On Sunday, 2013-03-24, Kevin Chadwick wrote: > On Sun, 24 Mar 2013 14:16:34 +0000 (UTC) > > Duncan <1i5t5.duncan@xxxxxxx> wrote: > > Of course they'll with a near-certainty still continue to work on > > xorg for a few years anyway, but once the general desktop moves to > > wayland, the X dependency gets moved to the "might not be installed > > for anything else" list, and suddenly the additional deps cost of > > running that old app go ***WAY*** up, which means fewer people run > > them, which means a bigger likelihood of an accumulation of serious > > bugs over time. > > > > That's the wayland upset of the status quo I'm a bit worried about. > > OpenBSD is getting KMS drivers just now (except GPL'd drm drivers) so > wayland should be possible for OpenBSD atleast on a fair chunk of > hardware, as it is apparently very compatible with Linux, That is most likely only necessary if one uses the reference implementation. It would be weird if a protocol on user space level could somehow depend on certain kernel features. > especially as I am skeptical of it being implemented well considering > freedekstop.orgs history such as disregard for other opinions and > anything non linux Hmm, can you point out a freedesktop.org spec which is bound to Linux? Qt and KDE implement most of them can run quite nicely on BSDs as far as I heard. > Don't get me wrong I welcome the benefits of wayland which has been in > development for so long and so hope that will mean it will be good for > the whole eco system but I worry about recent stories of Gnome and KDE > rushing towards it due to Mir. That seems illogical. Indeed. But that illogic is easily enought explained. Somebody misinterpreted developer communication. Happens all the time, one of the drawbacks of open development. Work has been going on for a long time already, on the Qt side of things (no idea about the GTK+ side) the respective QPA backend has developed in parallel with wayland's efforts for years, inspite of that requirng a lot of work in the end (having to update to protocol and API that were in flux). However that was deemed necessary so one could early one see if things needed to be different to fullfil the requirements of the toolkit side. However, some people seem to have got the impression that no such work was being done, so the discussions around that topic were emphasised more. Which, as we have seen, led unfortunately to the misinterpretation that work had only recently started. If you find anyone who got tricked by that misinterpretation, most likely because they didn't have access to respective developer communication themselves, you can easily show them a video: http://www.youtube.com/watch?v=jLiSEmtRvGs It is from a developer on the Qt side of things, showing things that can be done with a Qt based wayland compositor. In November 2012, obviously way earlier than the recent Mir announcements and development for that compositor library did almost certainly start many months prior to the presentation. Cheers, Kevin -- Kevin Krammer, KDE developer, xdg-utils developer KDE user support, developer mentoring
Attachment:
signature.asc
Description: This is a digitally signed message part.
___________________________________________________ This message is from the kde mailing list. Account management: https://mail.kde.org/mailman/listinfo/kde. Archives: http://lists.kde.org/. More info: http://www.kde.org/faq.html.
