💾 Archived View for gem.sdf.org › jdd › posts › 20220312-desktop.gmi captured on 2022-05-22 at 21:19:24. Gemini links have been rewritten to link to archived content

-=-=-=-=-=-=-

The year of gemini on my desktop

<2022-03-12 Sat>

Any time Linux makes any sort of gain in the consumer marketplace (eg. the Steamdeck) some wag will inevitably proclaim "The year of Linux on the Desktop!" Mostly this is in the spirit of poking fun, a light-hearted reference to the early days of Linux when one could dream that it would supplant Windows in the corporate and consumer desktop space, and in so doing would plunge a silver dagger into the very heart of the evil Microsoft empire. Now of course, older and wiser, we realize it will never be the year of Linux on the desktop ... just the year of Linux on pretty much everything else.

The thing is, I'm not entirely sure why we ever thought "the Year of Linux on the Desktop" would be a good thing. Had Linux ever become truly popular on consumer desktops it seems very likely that its most popular variant would be little better than what it replaced. Linux would need to be so user friendly that your average non-technical computer user would willingly dump Windows for it. Getting Linux to that point would require a fairly significant investment not just in development but also promotion and support. Which means serious private investment. And since it's doubtful such a product would sell, there's no very good way to monetize it aside from the tiresomely familiar avenue of surveillance-based advertising, so guess what would get baked into this new Linux distribution? At this point, desktop Linux would be subject to the same kinds of market forces that lead to walled gardens, and cloud spyware, and all the rest of it. Of course, there would still be other distributions with all this stuff stripped out, but "normal" users wouldn't even know they existed, or go to the effort of installing them.

If you had told me in 2002 that in 20 years time the dominant browser would no longer be Microsoft's Internet Explorer, but rather a (mostly) open source browser called "Chrome", that would have sounded almost too good to be true. And if you had told me that Microsoft themselves would build their new browser, Edge, on top of Chrome I might have wondered what you were smoking. But again, this turns out not to be the utopian scenario I might have once thought it would be. For Chrome, as we know, is a surveillance nightmare, engineered to maximize ad revenue for the corporation that spawned it.

So why is Linux (despite its increasing corporate-ization) still reasonably user-respecting, while Chrome is not? I would argue the main reason is precisely because Linux is not widely popular as a consumer desktop OS, and mostly exists to run on servers and other specialized devices. So it remains open, customizable, and unencumbered (at least out of the box) with telemetry, surveillance and advertising. A collection of tools, rather than a collection of products, built for folks who want to understand it rather than those who would rather its inner workings be hermetically sealed away so they don't have to.

Thoughts like these lead me to hope that gemini never becomes truly popular.

Now, frankly I doubt there is much danger of that. There is little in gemini that would, I think, appeal to more than a smallish group of computer hobbyists and disgruntled oddballs (not that the two groups are mutually exclusive, citing myself as an example). But it is worth contemplating what might happen if gemini ever did gain significant traction.

The first thing that would happen, I think, is that it would collapse into the web, much as gopher did in the 1990s. If major web browsers ever added support for the gemini protocol, if major search engines started indexing gemini space ... it would be pretty much game over in terms of maintaining gemini as its own thing, distinct from the web. If you could transit between the web and gemini space seamlessly, gemini capsules would become little more than curiously retro-looking web sites. And how long do you think it would be until web browsers started supporting non-standard markup in gemini pages? You know, to better allow you to monetize your content by including a few ads and some tracking cookies?

But wanting gemini to remain unpopular, reserved for those who are willing to put at least a little mental elbow grease into understanding what it is and how it works and why it is the way it is ... isn't that kind of elitist? Why shouldn't gemini be as accessible as, say, Twitter or Facebook, so everyone can participate?

Maybe it is elitist, a bit, but I can't see that as a problem. It might be if gemini was the only game in town, but there is no shortage of other options out there. Gemini has the character it does precisely because setting up a capsule involves a small amount of technical overhead, and browsing gemini space involves (a) knowing it's there, and (b) installing a gemini client. As long as these conditions persist it is unlikely gemini will ever become very popular, and so will retain value as an alternative to the web. Should it grow too much beyond its current niche, there is a very real possibility the web will eat it alive.

Up to Index

The year of gemini on my desktop was published on 2022-03-12