Current Tech Challenges and Plan - August 2026

Overall

For those of you coming straight to this from a social media post - I have been planning, for some time, to run an online game based on flooding taking place in East Anglia in 2048. This would be played over weeks or months, potentially representing months or years in the game. The players would represent many of the affected parties, from national authorities to displaced citizens. The game would be played on an online platform, enabling players to fit the game into their lives, and meaning that can give it as much or as little time as they see fit. I would like to get it running in some fashion to investigate some concepts and ideas I have around serious games.¹

I’ve been working with a couple of people on my planning, on breaking down my ambitious tasks to make them seem less daunting, and to make progress easier to track. So let’s try that here, with this project.

You can judge how poorly this project is doing by the last update being four months ago. But life has been a little too interesting over those months, which explains some of the delay - but also I’m being so ambitious with this that it can take a good few hours to get back into figuring out where I am, and what I need to do next - and by that time I should probably be getting on with something else.

So this log entry should tell you what I’m doing, but also will make it easier for “future me” to get back up to speed and progress the project.

And this is all part of “working in public”, and also to remind myself how far I’ve come while the actual game isn’t running yet. Also it means that anyone following the project can forewarn me if I’ve made any especially bad choices, or if they know better options that I should use.

Overall technology to be used

A brief summary of the intended platform:

• Everything self-hosted on a VPS.
• Overall authentication and authorisation, so players can confirm who they are, and be restricted to the game information they should see: Authelia
• A mailing list manager to push updates to players: Mailman.
• A wiki for game administrators and players to manage the game information: MediaWiki.
• A forum for players to discuss their orders for the next turn, and to submit those orders: Discourse.
• A “non-diegetic” website to give updates about the game: built using Hugo, hosted on Microsoft Azure.
• A map that can be updated by game administrators and players: to be determined, suggestions welcome.

I was planning to put a summary of what I need to do, and in what order, but I think this should be the next post, as that’s going to be disappointingly complex. Advice welcome on the absolutely simplest way of listing and displaying that, what’s the next level up from a bulleted list but that isn’t veering towards being an actual GANTT chart?

Feedback is welcome, preferably on whatever online forum or social media platform brought you to this page in the first place.

Self-hosted

I have some experience with Linux, so this is favouring what I know rather than maybe what is the ideal situation if I could just throw money at a desired architecture.

Also while this does put more pressure on me, it avoids issues caused by SaaS services disappearing or changing configuration, which can break established integrations.

Overall Authentication and Authorisation

Authelia looks to like good open-source software I can use to manage all the accounts this game needs, but in one place. It should integrate with all the other software I’m considering. That means that I just need to add a player to a team role in Authelia and they’ll have that role in all the other software I’ll be using. This minimises the time I spend on administration, and the possibility of players having the wrong permissions.

On the one hand trying to use a “corporate-scale” IAM² solution is far too ambitious for me, a single technical person, to put together - especially leaning on LLMs for advice, that are only mostly right most of the time. But on the other hand, if all this software can be configured to work together it gives me a reference implementation for more versions of this game, more online games in general, and solves many problems in one go. It also reduces the administration required for any specific implementation.

Of course this could all just be run on a Discord server, but at any significant scale I think a Discord server will “break”. Not that it will stop working, but it’ll be clear that it’s no longer up to the task of keeping information limited to the relevant players while requiring minimal administrative overhead. Also I’ve had issues with Discord servers before, and the company’s choices towards restricting user activity or requiring proof of identity will only get more restrictive.

Mailing list server for updates

I think the game needs to “push” notifications to the players via email, which will then prompt them to come to the forum to discuss the game or to plan their next turn. As I’m not using a service that people will be on anyway, such as Discord or WhatsApp, I can’t expect players to come to the forum every few days, or see game alerts among their other notifications. Therefore the best way to manage that is a mailing list server. This should be a self-hosted Mailman instance, as then it will integrate with the ambitious “single sign on” concept from the previous section.

I was hoping that an online service could be used here, as players or observers or other interested parties being able to add and remove themselves from lists is key. But all of the solutions I’ve looked at seem to be aimed at newsletter distribution, which is more open and simplistic than the requirements for a game.

As of August I’m still fighting that Mailman configuration, just to get it to work, and to integrate it with everything else.

If you’ve read this far and want to be informed when the game is ready to play, please email wwe1-signup at conflict.games . Emails to that address won’t be read, but directed to a folder I’ll work through once I’ve got Mailman working in some fashion.

I can see it only being used for emails from game administration to the players, the deadline for the next turn, things like that. But also it might end up being where players communicate with each other, and where significant announcements are made within a team. Regardless of where the majority of the game happens, such as Discourse or a similar online forum, I think email is the best way to “push” notifications to players and keep them engaged.

Wiki for a static description of the game’s current situation

A wiki that players can use and update seems essential for this kind of game, ensuring that everyone has a “common operating picture” is crucial.

The game runner will set up the scenario on this, and modify a lot of the information as the game takes place - but also a lot of the access can be delegated to players. That means that this needs to be easy for anyone of any technical level to use, and also to support roles and delegation.

Also this will be a good place for teams to share private information, or to have pages they can share with limited groups of other players. So using popular software is probably a good idea, and whatever has a simple interface so that players can get on with playing, rather than figuring out a platform. I’ve started on this using MediaWiki, I’ll share links once it’s got a little more material on it and configuration around it so anyone on the Internet has limited access.

Ideally this would also form the forum detailed in the next section, but I think that there’s two distinct functions here: the Wiki describes the game world as it is, the Forum is where players meet to discuss plans with allies and enemies. Counterpoints are welcome here.

An online forum, where the game, and discussion around the game, happens

Optimistically this should support both real-time and asynchronous communication ( like an online web forum, or a WhatsApp discussion group ). Discord is the most obvious choice, but there are many questions around that platform, who and what is allowed on it, and how it might develop in future.

Therefore this would be something like Discourse, which seems to meet all the requirements. Especially as limiting what players can see which information is key, and is one of the benefits of this being an online rather than in-person game. I need to remind myself of what real-time communication that supports.

( just as an aside, I was recently reminded of the concept of Medium Specificity, which feels like something game design should lean into… as much as I understand it. So I interpret that as meaning that online games should emphasise the benefits of being run online, such as asynchronous play, anonymity, covert activity, multi-threaded conversations - rather than trying to replicate an in-person experience. )

Of course this is all very ambitious, rather than just throwing ideas about the game onto a Discord server and getting started. But I’m very conscious of how setting up the right platform choice now could affect all sorts of workloads and options down the line.

Development website and Azure

The game’s development website, that describes what the game is and how it’s built, is the one you’re reading now. It’s a static website hosted on Microsoft Azure, written in markdown which is then turned into web pages by Hugo. The website is updated on my local machine, then the HTML is pushed to GitHub. That is then pulled onto the website, but only when Azure feels like it, there doesn’t appear to be a way to force the process to happen.

I’m on Azure because I can see the benefits of hosting this kind of simple static website online, especially if several games are run in parallel. That would mean that I can “copy and paste” the configuration into a new setup.

As per my comment back in March, the DNS for this specific game is also hosted on Azure. I still need to write up how the DNS works, as it took quite a fight, and then seemed to work by accident. Maybe if I run a second game, which will be at wwe2.conflict.games, I should document how I do that, and what problems I solve, rather than trying to “reverse-engineer” my own work for wwe1.

Flood map

A key part of the game is an authoritative map of the current state of the flooding, which feels like it’ll need something more than a static image on a wiki page. This map has the following requirements:

• be approximately true to reality, but minor discrepancies can be waved away.
• it can be edited by the game runner before we start, as the game will be set in 2048, so I’ll make some assumptions about how the landscape will have changed and add those in
• alterable by the game runner during the game, so it can react to player actions. In one of the playtests a team tried to build a “bridge” from where they’d evacuated to - in the emergency - back to the village they’d evacuated, a mile or two away. If successful, that would have made a difference, and would need to be reflected on the map shared by everyone.
• potentially have “local” or user-specific versions, so that a user can track future estimations of where the water might be.

I’ve no real idea where to start with this. On the one hand it feels like something I could “vibe code” together relatively quickly, that would be good enough… but also that might reveal bugs or errors only when the game has come to rely on it. Pointers to anything online would be welcome, although fitting in with everything else it does feel like self-hosting this, well, running it on a VPS, would make it easier to use and manage.

Other technical bits

This is all currently being built on a VPS hosted at Mythic Beasts. Their pricing and support seems ideal, and I like them as an ISP. If that changes for some reason my current alternatives to check out are FastPanda, or FarByte.

Nginx will be a proxy server sitting in front of all of this. That will use certbot to keep the SSL certificates up to date.

Being especially ambitious, hosting this all on VPSs also means that I can configure security and monitoring to whatever degree I see fit, rather than relying on online services.

Epilogue

If you want to see the previous summaries of challenges, please look over at the menu on the left.

If you want an idea of the other projects I’m trying to juggle alongside this, please look at my list of current projects and interests.

If you’ve got something very specific to our existing relationship to mention, then please do contact me directly - if that applies to you then you’ll know how. But I expect you’ve come to this post because I’ve mentioned it online somewhere, in which case, please do respond to this on that forum or social media platform. Not only do those “discussions in public” help put the game in front of new people, but also they might prompt others to see and respond to the point you’re making.