Skip to main content

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: [eclipse-pmc] Eclipse deliverable - food for thoughts

I'm still a bit skeptical on how this will work in practice and wether it won't be too restrictive for daily usage - and thus won't gain enough adoption to actually allow SWT to reduce its list of 13 different Gtk versions. (Btw. is it really 13 ?)

But if it can be made to work then definitely something to at least use as the *actual tested and supported* mechanism.

/max


----- Original Message -----
From: "Max Rydahl Andersen" <manderse@xxxxxxxxxx>
To: eclipse-pmc@xxxxxxxxxxx
Sent: Thursday, 18 June, 2015 2:38:08 AM
Subject: Re: [eclipse-pmc] Eclipse deliverable - food for thoughts

On 17 Jun 2015, at 16:48, Aleksandar Kurtakov wrote:

The official entry point is
https://wiki.gnome.org/Projects/SandboxedApps .
But if you want to try it out please read
https://blogs.gnome.org/alexl/2015/06/17/testing-rawhide-apps-using-xdg-app/
. How does shipping/generating xdg-app sounds in terms of the Eclipse
IDE?

It sounded interesting until it states the app will have limited host
processes, network and filesystem access.

By default. But in the application metadata access to these can be granted. See https://wiki.gnome.org/Projects/SandboxedApps#Application_metadata .


Won't that make the IDE of very little use if I can't easily use my
existing projects, debug my java vm's nor access the network ?

Using existing projects and network access is supposed to just work assuming correct metadata in eclipse xdg-app. If by debug you mean attacthing to already running external process (not launcher from inside eclipse) this is a bit of a grey area for me still and has to be tried. Such limitations and especially how eclipse plugins can request additional permissions as some of them would need is something we will never get the answer of if we don't try.


And beyond that what does this feature provide Eclipse ? will it be
easier to install ? run on more distributions without having different
SWT bugs dependent on the host OS or ?

Exactly, this should give eclipse the possibility to say e.g "runtime=org.gnome.Platform/x86_64/3.16" and have a single set of GTK/Glib/Webkit/... versions to test with no matter what the distribution is. And when projects like Gnome start shipping/creating the runtime themself there runtime should be identical (thus the bugs too) on Fedora/Suse/Ubuntu/Mint..... This is the kind of problem I'm looking for a solution for as currently there is one set of problems on latest Fedora, different on latest-1 Fedora, entirely different on latest Ubuntu and so on. With runtime definition like org.gnome.Platform/$ARCH/$VERSION Eclipse will have the freedom to jump to newer GTK version when ready not urging like now because users start getting new version on their machines. This should also free some resources as currently SWT tries to support 13 different Gtk versions with workarounds for many of them which is not feasible solution and this creates even more bugs when old workarounds s!
tart to i
nterfere with newer versions workflow.

Alexander Kurtakov
Red Hat Eclipse team


/max

Let's discuss either here or on the call for anything that comes to
your mind.

Alexander Kurtakov
Red Hat Eclipse team

_______________________________________________
eclipse-pmc mailing list
eclipse-pmc@xxxxxxxxxxx
To change your delivery options, retrieve your password, or
unsubscribe from this list, visit
https://dev.eclipse.org/mailman/listinfo/eclipse-pmc


/max
http://about.me/maxandersen
_______________________________________________
eclipse-pmc mailing list
eclipse-pmc@xxxxxxxxxxx
To change your delivery options, retrieve your password, or unsubscribe from
this list, visit
https://dev.eclipse.org/mailman/listinfo/eclipse-pmc

_______________________________________________
eclipse-pmc mailing list
eclipse-pmc@xxxxxxxxxxx
To change your delivery options, retrieve your password, or unsubscribe from this list, visit
https://dev.eclipse.org/mailman/listinfo/eclipse-pmc


/max
http://about.me/maxandersen


Back to the top