Description of problem: From upstream bug: I understand that GTK3 support is underway, but it'd be great to be able to get it as the default on supported systems. Version-Release number of selected component (if applicable): 0.6+ How reproducible: yes Steps to Reproduce: 1. install latest development branch 2. 3. Actual results: Gtk2 will go obsolete. Expected results: Use Gtk3 Additional info:
In the past gtk3/webkit2 support was lacking a number of things. We will of course move to it with the next version as that will be gtk3/webkit2 only. I'm not sure what all you are asking here... the above is the plan, just need upstream to release 0.6.0 and rawhide will get it. ;) I don't really want to drop a snapshot in...
(In reply to Kevin Fenzi from comment #1) > In the past gtk3/webkit2 support was lacking a number of things. Because there's no official 0.6.0 release currently, those issues should be reported to upstream. > We will of course move to it with the next version as that will be > gtk3/webkit2 only. Nice. But it will break with the gtk2 enforcement as a (nice to have?) requirement of the Xfce spin. Since Xfce is delayed further with official(!) Gtk3 support, I am not sure if we could still promote Midori/Gtk3 as a featured part of it. > I'm not sure what all you are asking here... the above is the plan, just > need upstream to release 0.6.0 and rawhide will get it. ;) I don't really > want to drop a snapshot in... Well, it's requested for rawhide only, at the moment. Maybe it's even worth a RFE for Fedora 22? What do you think? It may depend who is releasing faster, upstream or Fedora … :)
(In reply to Raphael Groner from comment #2) > (In reply to Kevin Fenzi from comment #1) > > In the past gtk3/webkit2 support was lacking a number of things. > > Because there's no official 0.6.0 release currently, those issues should be > reported to upstream. Sure. But landing it in rawhide when it's lacking a bunch of things is more likely to cause people to just use some other browser. > > > We will of course move to it with the next version as that will be > > gtk3/webkit2 only. > > Nice. But it will break with the gtk2 enforcement as a (nice to have?) > requirement of the Xfce spin. Since Xfce is delayed further with official(!) > Gtk3 support, I am not sure if we could still promote Midori/Gtk3 as a > featured part of it. I'm not sure where you see that requirement. We have a number of gtk3 items on the Xfce spin right now... this would simply be one more. > > I'm not sure what all you are asking here... the above is the plan, just > > need upstream to release 0.6.0 and rawhide will get it. ;) I don't really > > want to drop a snapshot in... > > Well, it's requested for rawhide only, at the moment. Maybe it's even worth > a RFE for Fedora 22? What do you think? It may depend who is releasing > faster, upstream or Fedora … :) I'm not going to land a snapshot in rawhide, at least without more data on how functional it is. This seems like a fine use for a copr... would you be willing to make one? Or I could do so.
(In reply to Kevin Fenzi from comment #3) … > I'm not going to land a snapshot in rawhide, at least without more data on > how functional it is. > This seems like a fine use for a copr... would you be willing to make one? > Or I could do so. I am already working on some copr builds. It turns out that Fedora20 is too old with the versions for all those dependencies: vala, gtk, webkit. So adjusting the spec file for F21+, please stay tuned.
(In reply to Raphael Groner from comment #4) > I am already working on some copr builds. It turns out that Fedora20 is too > old with the versions for all those dependencies: vala, gtk, webkit. So > adjusting the spec file for F21+, please stay tuned. Great. Please do let me know when you have the copr ready, I'd be happy to help test. I might also make a local rawhide build to play with it some...
Trunk upstream has planned a further release … Midori 0.6.0 "All for one, One for all" https://launchpad.net/midori/+milestone/0.6.0
I can confirm that there are issues with responsiveness on javascript pages. Therefore my suggestion to delay any new builds and continue to monitor upstream activity.
(In reply to Kevin Fenzi from comment #1) > In the past gtk3/webkit2 support was lacking a number of things. Hi, I work on WebKitGTK+. We only support GTK+ 3 and WebKit2 nowadays; WebKit1 doesn't get security updates anymore, so I recommend moving to WK2 as fast as possible. WK2 nowadays is more mature and robust than WK1. If there are any particular issues with migration to WK2, I'd appreciate a heads-up.
midori will move to gtk3/webkit2 as soon as it does upstream. ;) They planned to last release, but it was taking longer than they expected. Hopefully the next one will be.
FYI, there have been 126 remote code execution vulnerabilities discovered in the past year since security updates for WebKitGTK+ 2.4 ended. I wouldn't consider upgrading to be low-priority....
Michael, don't confuse priority with severity. But your point is valuable as you mention about a lot of vulnerabilities, I'll remove low severity.
I'm aware of the urgency... but we can't (nor should we) force upstream to do what we want.
In assumption we can not do anything here in downstream, I'll change status back. Feel free to take this bug for any further work with Gtk3 and Webkit2. Although, I don't expect any big progress in near future cause WebKit major development is stalled in favor of new WebEngine project (- at least if we talk about Qt).
I've given up waiting on a upstream release and switched to the webKitTwoOnly branch that builds against webkit2. It's a bit unstable under wayland (unless you run it with 'GDK_BACKEND=x11 midori'), but it does work. Hopefully upstream will release a real version soon.