Purpose / Point of Glitch: A theory

Over the years (oh btw it was my anniversary on the forum when I starrted writing this), I’ve wondered exactly what sort of product glitch was meant to be. Clearly it doesn’t seem to be a sort of “free hosting” platform as we’ve seen through the rise and fall of the ping services along with discord bots. Rather it tries to differentiate itself from other platforms. Some platforms try to be multipurpose and check all the boxes like https://repl.it and https://gitpod.io .
But glitch seems to be different in the way that it is focused more on the web. I’m reminded of this by the recent blog post “Wordle Mania and the Remixable Web”
So glitch is a web app hosting platform, but why does it give us limited amounts of disk and memory? Let’s think about it from a different perspective.

When building an open-source web application you have two options:

Build a fully functioning application for the masses

This requires lots of server power, requires you to figure out how authentication for the masses would work. In addition if users could create content you would have to monitor for abuse.

Build the essential part and let the users deal with their own data and hosting

This is like having an actual server for a discord guild instead of discord’s giant databases handling everything. You just focus on building your app. The app doesn’t need to scale!

I believe glitch is targeted at people who build the second sort of app. Want to have your own forum? Just remix the project you’re interested in! You don’t need to build some sort of megaservice that handles requests for thousands of forum each day. You can just comfortably focus on your app writing. As a tweet I saw the other day put it:

My thoughts do align with some later developements on the product. For example, we were motivated to make things more static back 2 years ago.

Users like @wh0 later discovered that upon closing an editor session, static sites would no longer use the lws web server. Instead, they would be served from an AWS S3 compatible server (perhaps actual Amazon S3?), thus leading to some ridiculously good autoscaling.

Now another development is approaching that extends beyond static sites: You may have seen the new Fastly Compute@Edge banner, but here it is again.

We’re building an all new integration with Fastly that will let you build more sophisticated Glitch sites that scale infinitely, with 100x faster code execution startup time than other serverless solutions.

If this gets pushed out, this will mean writing non-static apps that scales becomes tremendously easier. Load balancing would be done automatically with the extra benefit of low latency as a result of code being run on the edge. A single request handler connected to a cluster of databases could easily scale as needed without any human intervention!

So, in conclusion glitch seems like the sort of platform where people try to focus on building the app they want with the ideas they dream of without the worries of scaling and maintenance!

Cited Sources: