Category: kafka

Introducing Kafka…

Posted by – August 2, 2009

In my last post, I mentioned that I was working on a pet project in Ruby on Rails. While I’m only in the early stages of the project, I’m discussing it with anyone who will listen to help gauge interest and inform the requirements. So, here goes nothing…

The project is a social web application called Kafka. It’ll help web development teams and their customers keep track of frontend changes to their sites. Kind of like an archive.org, built specifically for web geeks.

Upon deploying a frontend change to your site, Kafka will automagically take a snapshot and catalog the URLs that changed.

Seems kinda basic, right? In a way, it is. It’s already quite easy to take a screenshot, and there are plenty of services to manage your photos. Yet, I keep finding myself deep into a project thinking “holy crap, this has come a long way”, yet I have no visual proof of that. As much as I’ve always wanted to, I’ve never paused after each new release to capture the frontend changes.

In these days of release early, release often development, we’re constantly reworking our frontends, adding features, and tweaking the UX. Our sites are evolving faster than ever. Assuming you’re sober, you know how much your site has changed, but do your customers, colleagues, and stakeholders?

We have source control browsers like Warehouse and Github to keep track of the ‘when’ and ‘what’ of backend changes. And we have ticketing systems like Sifter and Lighthouse to document the ‘why’. But these tools are usually too private for your customers and too granular for your stakeholders. Kafka uses simple visuals to tell the story of your site’s evolution to anyone.

You can use it to generate executive reports on the site’s progress, to collect feedback from users as soon as a change is deployed, to show new users how often the site is updated, to start a conversation on your blog, or to fuel your portfolio, etc.

Like the majority of web app ideas, the execution is what will make or break it. It must be dead simple, performing the core task of automatically capturing screenshots with virtually no effort. It needs to display the catalog in a variety of ways, satisfying the needs of each user type. It needs to improve your feedback loop and help manage expectations with your stakeholders.

I’m on it. All of it.

Wish me luck!