UDS sponsoring in December 2008

Area linked to the brainstorm ideas and Ubuntu ML discussion

Desktop Experience

I heard a lot of parents asking for a Parental Control Tool for protecting their child from what they can find over the Internet without having to look after them all the time. The backend tools (with CLI) exists for years now and are reliable, but let's confess that, as of today, there is not really easy-to-use and human GUI tool for that. http://brainstorm.ubuntu.com/idea/1244/ and all its duplicate show cleary that need.

Some people thinks that this is some kind of DRM. I completely disagree with that vision as this tool is on the power hands of the administrator (which will be typically the parents in this case) and not in a third-incontrable hands.

That's the reason why I started a project, with mockups. I am currently finishing the specifications. GChildCare tends to be a simple, user-friendly et secure parental children protection software for ubuntu (first, and then, back to other distribution). Actually, there is no easy frontend for user to enable/disable and configure parental control.

This project aims at fixing this using already existing components but providing an easy-to-use GUI so that parents may not fear to use Ubuntu for their children because of Internet site control. I am really convinced that providing easy to use stuff "for human being", is really in the Ubuntu spirit.

The first deadline release of a usable version (0.1) is for Jaunty (9.04). All the work will be focused on the GUI, following HIG, as all the needed tools are available by command lines and configuration files.

This GUI will have two sides:

• one frontend (GUI, etc.) which will try to be as neutral as possible (ie GTK-based in a first attempt, and then, try to configure to make it also compatible with QT, switching between the libraries or providing 2 binary versions)
• one backend (connection to several solutions, one default will be automatically chosen).

I have made a lot of research on similar projects and found the 4 links mainly dealing with this:

• http://ubuntuforums.org/showthread.php?t=843510

• http://linux.softpedia.com/get/Security/Parental-control-24501.shtml

• https://launchpad.net/webstrict

• https://launchpad.net/willowng

So, why another project? The main idea here is to follow strictly GNOME guidelines and HIG for the GTK (which will be the first step) version. That is to say: - Use GTK2 (1 & 3 doesn't) - KISS (2 is totally out): 3 buttons should be a maximum for the frontend (the first window, of course). I think the average user needn't to specify himself a word list, he wants to use whitelist/blacklist (list automatically found over the Internet) or access/deny the Internet (for each user). - it should be only a GUI, not a proxy nor an dansguardan equivalent. So 4th would not fit here. I will of course, take some ideas from those projects.

For the QT version, we will see that once the first (but keeping in mind to make it as flexible as possible so that most of the code can be in common and reused)

I think that working on this area with people (using backend likes dansguardian + squid, do not reinvent the whell) will be a great help to see and reckon what people are really waiting for.

Drafts and blueprints are available in the project pages at https://launchpad.net/gchildcare

Blueprints summary on the wiki page, available through LP more detailed blueprints:

• https://wiki.ubuntu.com/gchildcare/Specprofile

• https://wiki.ubuntu.com/gchildcare/Specfilteringaccess

• https://wiki.ubuntu.com/gchildcare/Specrightaccess

• https://wiki.ubuntu.com/gchildcare/Spectimemanagement

• https://wiki.ubuntu.com/gchildcare/Specshareconfiguration

Of course, this software will be made using the rocking python language!

Miscellaneous linked to UDS involvement

Here are the areas where I want to get involved into for next Ubuntu release (jaunty), that are other contributions & interested topic on which I want to exchange and focus on :

• I am totally found of VCS softwares and especially with Bazaar.

Being convinced that this is the right path to choose for packaging, I am really interested to help out for jaunty in this particular area to make the work for MOTU and core-dev easier, more reliable, effectiveness and with enhanced traceability. James Westby's (james_w) current work video confort myself in this feeling. So, if you need me to jump into this task, it will be with a great pleasure.

• I worked with Jamie Strandboge (jdstrand) in ufw, completing integration with popular packages (apache, nis...), backporting the new version to hardy and making it case nonsensitive (python really rocks, have I already told that ? ;)). Going further with ufw will be great (completion and stuff like that). I can give an hand for that also.
• Also, I removed the stop linked to init.d for a buch of package to make stop and restart quicker (and it has been accepted by debian !), working on this with James Westby. I think that the currently main focus on jaunty to speed booting sequence is a great direction to go with and I will be enthusiasmed to work on this as I followed from the very beginning the progress of upstart and all related stuff to it.