Posted February 8th, 2020 by Zack Spellman
In which we discuss the future of this blog
Altmeta.org has been around as a blog since the end of 2013, a bit over seven years ago. The posts have, in general, been about the infrastruture I set up to run the site. In a way, it has always been a guide should I ever need to replicate my setup; a record of decisions made and their justifications. It has also been an attempt to build a tiny fraction of what I work on at my day job without all the tools and support a large company provides. To that end, I've learned a lot even if a lot hasn't been accomplished.
As this is the twelveth post, and six of those (including this one) have been since the move to Gatsby, I feel like it might be time to talk about something more than infrastructure. One could argue that talking about how the example Todo app was built was already a step in this direction, and it is in this direction I intend to head for a while.
A few weeks ago I knew I was heading towards this pivot and wrote a list of posts I wanted to write. The first, three, of course, are already done:
After these three, I knew I would be here, in a good position to pivot, so:
It's always dangerous to talk about upcoming plans, since no plan survives contact with the enemy (or reality), but I'm willing to preemptively accept that my prognostication will be wrong somehow and that we can revisit this later and share a good laugh. What I'm going to try to avoid is stopping altogether after this post - that would be the real travesty.
Which brings me to the 20x20 preview which makes up the other half of this post (per the plan). In honor of 2020, I formulated my goals for the year as a list of 20 goals, each of which is 20-themed somehow. Relevant to not stopping, one of my 20 goals was to add 20 posts to this blog before the end of the year. So in seeking to make that happen, we'll need 16 more after this one. More on 20x20 to come shortly though.
I wanted to have one last infrastructure post (haha, this "last" won't survive to the end of this post) since I don't feel like I've done the tweaking of the base style selected that I hoped to do. So this will be a time boxed overhaul of the layout component and the front page, plus the standard description of what I tried, selected, and learned.
Coming back to the 20x20 goals, I'm currently tracking these in a private Trello board which I have bent to this purpose. However, it occurred to me that it could be a reasonable step up in complexity from a Todo app with which to practice and demonstrate some web application patterns. So by building an app to visualize and record goal progress, I can also provide a bit of a backlog of posts to this blog (assuming I can write the app expediently).
I'm not actually sure now if this will be one post or two. If it is two, I'll spend the first one on the viewer (hopefully with a demo), and then the second one on the model that lies behind the view. I'm still working out what features I want the viewer to have (big question: do we support viewing the past or not), and from that driving how the model needs to be structured. If I go for something more complicated, then I'm likely going to need multiple posts to cover the additonal ground.
This is an infrastructure post, kinda. In order to support a data editor for the 20x20 app, we'll need to support storing the data somewhere and allowing it to be accessed later. I want to only allow myself to add data, but want much of it to be publicly visible, so figuring out how to enforce that securely is also part of the equation here. But so is mapping my data model to DynamoDB, or whatever other data store I choose to use (It'll be DynamoDB).
Finally, I'll create a post talking about the second "viewer" app which allows modificaiton of the data. This one is a bit tricky because I'll need to create a separate backend which simulates whatever mutation framework I'm using for the real deal, but I'm going to optimistically assume that's not a big deal and go ahead and assume this will be a relatively easy post. That just leaves the editor app itself to be the hard part, which again depends on how complicated a data model I choose to support.
After doign the 20x20 app to death, I want to take on a new app - a photo and video gallery. Long story short - I think the paid photo gallery options are overpriced, the free ones aren't private enough, and none of them are self-hosted. So I'd like to take a stab at creating my own, complete with React Native client so that I can take pictures on my phone and have them automatically upload to my cloud-based photo gallery.
I'm hoping my experience with the 20x20 app will give me some guidance on how to do this - authZ, data modeling and access patterns, building an effective viewer are all included in both projects. But we'll see how many other things I need to support - blog storage for photos and videos, automatic thumbnail / various resolution creation, how to keep costs down and avoid running any servers I need to maintain, etc.
And speaking of maintaining servers, the big bugbear I'm afraid of for running a gallery is indexing. When you have infinity money, running a cloud-based search server is fast, super effective, and easy. However, it isn't cheap - you're basically guaranteeing you'll need to run two or three servers 24/7 in order to support your use case. And, if you're talking about a private gallery which will only be searched occasionally, paying for that to be ready all the time is a waste of money and computing resources. So I'll (likely) need to figure out ways of incrementally computing and updating indexes without runnign any servers, which may prove a bit challenging. But I like challenges, so it should be fun too.
So, that lays out the next 6 or 7 posts, as well as two separate applications I'd like to write this year. I'll still need to come up with another 10 to make my goal of twenty, but I suspect the gallery will be an endless font of opportunity as I tweak and add features. For many reasons I'm looking forward to getting started, which is what I'm going to go do now.
Last edited: February 9th, 2020 06:39:30
Previous: How I Learned to Stop Worrying and Love Typescript