More Of KDE Is Getting Hooked Up, Running Properly On Wayland

[phoronix]

Phoronix: More Of KDE Is Getting Hooked Up, Running Properly On Wayland

KDE developers continue making progress on porting to Wayland...

http://www.phoronix.com/scan.php?pag...nd-Blur-Hookup

[Awesomeness]

I had hoped for a more usable Wayland session in 5.4, however KWin does absolutely no window management in that release.

[mgraesslin]

I had hoped for a more usable Wayland session in 5.4, however KWin does absolutely no window management in that release.

Why did you hope for more? We never said it would do more. Quoting from the release announcement:

The currently supported feature set is driven by the needs for the Plasma Mobile project and more desktop oriented features are not yet fully implemented.

Apart from that window management for XWayland based windows is working fine. Only for Wayland windows the window management is not yet complete. Nevertheless we decided to go for all apps on Wayland, while we could have tricked people into thinking all works fine by just forcing all apps to XWayland.

[Awesomeness]

Why did you hope for more?

One can always hope? and hope is something different than expectation.

We never said it would do more.

Quite frankly, you didn't even say much. You never do. From time to time to release blog posts with technical mumbojumbo about very specific bits in the implementation but you are shy of answering straight questions of when will the implementation be how good. Even ?Wayland support will improve in future releases with the aim to get to a stable release soon? does not say much.
So what's the roadmap? Movable windows and usable for early adopters by 5.5? 5.6? Sometimes after Unity 8?

If you don't want to answer those questions: Fine. You can do whatever you want but don't act surprised if you get such user reactions.

The currently supported feature set is driven by the needs for the Plasma Mobile project and more desktop oriented features are not yet fully implemented.

Exactly what I meant. Few people know the exact feature set of a mobile compositor and how is any normal person supposed which user-facing features of KWin have been ported to KWin_Wayland, anyway?

[scionicspectre]

you are shy of answering straight questions of when will the implementation be how good.

Just because it's a straight question doesn't mean it will automatically have a straight answer. You're essentially asking him to lie. This isn't a usual development task, and even rudimentary work in such a complex piece of software can be unpredictable and unexpectedly burdensome.

I would think you should expect users who care about this sort of thing to understand the difficulty and uncertainty involved, not to react in a condescending manner. To assume some of the hardest-working developers are just being coy about their work probably isn't the best way to go. We're all excited for Wayland, but unless you're working on solving the problem it's probably best to be patient or perhaps even grateful for the work being done on your behalf.

[Awesomeness]

Just because it's a straight question doesn't mean it will automatically have a straight answer. You're essentially asking him to lie. This isn't a usual development task, and even rudimentary work in such a complex piece of software can be unpredictable and unexpectedly burdensome.

That's bullshit and you are an idiot.
Even for complex tasks roadmaps exist but he likes to keep others in the dark and prefers to play stupid games instead:
https://plus.google.com/+MartinGr%C3...ts/C1RDQP8z35t
https://plus.google.com/+MartinGr%C3...ts/FenknJX5Ls3
https://plus.google.com/+MartinGr%C3...ts/j6VnxThuPhp

[scionicspectre]

Sorry for the late response, I don't sign into Phoronix often enough.

That's bullshit and you are an idiot.
Even for complex tasks roadmaps exist but he likes to keep others in the dark and prefers to play stupid games instead:
https://plus.google.com/+MartinGr%C3...ts/C1RDQP8z35t
https://plus.google.com/+MartinGr%C3...ts/FenknJX5Ls3
https://plus.google.com/+MartinGr%C3...ts/j6VnxThuPhp

He's just trying to enjoy the work and have a playful dialog with the community as progress is being made. The problem with making a concrete roadmap for something like this is that obstacles will inevitably come up and destroy your expectations of what can be done by when.

GNOME's initial projections put full Wayland support at roughly a year ago and it's still got one or two rough patches in 3.18. The goalposts kept shifting for them and they had support from third parties to get the work done. I think it's prudent for Martin not to set any unrealistic expectations or promise something he can't be certain of.

Of course, that doesn't mean there's no point in asking him more detailed questions about Kwin's implementation of Wayland support- in fact, that would probably yield more favorable discussion. But asking him to boil it down to some newspaper headline simply isn't productive. The release announcement is the most basic, straightforward description you're likely to get.

[Awesomeness]

He's just trying to enjoy the work and have a playful dialog with the community as progress is being made.

I don't think Blue Systems pays him to play those games.

The problem with making a concrete roadmap for something like this is that obstacles will inevitably come up and destroy your expectations of what can be done by when.

Don't put words into my mouth. I never wrote "concrete". In development roadmaps never are concrete.

GNOME's initial projections put full Wayland support at roughly a year ago and it's still got one or two rough patches in 3.18. The goalposts kept shifting for them and they had support from third parties to get the work done. I think it's prudent for Martin not to set any unrealistic expectations or promise something he can't be certain of.

A statement like "We hope to get something somewhat usable by Plasma 5.6. Let's see how that works out." is more than enough.
They had no problems making a similar statement for Plasma Mobile: to hopefully release a usable product within one year.

[scionicspectre]

Don't put words into my mouth. I never wrote "concrete". In development roadmaps never are concrete.

Okay, I suppose I had a misunderstanding of what answers/information you may consider 'straight', to take a word from your mouth. I'm glad you've allowed for a reasonable amount of leeway in regard to roadmaps.

A statement like "We hope to get something somewhat usable by Plasma 5.6. Let's see how that works out." is more than enough.

Yeah, I think Martin could be persuaded to give a statement like that fairly soon given how far Wayland support has come, recently.

Also, when it comes to Martin not being employed to 'play games', I don't think a playful attitude and a love for your work necessarily constitutes playing games with people. Most who responded to those Google+ posts seemed to enjoy the back-and-forth on those small features, so I'd say very few people were bothered by it. Communities react badly when a contributor's behavior changes after being employed to do what was previously volunteer work; It's generally considered taboo for a company to ask contributors to do anything differently in their development activities or political interactions.

I get the feeling that you don't think he's being professional enough. It could be that you're not very attached to that idea, but you're using it to legitimize your feelings that Martin has disregarded what you feel is important. I think you care a lot about Wayland adoption and seeing desktop Linux thrive, and those emotions are making obstacles to that vision of the future easier to criticize or feel spite towards. Of course, this is just an armchair analysis and now I'm guilty of putting thoughts in your head as well as words in your mouth, but I don't generally think you would take an issue like this so seriously unless you genuinely cared about the consequences.