Does The Display Server Matter? The Latest Mir vs. Wayland Argument

[bison]

well FOSS dev comunity means the developers/packagers/distro representatives/bussiness/foundations/contributing individuals and the like not every living been in the universe as those are user communities and those are not related to the development process due to the lack of knowledge in technical matters.

everyone here refers as comunity the FOSS dev community not the user communities for ovbious reasons

If you exclude users who are not involved in the development process from your definition of the community, then you are excluding those who ultimately make the decisions on which software is widely used, and that which never catches on, or is used for a time until something better comes along. And there is nothing obvious about such a definition of the community -- it is simply a definition that you appear to like.

[toka]

Why do you all compare Wayland, a protocol with Mir, a display server?

As far as I understood, Mir is a display server, like Weston or like X.Org Server, but, while Weston supports only Wayland and X.Org Server supports only X11, Mir would/will support a multitude of protocols.
Correct or not?

Also, is there a Protocol, that is competing with Wayland and X11, called mir?

[bison]

That is completely false. They have agreed on dbus (KDE dropped their in-house dcop for it, for example), they have agreed on systemd (there are a few outliers, mostly Slackware and Gentoo, but Gentoo is working on making systemd a first class citizen on the same level as OpenRC, and even Canonical decided to transition from upstart to systemd). And they have agreed on Wayland. That's just a few examples, I doubt they're the only ones.

And by "they" I mean the people who actually write the code. Which is important, first and foremost you need buy-in from them, because without them there is nothing for the users to use. In the case of Wayland, "they" is driver developers, toolkit developers, DE developers. They all agreed to move in the same direction and have been working for quite some time to make Wayland happen. There is a phone out there that uses Wayland by default already - the Jolla. But then came Canonical and announced their project that they have been working on behind closed doors for nine months, and they seemed to just expect everyone to be on board with it. Well, not so fast. Canonical is not getting buy-in for Mir, and with good reason.

Very little is "completely" false. If you select definitions that appeal to you, then nearly anything can be made to be true. In this case, restricting the definition of "the community" to those who made the decisions on dbus and systemd. And while it is true that there would be no software for users to use if someone didn't write it, it is also true that there would be little motivation to write software if no-one were to use it.

[bison]

There is no entitlement. Canonical made a decision. They are free to do that. Martin never says otherwise. However, those decisions have consequences. Pointing out those consequences, and pointing out that Canonical is trying to downplay them, is not "entitlement".

If you disagree with his analysis of the consequences, feel free to point out the mistakes. But I don't see how merely explaining the flaws in Canonical's assessment of the effects of their decision could possibly be considered "entitlement".

I didn't associate "Canonical's assessment of the effects of their decision" with entitlement. In fact, I didn't address "Canonical's assessment of the effects of their decision" at all.

[TheBlackCat]

I didn't associate "Canonical's assessment of the effects of their decision" with entitlement. In fact, I didn't address "Canonical's assessment of the effects of their decision" at all.

Huh? I didn't claim you did. I said you associated Martin's assessment of Canonical's assessment with entitlement.

[TheBlackCat]

As far as I understood, Mir is a display server, like Weston or like X.Org Server, but, while Weston supports only Wayland and X.Org Server supports only X11, Mir would/will support a multitude of protocols.
Correct or not?

No, Mir has a protocol, it is just not public and not stable. It requires that all clients access it through an API, but it nevertheless still has a protocol working behind-the-scenes. Wayland has an API as well, but also allows clients to access the protocol directly. And, although they have the same goals, the architecture of the two is fundamentally different. Even if Mir could support multiple protocols, it would almost certainly require major fundamental changes to make it compatible with Wayland.

[toka]

Wikipedia adds to the confusion

This article here: https://en.wikipedia.org/wiki/Mir_%28software%29 only mentions the fact, that Mir is a display server. The same is true for its talk page: https://en.wikipedia.org/wiki/Talk:Mir_%28software%29

BTW, there is an article for this kind of software:

https://en.wikipedia.org/wiki/Display_server

The mir -article omits the fact, that mir __also__ introduces a new display server protocol.

[jrch2k8]

If you exclude users who are not involved in the development process from your definition of the community, then you are excluding those who ultimately make the decisions on which software is widely used, and that which never catches on, or is used for a time until something better comes along. And there is nothing obvious about such a definition of the community -- it is simply a definition that you appear to like.

when you wanna debate about a piece of engineering that probably rank among the hardest 3 to do in the entire industry, you ask the expert and higly technical asset in an technical review board to paint the scenario, then after the goal is clear then you present it to the user and consumers(still is not the final user at this point but DE/toolkit devs) and retrofit the feedback, then you release a product based on it and retrofit again against the layers of that product(here is granma say it likes or not).

so for the users standpoint any technical difference between X11/Wayland or mir is no more than alien language jibberish, hence you have a point if this articles were in ubuntu/arch/suse/fedora user forums asking about how it feels either protocol/server implementation but when an user come to a technical discussion to drop fanboism about something can't possibly understand is implicit the above definition come in to place, hence community in this technical sense exclude automatically any regular user because its ovbious, that is why NASA don't call your grandma to ask if she like the formula they are running in a supercomputer to predict if tomorrow is gonna rain, your granma judge the android/ios/etc application that show her the result

[toka]

granny speaking here: is the "display server" = mir the problem? or is the "display server protocol"=libmir-client + libmir-server?

A dispute with less fanboys: http://lwn.net/Articles/541336/

[GreatEmerald]

As far as I understood, Mir is a display server, like Weston or like X.Org Server, but, while Weston supports only Wayland and X.Org Server supports only X11, Mir would/will support a multitude of protocols.
Correct or not?

Also, is there a Protocol, that is competing with Wayland and X11, called mir?

This was already answered, but it's actually the main issue with Mir. Everyone would have been happy if they made Mir a Wayland client (or whatever is the right term for it) instead of a display server.

And while it is true that there would be no software for users to use if someone didn't write it, it is also true that there would be little motivation to write software if no-one were to use it.

Hardly. At least for me, all the motivation to work on anything is for my own gain. Being able to share it with users is just a bonus. Collaborating with more developers is another bonus. But not motivation.