Wayland, Mir, and X - different projects

The longer I think about Mir and the new Mir dependent Unity, it looks like Canonical wants its own unique display stack a la Quartz and Aqua. Quite possibly to create a space where apps need to speak Mir/Unity to be able to run on Canonicals platform. Hence their SDK.

If that is the strategy, it is a brilliant one from Canonical's standpoint. Even if they keep it FOSS (not necessarily with the CLA, they could close it up), it will be so uniquely tied to Canonical and Ubuntu, that users and developers buying in to that platform are virtually locked in through the Canonical specific API's.

Other distro's are highly unlikely to adopt Unity/Mir and Unity/Mir apps, as that would make them second rate Ubuntu clones. If Canonical succeeds in this endeavour, they wil have successfully transformed themselves into an independent OS supplier and they will indeed be serious competition for MS and Apple. Their userbase seems to be highly uninterested in the broader Linux landscape and very much interested in "Works for me" and "I don't care how Canonical makes their money, as long as I get my Ubuntu as a freebie".

Me, I'll keep using what is known as a GNU/Linux distro. I have no need for another lockin platform. Manjaro looks "tastier" by the minute.

The article's relationship with the Mir project -- whether directly as a developer or indirectly as an Ubuntu/Mir cheerleader -- wasn't apparent until well into the article.

I fear that these projects are going to do to the Linux desktop and its adoption what the proprietary hardware vendors did in the '80s and '90s with their distinctly different and slightly incompatible flavors of UNIX. At some point, when Gnome and KDE rewrite all their applications for Wayland, the rest of the Linux desktop community will be ghettoized and applications that don't require (or need) those desktop environments will stop getting updated and eventually die out. Sure, the Wayland project says that there will be X11 compatibility and network transparency via a "plug-in" or some such mechanism but something tells me that once the major desktop projects are fully in line with Wayland, the delivery of that plug-in will be get pushed out to the future again and again and again until -- and I'm taking bets on this -- the response to the end-users being something like: ``Hey! The protocol is published... just write your own. Now quit bugging us.''

But... I'm certainly willing to be -- even hoping to be -- proven wrong but given what we've been seeing happen to Linux as pushed by the major commercial Linux vendors I'm not encouraged by projects like Wayland and Mir.

Lets not get our panties in a bunch just yet... This is all more ideas than code at this point.

Wait and see how everything pans out - I'd be willing to bet people are over reacting to this whole thing.

~Jeff

Wayland and now Mir are local display systems designed for gamers and cellphones. I haven't seen anyone suggest they have any concept of how to communicate over a network in native mode. That is what X does very well and why Wayland has a specified X-Windows layer to communicate over networks and make use of the huge volume of applications written to use X. Over networks X is the way to go and that isn't likely to change anytime soon.