OnlineOrNot Diaries 17

It's a foggy Sunday morning here in Toulouse, time to close out the year.

In the last two weeks I:

• Learned screencasting
• Cleaned up tech-debt and realized I have a long way to go
• Had thoughts on focus

Learned screencasting

As I mentioned in the last diary, I started a YouTube channel for OnlineOrNot, and I'm still learning. In fact, in the last two weeks I learned enough from the egghead instructor guide that I felt the need to re-record some of the original videos I released.

To start off with, I was having issues with:

• un-normalized audio (one video would be significantly quieter than another)
• weird output size (I think I was recording in 4k)
• my pronunciation was off

Though there are some meta-skills that only apply to YouTube that I refuse to adopt:

I'm treating YouTube like I treat other social media: it's just a place to distribute my content, not the source/single means of hearing from me.

This is why I try get folks to subscribe to my email newsletters instead.

Cleaned up tech-debt and realized I have a long way to go

I continued the theme of finishing my MVPs:

• Since Status Pages now use JavaScript, I was able to finally make the graphs interactive (you can mouse-over the data, and inspect each individual data point)
• The emails that Status Page subscribers get now actually tell subscribers what went wrong (before it would just say there was a new incident), without having to visit the status page
• I took a few shortcuts while building the UI for creating status pages, incidents/components/subscribers, and in certain cases it would take ages to load, so I cleaned up those shortcuts

With Status Pages more-or-less at a point where I'm satisfied with what I've built, I took the time to research how folks buy status page software, and actually got really lucky in finding a public procurement report where a CTO outlined every decision they took to pick a vendor.

It made me realize OnlineOrNot's Status Pages have a really long way to go to support big software teams that already use other software as part of their ops stack.

The OnlineOrNot Status Pages MVP is in-fact, minimal. You can't customize the look-and-feel, there's no API for it, so no terraform support, and almost no integrations. Despite this, I somehow have paying customers that use just status pages.

So in short, I'm feeling inspired for 2024. OnlineOrNot's Status Pages will meet folks where they already are, and integrate with software that customers already use.

Had thoughts on focus

In last year's review post, I wrote about how OnlineOrNot's success comes from showing up, doing the work and releasing it. Particularly having a single-minded focus on OnlineOrNot, rather than a portfolio of several distracting small bets.

The neat "hack" keeping me focused on "just" OnlineOrNot is that it's now 3 separate products with a variety of interesting tech powering it, so I never get bored:

• All parts of the product involve a Next.js web app and a Node.js API server talking to a plain old Postgres DB
• Uptime monitoring involves various microservices in a Highly Available configuration on AWS
• Status Pages involves using the latest frontend tech with Remix
• Heartbeat monitoring involves mixing microservices with my API