Re: MATE desktop , compiz does not work properly after disable gconf support

In reply to: Minoru TANABE : "MATE desktop , compiz does not work properly after disable gconf support"
Go to: [ bottom of page ] [ top of archives ] [ this month ]

From: Tomoaki AOKI <junchoon_at_dec.sakura.ne.jp>
Date: Thu, 16 Jun 2022 01:11:18 UTC

On Wed, 15 Jun 2022 23:50:24 +0000
Minoru TANABE <kotanabe3@gmail.com> wrote:

> Hi all.
> I'm using MATE desktop and compiz. For several years, they work fine.
> But after commit below
> ---
> commit d631449922df1f12590bdd282ea21f2cbf885109
> Author: Baptiste Daroussin <bapt@FreeBSD.org>
> Date:   Wed Jun 8 14:49:53 2022 +0200
> 
>     compiz: disable gconf support
> 
>     Approved by:    maintainer (private email)
> ---
> X display freeze, (System is still working).
> I wonder Why disable gconf, and I want to know how to fix it.
> Thanks.
> -- 
> Minoru TANABE
> 15-1-406, aoto 1-chome, tokyo, japan
> EMail  kotanabe3@gmail.com

It is because only gconf holds the working configuration of compiz.
I've already commented about it at dev-commits-ports-main ML [1] and
the maintainer is working on it.

But the current approach (port again x11-wm/ccsm) requires working
gconf backend, maybe.

If something is NOT already broken, applying attached patch (partially
reverts the commit, excluding MOVED entry) onto latest ports tree and
forcibly rebuild / reinstall

  x11-wm/compiz
  x11-wm/compiz-plugins-main
  x11-wm/compiz-plugins-extra
  x11-wm/compiz-plugins-unsupported

would hopefully help.

If you already deinstalled x11-wm/compizconfig-backend-gconf on update,
you should build / install it again.

If you want to backup your configuration from gconf backend, run

  gconftool-2 --recursive-list /schemas/apps/compiz > somewhere1
  gconftool-2 --recursive-list /apps/compiz > somewhere2

would export them to XML data.

Note that I've locked affected ports before update, I'm on gconf
backend. So if the update actually break something, the above would not
help.

One advice:
  Checking dev-commits-ports-main ML whether there's any
  dangerous commits like this exists or not would bless you.
  I'm checking it via ML archive. (Not subscribed.)
  Or check via cgit [2].

[1]
https://lists.freebsd.org/archives/dev-commits-ports-main/2022-June/026094.html

[2] https://cgit.freebsd.org/ports/log/

-- 
Tomoaki AOKI    <junchoon@dec.sakura.ne.jp>