GitLab

Chain up here instead of doing our funky only chain-up after setting
phase. This should make things more predictable for subclasses that
need to do anything funny, and Vala.

https://bugzilla.gnome.org/show_bug.cgi?id=779891

Commit e874bb7902cc06f9f6d4427d99ec33e3757304e4 broke a bunch of
inheritance in vala (that wasn't very good to begin with, but at least it
worked). In particular, Vala has a very hard time determining if an object
properly meets the interface requirements. It does not handle relaxed
write checks vs the parent very well (and has strange requirements for
vfuncs to be available that are completely unnecessary).

However, we can stop inheriting from Ide.Object and implement the context
property ourselves to work around the issue. Incredibly annoying, but good
enough to fix the build a week before release.

https://bugzilla.gnome.org/show_bug.cgi?id=779891

We don't need to require IdeObject for a bunch of things, we can just rely
on GObject + the context property.

This will let us work around some language breaking checks in Vala.

https://bugzilla.gnome.org/show_bug.cgi?id=779891

This reverts commit edc8381c.

This reverts commit 5c05725f.

This is an import and update of GsProgressButton from GNOME Software. We
can use it as a base for our transfer button so that progresss is shown.

Commit e874bb7902cc06f9f6d4427d99ec33e3757304e4 broke a bunch of
inheritance in vala (that wasn't very good to begin with, but at least it
worked). In particular, Vala has a very hard time determining if an object
properly meets the interface requirements. It does not handle relaxed
write checks vs the parent very well (and has strange requirements for
vfuncs to be available that are completely unnecessary).

However, we can stop inheriting from Ide.Object and implement the context
property ourselves to work around the issue. Incredibly annoying, but good
enough to fix the build a week before release.

We don't need to require IdeObject for a bunch of things, we can just rely
on GObject + the context property.

This will let us work around some language breaking checks in Vala.

This should probably be /app/etc, without this change the build fails due
to trying to write to /etc without permission.

https://bugzilla.gnome.org/show_bug.cgi?id=779885

This dir does not exist.

https://bugzilla.gnome.org/show_bug.cgi?id=779885

We want to keep Nightly builds working, even if that means vala is
disabled. Vala plugin is broken due to bug 779038.

Tasks were not getting completed when they were successful transfers due
to only returning an error and not success.

This will check for flatpak-builder on the host by using:

  which flatpak-builder

as that is more reliable than checking the package manager. We will still
use pkcon to install the package, but we don't need the D-Bus service to
check for it.

Lower is better, and we want autotools to always be the preferred build
system until we have either 1) rearchitected multi-build systems or 2)
decided on another "blessed" build system.

Builder only allows vim like substitute commands of the form :s/foo/bar/
but vim allows :s/foo/ or :s/foo (which just replace with the empty
string). This commit makes those formats work by changing how the search
text and replace text are calculated.

https://bugzilla.gnome.org/show_bug.cgi?id=746028

After we have installed the given package, ensure that the build pipeline
will be re-setup.

Instead we'll rely on the user to install the package using the message
in the workbench. That provides a more clear flow of what they are doing
so they are not surprised by a pkexec dialog.

We need to do one at a time, so we can default to flatpak-builder which
will require flatpak too.

This checks for flatpak-builder when the flatpak plugin is loaded. We
might want to delay this until the current build configuration is changed
for those that are not using flatpak.

This is a new widget that can be pushed onto a workbench so that toplevel
messages are visible to the user. You might use this to show a warning
that a package needs to be installed, such as flatpak-builder.

This is a new transfer that can perform installations of packages using
the pkcon command on the host. Such is useful for installing packages
like "flatpak-builder".

If we need to install flatpak-builder on the host, we can do so with
pkcon instead of the dbus service which doesn't seem to do what I thought
it did.

We should also be able to make an IdeTransfer around this for progress,
but that is a future exercise.

This will do some cleanup work like strip and splitting debug symbols.
That results in the .flatpak files being much smaller (3mb vs 13mb for
the C++ template).

Krzesimir Nowak authored Mar 09, 2017 and Christian Hergert committed Mar 09, 2017

It was aborting gnome-shell, when downloading the flatpak failed.

https://bugzilla.gnome.org/show_bug.cgi?id=779802

Krzesimir Nowak authored Mar 07, 2017 and Christian Hergert committed Mar 09, 2017

We should return the error immediately if downloading the archive
failed. Also, we shouldn't try to open a flatpak project if download
failed.

https://bugzilla.gnome.org/show_bug.cgi?id=779802

The first time we show the panel, it is fine to raise it (without
focusing).

This does a bit of cleanup and bumps some of the package versions to be
release-candidate versions rather than pre-release tarballs.