Apple rumors are perennially popular for two major reasons: First, the company’s actual announcements are so few and far between that only “what’s next?” speculation keeps its fans excited year-round. Second, the company is famously iterative, rarely entering a new product category without years of research and a multi-year “what’s next?” product pipeline in mind, constantly feeding the speculation.

If you follow the rumors, augmented reality hardware has been Apple’s next big thing for at least a couple of years. Following the release of ARKit software for developers, Apple CEO Tim Cook teased that Apple was working on something bigger, and this summer, a beta release of iOS 13 all but established that an Apple AR headset is headed to market. While it’s expected to be iPhone-dependent, no one’s quite sure how the headset will work, or what it will look like.

This week, hit-and-miss Taiwanese supply chain publication Digitimes reported that Apple is working with software and VR giant Valve on an AR headset. While this is the same Digitimes that said Apple cancelled its own AR device earlier this year, the latest claim is that the companies are now collaborating on a design that will be “released in the second half of 2020 at the earliest,” with assembly to be handled by two Taiwanese contract manufacturers.

Given almost everything — Digitimes’ reputation, Apple’s do-it-ourself approach, and Valve’s own “we’ll do it on our timeline” track record — it would be easy to write this report off as fantasy, or more generously, as wishful thinking that will likely bear no fruit. Unsurprisingly, there’s been no confirmation from either Apple or Valve, and there may never be any.

However, something weird in the aforementioned beta release of iOS 13 lends credence to the report. Instead of finding references to a single AR headset in the beta, developers spotted codenames and specs for not just one but multiple AR headsets, each with significantly different display characteristics. They were all tied into a secret feature called “StarBoard,” or “stereo AR board,” which allows an iPhone with an A13-series chip to output stereoscopic 3D objects to an external display device.

In the apparent leadup to the release of an AR headset, Apple has researched and attempted to patent all sorts of high-performance AR headset display solutions, including components with retinal projectors and micro mirrors that will prove tricky to produce at scale. Considering all the publicly released patent applications together, it’s clear that there isn’t just one Apple vision for AR, but rather many variations on the theme, some of which would be much harder to make than others. Meanwhile, companies such as Nreal are months away from bringing viable Android phone-tethered headsets to market for around $500.

That’s one reason that working with a partner — or partners — could make sense. Apple might co-develop one headset with a gaming company such as Valve, while it works with an optical company or an athletic fashion company on another headset. Each company’s headset could be optimized for a different demographic, which might explain why the iOS 13.1 beta referenced “Luck,” “Franc,” and “Garta” headsets with 58, 61, and 68 degree diagonal fields of view, all larger than the displays in current AR glasses.

It’s worth noting that Apple typically only works with partners on products when it determines it can’t bring the “right” design to market on its own for one reason or another. Years ago, Apple designers collaborated with Harman Kardon on Mac speakers, and Belkin on early iPod accessories, foreshadowing the growth of a robust and successful Apple accessory ecosystem.

More recently, Apple attempted to corral third-party developers to create game controllers for iOS devices. The company knew the features it wanted to support, but didn’t want to actually make the controllers itself. So it created technical documents outlining several different controller options, then roped third-party developers into manufacturing and selling them.

Though it’s mostly forgotten at this point, that experiment was a disaster. The competing alternatives hit shelves, sold poorly, and were discounted until they disappeared. Then Apple expanded controller support from iOS to Apple TV, encouraging followup models from SteelSeries and Hori. After years of going nowhere, Apple gave up on its own requirements and began supporting Sony PlayStation and Microsoft Xbox One controllers. What was the point of all that experimentation if it was just going to lead consumers to use the same hardware rivals had already perfected years earlier?

Another largely forgotten example is Motorola’s ill-fated ROKR E1 phone. Apple partnered with the cellular giant on a famously terrible candybar-shaped handset — including iPod-ish music support — just before it revolutionized the industry with the iPhone. Anyone who excitedly bought ROKR in September 2005 not only had a seriously mediocre experience, but was effectively abandoned in favor of “what’s next” when the iPhone was introduced in January 2007.

Assuming the Digitimes report is correct — a stretch at this point — it’s unclear what Apple’s angle with Valve would be. On one hand, it might be relying on a proven VR hardware developer’s expertise to bring something especially ambitious to market. Perhaps Valve has some patents or other intellectual property Apple needs to succeed. On the other hand, Apple may indeed have given up on some technically infeasible AR device and just want to get some hardware into the marketplace. Valve has been working on AR for years, and may well have had something more viable than whatever Apple’s been working on.

As a potential customer for this product, my question is whether the end product of an Apple AR headset collaboration with another company would go down the game controller or ROKR path — in other words, a stopgap product until a better competitor or the real Apple design is available? Or might Apple partner with another company in a long-term fashion, achieving something it cannot and will not be able to do alone?

A deal with Valve would give Apple the sort of mixed reality and game industry credit it seemingly hasn’t been able to either build or buy despite years of effort. It remains to be seen whether that’s a Digitimes fantasy, one step down a much bigger path, or the end goal of Apple’s AR efforts.