[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

d-e-c git changes



Hi Wolfgang,

kudos and thanks for your latest commits in src:debian-edu and
src:debian-edu-config!

Some questions:

commit c08ad7e42c9c1e6b38c92486e9be650da5eef2f0
Author: Wolfgang Schweer <wschweer@arcor.de>
Date:   Mon Apr 30 21:28:06 2018 +0200

    cf3/edu.cf: Add class desktopintern.
    
    This class matches a stationary workstation with package education-desktop-other
    installed.

+    # Set for networked workstation if the 'education-desktop-other' package is installed.
+    "desktopintern"  and => { fileexists("/usr/share/doc/education-desktop-other"), fileexists("/usr/bin/ntpq") };


why use the existence of /usr/bin/ntpq ?


commit d5a6ec5c48dae5845e200db69f48467b5a9ec584
Author: Wolfgang Schweer <wschweer@arcor.de>
Date:   Mon Apr 30 21:37:18 2018 +0200

    cf3/cf.finalize: Add shell command for class desktopintern.
    
    This allows one to install some packages suited only for networked machines in
    addition to those installed via the education-desktop-other meta-package for all
    workstations.

+
+  # Install desktop related packages, suited only for networked machines.
+
+  debian.desktopintern.installation::
+
+    "/usr/bin/apt install -y jxplorer gosa-desktop krb5-auth-dialog"
+      contain => in_shell;


hm, the commit message explains a bit, but why are those packages choosen and anyway,
this feels wrong, like an attempt to rewrite configuration management systems. 
(which arguently this is not, but rather a configuration of cfengine, which indeed is a
configuration management system).

Now I read debian/changelog I understand the intend much better. I think this should
be made cleaer in cf.finalize (as comment in that file) as well as better packages should
be choosen. (Or explained better as well :)


-- 
cheers,
	Holger

Attachment: signature.asc
Description: PGP signature


Reply to: