> We should mandate that device manufacturers set aside a portion of the purchase price of a gadget to support ongoing software maintenance, forcing them to bud

> We should mandate that device manufacturers set aside a portion of the purchase price of a gadget to support ongoing software maintenance, forcing them to budget for a future they'd rather ignore. I think the problem is that any portion of the purchase price will eventually run out over time. Let's say that I purchase a $1,000 phone and the company sets aside $250 for ongoing software maintenance. How much should they budget when? If they spend the money too quickly, it will run out too quickly. But no matter how prudent, it will run out. I think the problem with software is that it's often unmaintainable by the end user. When you buy a house or a car, you don't expect infinite free maintenance of it. You simply expect the ability to have third parties maintain it for a fee. With closed-source software, you don't really have the option to hire a third party to maintain the software for your PSP. However, if we're expecting the original manufacturer to support it for free, that support can't be indefinite. Any amount of the purchase price put away (even 100%) would eventually get exhausted. > Or maybe they aren't ignoring the future so much as trying to manage it by speeding up product obsolescence, because it typically sparks another purchase. I think this is one of the reasons that so many companies have moved to subscription models. If there's ongoing revenue, there's an ongoing reason to support it. Even then, if consumers have moved on, there might not be enough revenue coming in for that device to justify working on it. If there are 12 people who still want to use a PSP and they're paying $10/mo, that won't cover one engineer to keep updating the software. I'm guessing there are more than 12 people who'd like to use a PSP, but you can understand that there is some point at which there aren't enough users even with a subscription model. In a lot of ways, it feels like the author is making two arguments. 1) Companies should provide free updates for their products forever via a single purchase. That strikes me as a weak argument. Nothing lasts forever. Too many things are abandoned too soon and that is a problem - one that could be handled by some more modest proposals like supporting devices for 6 years or something. 2) We should have a right-to-maintain. This is somewhat at odds with the original one. The right-to-repair isn't about free repairs from the manufacturer, but about our ability to maintain our own things (or get third parties to maintain them). Maybe a right-to-maintain could be useful without going the direction of totally free software. It seems like it might be difficult to mandate this, but one could imagine saying that consumers should have access to the source code to make modifications to maintain abandoned systems. However, I think there are issues in general. As more items rely on services that cost money to operate, at some point a device manufacturer will want to stop paying for those services if they're no longer used and for a product they no longer sell. I think it's also important to think about what this means for products that were financially unsuccessful to begin with. When we think of something like the Playstation, we think of something that makes a lot of money for Sony. What about a device that sells poorly and the company has lost a lot of money on the device itself? I'm not talking about some sort of loss-leader where their strategy is to make up the money some other way. I spend $100M creating a new smartphone and I sell 100,000 of them at $500 a piece with a cost-of-goods of $200. I've spent $120M and taken in $50M in revenue. Having to support these devices for decades would mean that companies wouldn't challenge incumbents. I'm not going to go up against Samsung and others if I might get saddled with a huge maintenance burden. You'd only be able to launch a product if you knew that you were going to be able to handle the maintenance burden. That would probably make the market a lot less competitive than it is now. So many companies wouldn't try things for fear of getting saddled with large, ongoing maintenance costs. You'd rather not try your chances at competing with incumbents and consumers would be left with fewer options. I think some of that could be accommodated by tying maintenance to number of units sold or profits achieved. However, that would mean the maintenance window for a product depends on how popular it is. Of course, we already see this a bit in the real world. Companies do maintain popular and profitable products longer than unpopular ones. If we're asking manufacturers to "set aside a portion of the purchase price" then we're giving cheaper and less popular devices less money for maintenance than more expensive devices. That seems to defeat the purpose of buying with the confidence that it'll be maintained. In fact, you'd want to spend your money with large incumbents rather than supporting a new competitor since the incumbent device will likely sell well and be supported longer. I agree that devices should have their software maintained longer. It's one of the things that annoys me so much about the Android ecosystem. I just don't see anything in this article that proposes a workable solution. Free maintenance forever isn't workable and would likely distort the marketplace in other ways with reduced competition since the burden of an unsuccessful product isn't just the immediate losses, but the ongoing maintenance losses. While the author briefly talks about a right-to-maintain, the idea is never actually explored and the article is dominated by the idea of the original manufacturer maintaining things indefinitely rather than us having the ability to maintain our own software. I think a more limited scope proposal might include things like a limited maintenance commitment in established categories where we've seen the lifecycle of devices and such (ex. one could argue for something in the 3-7 year range for smartphones); unlocked boot loaders so that third-party software could be run on hardware; a requirement that key parts of source code be released (even if the release only allows the source and modifications to be run on the original hardware, ie. not free "FOSS" software) if the devices are no longer maintained so that third-parties could offer service for it.