Cloud Computing Doesn't Matter in 2019
OK, I admit. This is a controversial statement. After that admission, I can honestly say that as a user, cloud computing doesn’t matter. It doesn’t make any difference where my apps live. Now, to the fun part…
I recently joined Macrometa for reasons that were fairly clear to me. My professional life has been largely spent waiting. Waiting for responses from customers, waiting for responses from colleagues, but worst of all - waiting for responses from applications and websites.
What really matters to me? Performance. Content availability. Dynamic data delivered in real time. What do all three of these things have in common? From an application perspective, the only way to deliver them is with edge computing.
Edge computing is all about running applications closer to where I am accessing them, and eliminating latency and hops between my Mac and wherever the app happens to call home this week… AWS, Azure, anywhere. If you imagine the internet and by extension cloud computing as a hub and spoke architecture, AWS and similar services are the hub. I’m at the far end of the spoke, with points of presence (PoPs) in the middle. The PoP is the first piece of the app-enabled network that I access.
Until recently, the PoP had very limited value. It connects me to apps living hundreds (if not thousands) of miles away, and also had resident Content Delivery Networks (CDNs) running locally. The CDNs cache the static content and graphics I use from commonly accessed websites, but as soon as I make a weird content request (which I’m known to do) it is forwarded light years away to a sub-basement of a cloud datacenter and served up with the most fantastic latency possible. If all I did was browse Buzzfeed articles, this would be fine. But I’m actually trying to use web applications to get stuff done.
Here is where edge computing comes in. The operative word is computing. Moving application logic, analytics, and content closer to me improves my experience. Moving cached content closer to me doesn’t help once I navigate away from Buzzfeed’s “30 Items for Valentine’s Day” article. Having said that, once I navigate to ‘open source games with Python APIs’ I’m on my own in a wilderness of cold storage and waiting for the page to load.
Macrometa is a company that really understands this problem. They’ve created an easily consumed PaaS for developers that blends a globally replicated database with all of the developer tools necessary to get a globally distributed application ready for distribution in minutes – not months. That application’s compute resources live 10 miles from my house or in my browser, not a thousand miles from me. The result is the best possible (and shamelessly selfish) user experience, while retaining enterprise management and distribution.
It may be too early to say that Macrometa is the “platform of the future”, but I can honestly say that it is definitely not the “platform of the past” that most of my web apps (including CRM) are currently delivering.
If you’re a developer, this is the new “it” platform. If you’re a consumer of apps, cross your fingers and hope that in the near future your apps will be delivered on this model. Cloud is no longer new, or optimal. It’s all about edge computing.
The king is dead, long live the king!