saturnian Posted April 3, 2014 Share Posted April 3, 2014 Blog post by Aaron Seigo: http://aseigo.blogsp...oes-matter.html My hope is that Canonical will see the same light when it comes to display systems and, instead of stretching Mir out into 2016, will commit to supporting Wayland along with the rest of the free software world. It is their right not to do so, but I still hold out hope that they will. 1 Quote Link to comment Share on other sites More sharing options...
Guest LilBambi Posted April 3, 2014 Share Posted April 3, 2014 Thought provoking article! Thanks for posting it, saturnian! Quote Link to comment Share on other sites More sharing options...
Hedon James Posted April 3, 2014 Share Posted April 3, 2014 yes, thanks for that article saturnian! interesting read, and thought provoking. while a lot of this developer-speak and technical talk goes over my head, I think I'm able to parse the main arguments. But what I just can't wrap my head around are the followings statements/issues: 1. If the display server is to be invisible to the user, why the need for a second project (Mir), when the first project (Wayland) appears to be headed in the right direction? 2. If "Mir and Wayland do not do anything of value significantly different. Neither design offers an advantage over the other in terms of user facing value", I must again ask the same question? It makes no sense to develop an alternative display server just for the sake of seeing if you can. I have yet to hear WHY Canonical is doing this. I hear Canonical state in a generic fashion that "Wayland will not suit their needs" without saying what those needs are, and how Mir will remedy that. IF TRUE, this would make sense to me. But I haven't heard a viable reason to believe this is true. Alternatively, I hear others speculate about Canonical's secret motivations to control this portion of the graphics stack. IF TRUE, this makes no sense, as I fail to see how Canonical will make $$$ from a proprietary license when a FOSS alternative is equally viable and readily available. And I don't believe that Canonical is in business to develop software technologies just for sport. Inasmuch as neither of these 3 scenarios make sense to me, I'm still wondering "WHY?" 1 Quote Link to comment Share on other sites More sharing options...
Guest LilBambi Posted April 3, 2014 Share Posted April 3, 2014 That is one of the concerns certainly, Hedon! And I am not entirely sure that I would go so far as to say Wayland is full FOSS either. Certainly better than Mir, but as open as X? And he also said that Wayland and Mir weren't appreciably better than X. Quote Link to comment Share on other sites More sharing options...
Guest LilBambi Posted April 4, 2014 Share Posted April 4, 2014 No need. Like everywhere else, competition makes for a better product. Whether you pay for it or not. We are by the way not talking about the backend protocols used to deliver the GUIs when we talk about things like OpenOffice.org and Document Foundation's LibreOffice. And the licenses are not exactly the same there either btw. Quote Link to comment Share on other sites More sharing options...
raymac46 Posted April 4, 2014 Share Posted April 4, 2014 Over 100 years ago F. W. Taylor (the father of Productivity Analysis) came up with the three attributes necessary to get things working better: Specialization Simplification Standardization Seems to me an ideal display server would have all 3 of these attributes. Specialization is part and parcel of the Unix way of doing one thing and doing it well. Simplification means getting rid of a bunch of unnecessary stuff - in the case of X this is leftover 1980s graphics technology which is redone by hardware and window managers. Standardization means that the display server can work nicely with any desktop environment, window manager, or hardware implementation. It's early days but as I see it there's only one candidate who can conceivably deliver the Triple S - and that's Wayland. Go to it, boys! Quote Link to comment Share on other sites More sharing options...
Hedon James Posted April 5, 2014 Share Posted April 5, 2014 (edited) Saw this in the linked article about Linus being angry with developer Kay Sievers, author of systemd software. http://www.muktware....ly-broken/19811 This article was very informative to me, and helpful in understanding the bigger picture. I saw this article as a bridge that ties 3 recent threads together on this board: Linus Angry...; Ubuntu and the Unspoken Rules; and this thread "why the display server matters." This is like peeling an onion, where removing a layer only reveals another layer, which looks like an orange; but peel the orange, and it looks like a banana; peel the banana... you get the idea. Referencing Linus' point of view, perhaps Canonical's CLA isn't as big of a deal as people are making it out to be? As I understand CLA in general, the whole purpose is for the "distributor" or the "aggregator" of contributions to be able to defend those contributions from copyright challengers. It isn't practical for hundreds/thousands of smaller contributors to be expected to defend each and every potential challenge, assuming you can even locate them to inform them of such a challenge. Without a viable CLA, I imagine patent trolls would view Linux development as the holy grail of panning for patent gold?!!! Based on this understanding, as Linus puts it "people love to hate on Canonical." (back to the tone of voice argument? not so much what Canonical says, but how they say it?) Conversely, in all fairness, I must question Canonical's motives for changing the language in the Project Harmony CLA that Canonical used as a model. Why not just use Project Harmony's CLA as is? I'd like to know Canonical's thought process behind that. In the absence of that, can I be a devil's-advocate? Even though Canonical's CLA provides for the ability to subjugate code and re-release as proprietary, when have they actually done that? To the best of my knowledge....never. In all seriousness, does anyone here know of an instance where they have? And in the absence of an example, what makes people think that Mir will be the first? Just because you CAN do something, but NEVER have, doesn't mean you WILL. Which brings me back to Canonical's objectionable tweak to their CLA...why? This seems to be the elephant in the room. All that said, I can't help but speculate on a few things, based on nothing more than application of Occam's Razor and conjecture. Somewhere in these 3 previously referenced threads, I referenced an analogy regarding Init, Systemd, and Upstart; someone else linked an article mentioning the same thing. Based on Linus' rant against Kay Sievers and allegations of "breaking other people's processes in an un-apologetic fashion", I would speculate that perhaps Ubuntu wrote Upstart to avoid these experiences with the fledgling systemd, and even beating them to a finished product. (Otherwise, why would Kay Sievers develop systemd to compete with an already established and well-functioning Upstart that had already been adopted by several major distros. And why didn't Kay Sievers get fussed about dilution/duplication of efforts?) Later, with systemd in a more mature state, the Linux community arrived at a consensus that systemd was a better solution. Either that, or "Ubuntu are bigger jerks than Sievers, so we'll use Sievers' systemd." If that is anywhere close to reality, this might explain what appears to be Ubuntu's duplicitous effort in developing Mir rather than contributing to Wayland. If that's not it, is it possible that Wayland addresses today's graphics issues in a modern fashion, while Mir addresses tomorrow's anticipated graphics issues in an extensible fashion? To the best of my knowledge, Ubuntu is the only distro pursuing convergence across devices. If this is the reason, how does Mir allow this in a fashion that Wayland will not? I don't know these answers, and my brain hurts from speculating all the angles and tangents. And maybe we'll never know the true story behind the smoke screen, but this much I DO KNOW...the stories told in public don't add up, and there is definitely something rotten in the state of Denmark! If only we knew from where the stench was emanating... Edited April 5, 2014 by Hedon James Quote Link to comment Share on other sites More sharing options...
sunrat Posted April 6, 2014 Share Posted April 6, 2014 Without a viable CLA, I imagine patent trolls would view Linux development as the holy grail of panning for patent gold?!!! No. That's what GPL is for. so we'll use Sievers' systemd. Sievers was a developer, not the designer of systemd. 1 Quote Link to comment Share on other sites More sharing options...
securitybreach Posted April 6, 2014 Share Posted April 6, 2014 If Ubuntu wants to use their own license; they need to remove the Linux kernel and develop their own kernel. As far as patent trolls, there is a nice legal team behind the gpl and open source: Free Software Foundation Software Freedom Law Center Some more reading... https://www.software...ance-guide.html http://gpl-violation.../legal-faq.html Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.