A Brief History
About three and a half years ago, after a rather disastrous loss of hosting, the active team and many retirees pulled together to rebuild from the ashes. For those around at the time, it's fairly well known that LoU did all of the backend work setting up the website, forums, and development infrastructure, among other things. Less well known is that our main hosting and beta testing servers were generously provided by Hexodious. Alas, all good things come to an end...
The team was given notice at the end of last year that this arrangement would need to come to an end soon, so we have been assessing our options. While the timeline for completing the move is undefined (no need to panic over an encore of 2015), the consensus of the team is to try and get it done ASAP to relieve Hexodious of the ongoing responsibility.
And of course we would like to formally, and publicly send our enormous gratitude to Hexodious for what has been his large personal financial contribution over the past few years. We as developers and players are forever in your debt.
The Future
While I am retired from development (and not planning to change that status), I've still kept tabs on goings on with the team and done a few things like recreating the web serverlist. Since the start of the year I've been sizing up hosting options and mentally spitballing the problem. Last month I volunteered to take on the task of migrating everything.
As already mentioned, this time around we have a bit of time to consider infrastructure and making improvements. While the current hosting has served us well over the years, its main weakness has been the lack of network capacity at release time. Any large updates (> 1 GB of files) usually result in very slow download speeds on release day up until enough of the new assets are cached with the CDN. While Defiant's refactoring of launcher downloading has greatly helped with asset corruption associated with that congestion, the update speed is still a problem. This means we want the new host to have both better upload speed and enough data transfer to avoid overage fees or throttling.
Another consideration is security. Historically it's not uncommon for official and community servers and services to come under DDoS attacks. As disappointing as that is, it's still a reality that demands planning for. This constrains us to only picking among seasoned hosting providers (won't fold the moment an attack starts) and carefully planning infrastructure to help compartmentalize the effects. All of this ultimately increases hosting costs.
Lastly, since we are essentially refactoring most of the infrastructure anyway, I am aiming to document as much as possible so it may be more easily maintained and modified in the future. This is definitely a non-trivial task, but in my experience it almost always pays dividends (such as making any future migrations like this much easier).
As for timing goes, the first major part of the migration (moving these forums) is planned to take place this Friday morning (GMT). There will be at least some brief downtime that day.
tl;dr: The new hosting infrastructure I've planned will hopefully bring some improvements for the trouble, but the end result will still be a doubling of annual project expenses relative to last year.
Step 3
????
Non-Profit
So what are those expenses? That's a question I had to research myself when I started looking at this project. Here is my projection for the next three years:
- Hosting: $19/month - $228/annual
- websites, development, master, regional officials
- Fees: $95/annual
- domain renewal, forum license maintenance
- Other: $300 (budgeted at $100/annual for expected minimum lifetime)
- hardware dedicated to offsite backup storage
- Total: $35/month - $423/annual
While we've had many donations over the years, the majority of hosting expenses have been covered by current and former developers. It's a great practical example of how passionate we are for the mod, but I don't see it as a sustainable model. Even if development stopped forever tomorrow and we cut things down to the bare minimum, it would still cost around $200 a year to keep the remaining servers running.
So as you might have seen coming a mile away, here's the pitch. Even if you don't have the time or knowledge to work on the mod directly, you can help with keeping the lights on. By our current estimates the mod's active population is in the neighborhood of 1000 unique players. That means if even only half the community contributed $1 we'd be set for the year!
We're considering some options like setting up a donation meter to make funding levels easily visible, but that won't be coming until after the migration. For now I've written this post to highlight the funding problem and try to make things more transparent. Feel free to let us know if you have any suggestions. And a big thanks to anyone contributing past, present, and future!
Last edited:
