANTIC podcast (38:51–41:53). Listeners of that podcast immediately flocked to Retro Roundup and began submitting RSS feeds. Even before I knew how they’d discovered the site, I realized that I couldn’t keep this cat in the bag any longer. So I spent a day off from work adding dozens more feeds to the site, many of which I’d solicited months ago on Facebook, before deciding I’d met some arbitrary, minimum quantity of content.

Despite this milestone, there are still more feeds and features to add. Mark Lemmert of 6502 Workshop was the first to use the "submit an advertisement" form, which I’d somehow overlooked in my testing. I was appalled by the results; an hour of my Friday night was spent bringing it up to spec. And a developer who contributed essential functionality to the a2.click tool is even now working on code that will make Retro Roundup even more usable.

Before last week, I had only two retrocomputing websites: Apple II Bits and Juiced.GS. I hope the former entertains its readers, but it’s primarily a personal outlet; while the latter is in support of an offline product. Retro Roundup is the first retrocomputing website I’ve built that I would call a resource for the community. I’ve learned scads about WordPress and project management during its development. I hope it is found equally rewarding for its users, who will discover new sources for retrocomputing content, and for publishers, who will see new visitors being sent to their site from Retro Roundup.