Heliodex
Joined 4 years ago
Joined 4 years ago
I'm puzzled by the -- @import comment syntax used in these main files. Did they have to build custom Luau tooling for type-checking across files for this, and why not use require()?
Other than this, the example games had easily readable and clear code.
Adjusting for educational attainment is an intriguing choice. If higher educational attainment were to decrease risk of Alzheimer's disease, wouldn't this adjustment optimise out the factor that they're looking for?
Otherwise very interesting results and conclusions. Also what's up with the outlying occupation on the left side of the 1st 2 charts on page 3: "Personal care and service workers, all other" with a mean age of death of 50.65?
Edit: https://web.archive.org/web/20260723015826/https://codeberg....
Unrelatedly, I can recommend Radicle, which hosts a growing network of crypto projects. https://radicle.dev/
My main concern with these types of programmes is that highly technical yet difficult to demonstrate projects (think programming languages, distributed systems, OSs, CLI tools etc) tend to get disproportionately bad results in voting. People new to programmes like this could be strongly put off by their perceived bad results, and people with knowledge of how the game works will just take their projects to other Hack Club programmes. If I were running this it would have 2 separate groups of prizes: larger ones based on number of hours spent (tracked with Lapse or Hack Hour, because I strongly dislike Hackatime), and smaller ones based on number of votes. However the big flashy prizes (in this case, the Framework laptops) really do attract people – recall that Siege <https://siege.hackclub.com/> was one of Hack Club's best performing efforts despite how easy it was to sink dozens of hours into it and win nothing.
Nonetheless I see the potential and I do hope for a lot of submissions – the programme clearly distances itself from the usual Hack Club status quo ;). Good luck with it!
Of course there are differences between LuaJIT and Luau that I think influence their decisions on possible ternary expression features:
- Luau users are disproportionately beginners to programming that I believe would find the if-then-else expression syntax easier to learn; LuaJIT developers have a larger user base of professional devs wanting to make their code faster, and they will probably be more familiar with the `x ? y : z` style since it's used in plenty of other languages.
- Luau is a lot faster moving in its development than LuaJIT in terms of language features, the Luau team just wanted to move people to ternaries from `x and y or z` because it's easier to optimise in a normal interpreter; LuaJIT, with their JIT, I assume would be able to more easily implement optimisations for constructs like `x and y or z` despite its slight semantic differences (my assumption on why the change is being considered now rather than earlier).
If there's any such evidence that I've missed, please do bring it to my attention by providing it.
We don't have to trust them. We can make an educated guess right here based on the analytical data Roblox provides, both for the entire platform & tailored to each experience, the price of Robux, and the DevEx rate. This is far from the data we need to make a precise prediction, though we can still get a ballpark estimate and check it against Roblox's own figures.
The most popular Robux package is the 1000 Robux one. Well, not really, it turns out it's actually the Premium 1000 subscription, which for the same sales price provides 1000 Robux alongside Creator Rewards for the experiences they join for the next 60 days. This could be anywhere from 0 to 3 * 5 * 60 = 900 Robux <https://create.roblox.com/docs/creator-rewards>, alongside a 35% share of what they spent on the package if they're a new or returning user.
The price for Premium subscriptions and Robux packages differs regionally. For me, it's 9.99 GBP for Premium 1000, whereas in the US, where most Roblox users are based <https://create.roblox.com/docs/production/roblox-user-base>, it's 9.99 USD. We'll go with the 9.99 USD figure as it's most likely what a paying user is actually spending and it's in the same currency as the DevEx return rate.
Assume a user that purchases such a subscription then spends all 1000 Robux on one or more passes or products inside of an experience. We'll assume that the creator of the pass/product is the same as the creator of the experience, so the 10% affiliate fee + 60% creator earnings makes for 70% for the experience developer, with the remaining 30% taken by the platform. This could just as easily be substituted for anything else with an equivalent 70% fee, such as Marketplace products at their lowest fee rate, or plugins/paid access priced in Robux (pricing these in local currency wouldn't be subject to the DevEx fees). This deposits a total of 1000 * 70%, or 700 Robux, into the accounts/ownership groups of the experience creators. Assuming the developers immediately take this Robux and exchange it, this will give them fiat. This results in a total of 700 * the DevEx rate of 0.0038, or 2.66 USD. This is profit for the developer.
Now for the Creator Rewards. We'll start with the maximum possible Daily Engagement Rewards of 450 Robux. This goes directly to experience creators, with no Marketplace/product fee, who will DevEx it to produce 450 * 0.0038 = 1.71 USD. This is the maximum possible figure and is almost never actually achieved on the platform. Next up, the Audience Expansion Rewards, which will be 9.99 USD * 35%, or 3.50 USD, if and only if the user is new or returning to the platform.
To sum it up, this is 2.66 from DevEx + up to 1.71 from Daily Engagement + maybe 3.50 from Audience Expansion. The DevEx figure is now correct, whereas the Daily Engagement and Audience Expansion figures could use some work.
From the most recent statistics available, Roblox has 36.7 million paying Monthly Active Users, or Monthly Unique Payers <https://ir.roblox.com/financials/quarterly-results>. However, only about 37.7%, or 13.8 million of these, are Daily Active Users. Thus we'll assume our average Premium subscriber is active 37.7% of the time, or about 11.5 days each month, so 22 of every 60 days. If they play only 1 experience a day, this is 5 * 22 = 110 Robux, or 110 * 0.0038 = 0.418 USD, and if they play 3, that's 3 * 5 * 22 = 330, or 330 * 0.0038 = 1.254 USD. To reiterate where these numbers came from, Daily Engagement is 5 Robux per day for 60 days after purchase, given to each experience played for more than 10 total minutes, with a maximum of 3. We're multiplying this by the DevEx rate as we assume that's what the developer does with it upon reception.
So 2.66 + 0.418 = 3.078 USD for the lower average, which we'll use for these calculation purposes, from DevEx specifically. That's the best estimate I can give with the time I have, the Audience Expansion figure I'll leave off entirely as it doesn't go through DevEx and I can't give any specific statistics for when it happens other than "sometimes".
So, normalised per in-experience dollar spent, that's a rough estimate of 3.078 / 9.99 = 30.8% which is exchanged through the Developer Exchange programme to be given to developers as profit. Compared to the 25-28% that Roblox shows on their Monetisation documentation page (pre-increase, this would be 27-31% now, also given there's some overlap with the Creator Rewards section of the chart which we also calculated the figures for), it's pretty close, without having to trust any of their figures (apart from their own earnings reports, which are heavily scrutinised and legally required to be sufficiently accurate). The remaining portion is split between Roblox and the developer, at a rate probably similar to that shown in the same chart. The same could be done for any developer which has a more accurate figure of Audience Expansion Rewards or Extended Services usage payments to work out more of the areas in the same chart.
> such a lopsided deal is still presented with such weasel wording.
It is very clear to any developer with experience with Roblox's economic system that this isn't at all lopsided, and is in fact fairly balanced. Based on an educated perception of an average user, I've derived similar figures to what is shown on their monetisation documentation, so I have strong reason to believe that they're accurate. If we want our very own estimates, the amount taken by the developer through DevEx is "about" 30%, and the amount taken by Roblox for their own investment is similar. This leaves "about" 40% to be spent on the developers for their own infrastructure. No real developer needs to calculate these figures for themselves, I've just done so here solely for demonstration purposes.
The combined payment of services and profit is the full package of what's provided by Roblox and it would be uninformed to claim it as anything else or try to say that the developer paying for their own similar services would be far more cost-effective. None of this is weasel wording nor exploitative of developers on Roblox's part. It's just a different deal with more supportive ways of earning and different parties paying for the same resources. If one were to prefer this deal, they should use the platform. If they don't, they shouldn't. It's not better or worse for either party, it's just different.
I see. I apologise for failing to pick that up.
> There is no single exchange rate between dollars and Roblox on a platform
> Meanwhile an account with 6k Roblox at time of purchase may have been filled by some gift cards and some App Store purchases at different ratios over time while the user is spending money on the platform meaning that 6k doesn’t actually correspond cleanly to either App Store or gift cards Roblox
Indeed there is not any single fiat -> Robux exchange rate, nor has there ever been. I don't have any reason to believe this is a bad state of affairs, and you've provided most of the reason for why there isn't. Even on platforms with no app store fees, there will still be payment processing fees, which the larger "better-value" packages exist to reduce these fees to give more money to developers in the case that there's a flat fee for each transaction.
> Ok sure you have some average numbers of Roblox per dollar and then work out an exchange rate sure that’s reasonable.
If the target is to have a consistent exchange rate, maybe this is doable for new Robux entering the platform. Perhaps moreso given that the price of Robux has remained remarkably consistent over the years, in no small part due to the various fees on the platform (mainly the Marketplace fee and experience pass/product fees) protecting it from hyperinflation. Yet there are billions of Robux already in the economy that don't follow these rules.
The Robux spent on an experience is probably earned from the purchase of Robux or Roblox Premium at a variable package exchange rate, also dependent on the currency of purchase. But it could be from Creator Rewards <https://create.roblox.com/docs/creator-rewards>, with the amount differing depending on when the player joined the platform and how many experiences they played that day. Or it could be old Robux from a package with an exchange rate that no longer exists. Or it could be from the Robux already paid out to an experience, which the experience developer decided to spend on another experience, thoroughly blending any previously distinct exchange rates. Or it could be from the legacy pre-2016 currency exchange system and not have any defined exchange rate. Or it could be from a user who previously owned a lifetime Builder's Club subscription, which is now grandfathered in to Roblox Premium, and receives Robux monthly despite only making a one-time payment.
My point here is to demonstrate that attempting to create such a fiat -> Robux exchange rate figure isn't really feasible, or that if one did exist it wouldn't be accurate. Nor would I think it to be particularly useful either; the main exchange rate that matters to developers is the Robux -> fiat one (for how much Robux is going into their experiences and how much they can DevEx from it), which is very clear and very well-defined – since September 2025, it's 0.0038 USD per Robux <https://create.roblox.com/dashboard/devex>.
> even if we aren’t saying what the actual formula is it’s fine. I can’t help but wonder what the actual numbers look like
The chart on the monetisation section of the creator documentation represents the "estimated utilization of each dollar spent in an experience on Roblox". Roblox has calculated this from their total user spending on Robux, the total Robux spent on games (which has also been through the experience pass/product fees clearly mentioned at the bottom of that page), and their total other expenditure (developer costs & support + their own share & investment), because as seen above, they don't have a consistent exchange rate to calculate player purchases based solely on the amount of Robux in circulation.
(comment continued in reply)