Congratulations

First off, congratulations to the Launchpad.net team for reaching bug #1000000. They’ve managed to build a huge platform that scales very well. Very few bug trackers live to that milestone and it’s amazing how they have managed to keep it snappy and also keep downtime so low by doing continuous roll-out.

1 000 000 x 67

A million bugs are a lot, but even more mind-blowing: for every bug filed in Launchpad.net, 67 iPads have been sold. Educational institutions everywhere are jumping on the iPad bandwagon, and in the Edubuntu project, we believe that the tools are quickly coming together that allows us to deliver a product that can be truly competitive with the iPad in educational environments.

We’re currently re-designing the Edubuntu website and will soon have a dedicated section to this project, but in the meantime, please join us on the edubuntu-devel mailing list and introduce yourself, or on the #edubuntu IRC channel on Freenode.

Edubuntu 12.10 Plans

Today at the Ubuntu Developer Summit we had a session to plan out the next release of Edubuntu.

For the Edubuntu 12.10 core product, we’re doing some typical Edubuntu updates and features, which include:

• Authentication step in the installer for AD/Samba4/LDAP
• We’ll be reviewing the installed apps, add gnote, refresh the kde-edu apps selection
• Dynamic installer slideshow, based on options selected
• Juju charms for educational web apps (Moodle, WordPress Multisite, etc)
• Remote Live Installer (booting an Edubuntu/Ubuntu livecd over the network
• Education-specific software highlights in Software Center
• Speed-up installation by optimising the way we ship language packs
• A variety of Desktop/Artwork tweaks and fixes

Edubuntu Labs: Get Excited and Make Things

On Tuesday I had 2.5 minutes to speak about Edubuntu during a plenary session where I presented some of our more ambitious plans in Edubuntu. We want to make it easier for people to work on their ideas and projects that might be good for Edubuntu, but that doesn’t necessarilly fit into our main product yet or in a 6 month release cycle. For that, we’re starting Edubuntu Labs (subject to namechange). A playground for experimental and exciting features that might one day make it as a supported Edubuntu product. Internally, we’re starting two of these projects to kick it off.

1. Edubuntu Server

Edubuntu Server is a product we discontinued a few years back. Due to popular demand, we’re considering reviving it as a product. Aspects we’re currently investigating:

• Zentyal Small Business Server
• A built-in disk-cloning tool using LTSP
• A remote installer for Ubuntu based installer media
• Schooltool
• Schooltool integration into Zentyal
• Samba4

2. Edubuntu Tablet

All of this is still early work, but I wanted to get it out there as early as possible. Over the next 2 weeks there’ll be more official announcements on the Edubuntu website. We’re looking for more contributors to help us out with this, please join us on #edubuntu and add it to your autojoin and introduce yourself on the edubuntu-devel mailing list

Greentown School Kindergarden iPads

Earlier this year, I was a bit surprised to see that a school is planning to buy iPads for its entire Kindergarden. I’m interested to know how that went and how they’re using those devices. The school dipped into it’s long-term savings to buy those iPads, and to me that sounds risky. If I were a parent at that school I don’t think I would approve of the school risking long-term security to provide technology to kindergarden kids. The article I link to states that the iPads are bought because it aims to improve reading skills. I’m pretty confident that it would make at least some difference, but I’m also confident that it doesn’t justify the price and risk associated with it. Also, how does the school manage the apps installed? Are there tools for that (I’m honestly asking)? Can educators monitor scores? The child’s progress? Do the educators receive sufficient training on these tools? Who helps when things go wrong? The cost of iPads doesn’t end with just the price of the devices and the direct maintenance costs. Educators have to change the way they teach. They have to learn how to provide education via a new medium and redevelop some of their materials. When I first read that article I was thinking of all these things and wondering if that school could really justify buying them.

$18.2m Fort Bend Independent School District iPads

Today I stumbled across a website for students who are against their school board spending $18.2m for an iPad roll-out. The school board is rolling out iPads while their traditional computing IT infrastructure is aging, reportedly the school still has some outdated technology in place, such as Windows 98 machines. The students are making the case that money would be better spent upgrading their current infrastructure, like upgrading all their machines to Windows 7 and deprecating all the old hardware that can’t run it. I agree with them, it would indeed be an improvement on buying a bunch if iPads, and that investment is likely to also last a lot longer than the iPads will.

While $18.2m is a lot of money, I think it’s the tip of the iceberg of what’s being spent on iPads in schools and I can’t help wondering if it’s the best way for schools to be spending their money, especially considering the doubt of whether the use of technology actually increases grades. In my experience educators tend to think of computers as magical devices that will do a lot of the teaching for them and give them more time to focus on other things, but the reality is that it’s rarely the case.

The Case Against Technology in Schools?

I suppose it’s starting to sound like I’m trying to build a case against technology in education. I’m not.  However I do believe that spending that kind of money, especially for very young kids is a waste. They have much more important things to do, like playing. I can’t imagine how learning to read and do basic algebra can possibly be better for kids on iPads than interacting with your real life teacher and your classmates around you. These kids have several years ahead of them where they can learn how to be office drones and good little consumers, why put that on them now?  I’ve come across teachers and parents who believe that their kids need to start as early as possible in order to give them a competitive edge and to get them “in early”. I can’t predict the future, but I’m pretty confident that your kids will be completely fine if they learn about computers a bit later.

So why am I bringing all of this up?

Well, I’m not really trying to convince anyone of anything in particular, but I do think people should be aware of some of the current trends that are at least somewhat concerning. I also think that people should think about this, because I remember being in school and getting angry myself when I came across things in the system that just completely sucked and wondering… “isn’t there anyone out there who really cares about this!?”.

Edubuntu

I’m also thinking of solutions, because I can’t help doing that. I’m wondering of ways we could make Edubuntu a *lot* more better for schools. So far we’ve been pushing out a collection of the best available free software we could find every release and as a project it’s getting quite stable and we’re getting quite good at doing what we do. I also think it’s time we step it up, have an official, supported Edubuntu device. Have good management software where you can choose which software, ebooks, courseware and other tools are installed on which device. Have testing tools and feedback tools that are easy and secure to use. Make better use of the Ubuntu Software Center’s support for additional add-on applications and build an eco-system for people who develop software for education, free software and not.  We’re a small project and just keeping our current work going is already just about as much as we can carry, but we’re going to have to grow the project drastically if we’re going to develop something that is an appealing and interesting alternative to these iPads that people are so happily sinking money in to, often money that they don’t really have. I also want to see a good alternative that provides a solution using free software.  I have some more details on all of this to follow, but I thought I’d keep this blog entry shorter so that I can at least get it out.

I’m going to be presenting some cases for big changes in Edubuntu to the Edubuntu Council soon and I’m going to try to get us a lot of help for this. If you’d like to get involved in any way, please get in touch! More details will follow soon.

A Gnome summit just 2 hours away from where I live? Awesome!

A few weeks ago I noticed that the Gnome summit that usually takes place in Boston is happening in Montréal this year!  Being a Gnome user for more than a decade and having provided support for it to many people since then, I thought it would be great to pop in for a few hours and see what it’s about. The first session is just about to start and I hope to be back tomorrow as well.

Gnome Stuff in Edubuntu

In Edubuntu 11.10 (Oneiric Ocelot) that’s being released next week, we’ve had to drop Nanny, Pessulus and Sabayon. That’s quite painful since they were great tools, but they didn’t work so well with all the Gnome 3 stuff and they don’t seem to be very actively maintained upstream at the moment. I would’ve gotten involved with those tools if I wasn’t already over-committed, I even considered doing something from scratch that integrates with the Gnome System Settings manager. My goal for now is just to get to know some people and learn more about the Gnome project and its goals. Maybe in a few months my situation will be different and I can commit some time to it. There’s really a need for administrating what desktop systems look like, how they work and how they are locked down in schools and other large deployments.

In Edubuntu, we’ve had to make some tough choices regarding the Gnome desktop since the 11.04 release. We want to keep Edubuntu a great system for schools, but also not stray too far from what a default Ubuntu system gives you. For 11.04 (natty) we decided not to use Unity by default, but provide it as an option in the installer so that early adopters could give it a try. For 11.10, we now install Unity by default and provide the Gnome fallback session as an optional installation. Our aim for Edubuntu is to support Gnome fallback mode as well as Unity. I’m really glad to see that Debian has decided to support Gnome fallback mode as an equal to Gnome shell in Wheezy without having to activate it in system settings first. It would be nice if Gnome upstream could stop calling it the ‘fallback’, but I guess they want to push Gnome Shell as the default as far as possible. Happy hacking everyone!

Edubuntu at UDS

Today we had a session at the Ubuntu Developer Summit at Budapest covering the work in Edubuntu for the next release cycle. Not all of the items are assigned to someone yet (especially with the documentation), so if you’d like to get involved, please give us a ping on IRC or mailing list.

Fixes, improvements and some low hanging fruit

Translations:

• Make LTSP Live translatable
• Add ability  to set LTSP language from installer
• Add Ubiquity page to add additional languages at install time
• Edubuntu installer options itself should be translatable, main inclusion will probably be the easiest fix
• Fix a translation bug in LDM
• Investigate translation of Edubuntu.org, and whether we can migrate to Drupal 7 first before that to prevent future migration headaches

Installer /Upgrades specific:

• Offer to create a new LTSP chroot on upgrades
• Use the right wallpaper with the install-only environments (currently uses the Ubuntu default)
• Add wubi support to DVD

Documentation:

• Troubleshooting guide for common problems and questions found in schools and similar environments
  • Why can’t I access certain websites (who to install non-free flash, sun-java, etc)
• Howtos:
  • How to configure smart boards and similar devices
  • How to install and maintain Moodle (yes, we’re giving up on packaging web stuff)
  • How to install and maintain Koha
  • How to install and maintain open-school.org
  • How to install common wine apps used in schools that work, list those that we know don’t and link to wineappdb

Ongoing work through this cycle:

• Monitor the Gnome 3 transition, look at what’s broken in tools such as Pessulus, Sabayon and Nanny and get involved where possible
• Improve Edubuntu Live Welcome slides for the Live USB and WebLive environments, make it more generic so that other derivatives could use the underlying scripts as well

Big Hairy Audacious Goals

If our goal is to bring the best of educational free software available together in one easy to install system then I think this release will get us there, but where to next? We want to grow the Edubuntu community but we can’t do that unless we add some new and exciting things to our to do list. One thing that has come up over and over again and that I bought up in the UDS session but we got no answer for is in which direction we should start going next. Should Edubuntu focus more on making system administration and management of computer labs in schools easier and simpler? Or should we focus on education and pedagogy and make the system better geared towards being a great teaching tool? We could even alternate between a focus on each of those for each release cycle, but it would be nice to get some big dream ideal world goals down that we could chase in Edubuntu. I think that we’re past the point where we should ‘take it safe’ when planning new Edubuntu  features, I think we’re in the position now to take some big decisions and grow our community to make it happen.

Any thoughts or ideas?

Some people seem to just love talking about history. I prefer making it! Here’s something a bit more light to distract you from all the history lessons currently on Planet Ubuntu…

Current Edubuntu Wallpaper

The current Edubuntu wallpaper has served us well now for a few releases. It’s simple and powerful and works across a wide age group.

In Edubuntu 10.10, we had a bunch of Hitchhiker Guide to the Galaxy easter eggs to coincide with the 10.10.10 release, which I partially blogged about before. A small change was introduced to the wallpaper in this release; if you look really closely in the top right hand corner on the Edubuntu login screen, you’ll notice that it has the words “don’t panic” written on it. This was my favourite easter egg in the maverick release.

New Default Edubuntu Wallpaper

Martin Owens has done a great job of putting together a new wallpaper for us. When I first uploaded it to the Natty archives I thought that we’ll probably get some people who say that it’s too “cartoonish” (which we did in the meantime), but if Debian can get away with having the Space Fun theme for a universal operating system then the ‘cartoon’ critics can at least give a bit of leeway for Edubuntu here.

It was originally called “Perspective” when the first concept was put together. It’s a little bug on a little leaf with a view of a valley and a whole galaxy behind it. Even though it is somewhat cartoonish, the perspective theme is probably something that the older crowd will understand a bit better. I’ve tried it on Edubuntu for the last week and it works really well. It might still go through some tweaking so the final version may be a bit different than the picture above. Thanks to Martin for all his efforts so far!

What happened to the artwork competition?

Most of the submissions we received didn’t quite align with the specification we set out (some of them included text (hard to translate, doesn’t like scaling), some pictures of books or world maps (clichés)- things that we specifically wanted to avoid) and we were approached by Martin who said that he’d be willing to do a wallpaper and work it through various iterations. We notified the artwork team about this, there were some mixed feelings on both sides, especially considering that the submissions were still really quite good despite not being quite suitable as defaults. So what we’ll be doing instead is include the wallpapers that have been submitted to the competition as alternative wallpapers in the system. There aren’t many of them so at least they won’t take up too much space. I’ll do another blog entry especially for them when they hit the archives.

Simple Greeter for Live Systems

During the last development cycle we launched Edubuntu WebLive, which uses a Drupal module to create remote users on an application server and connect the user via NX. We thought of popping up a web page when the user logs in, explaining what it is that they have logged in to and how to get around.

We then moved to make it a bit more simpler with a Webkit/PythonGTK interface that displays the slides on login. It loads a bit faster than a full browser and there are less buttons and things that could potentially get in the way.

We need some nice slides

The messages we want to convey, as seen in the screenshots above are quite simple, but as you can see in these screenshots they are quite crude and definitely early work. I seem to be running low on creative juices on this and have asked for feedback on the edubuntu-devel mailing list but haven’t gotten much response yet, so I thought I’d post this to Planet Ubuntu and try to get some wider feedback. If you have some ideas or would like to propose some slides (it doesn’t have to be anything like mine), feel free to do so. Natty’s artwork hasn’t been finalised yet, which also makes it a bit harder, but it would be nice to get it right at least in concept for now. Feel free to comment here or through the usual Edubuntu contact points, or submit a patch!

Every now and again, educators ask me where they can get more fonts for Edubuntu. We include great desktop publishing software (scribus-ng, inkscape, gimp, etc) in Edubuntu, but our default font selection is rather dry and uninspiring. A few weeks back I looked whether there are some nice fonts in the Ubuntu archive that we could include. I figured that even if there’s one or two good ones available that we could ship, then it would at least be some improvement. The results were quite surprising, there are a wealth of fonts available in the archives.

I added many of them that seemed useful at face value to the edubuntu-fonts meta-package now available in PPA and soon in Natty. It installs quite a lot of font packages currently. The idea is to cut it down a bit and probably split it into 2 to 4 meta-packages, possibly in the categories I listed below. I’ll include some excerpts from package descriptions, and a few examples too. I can’t possibly list them all, it would make this post way too long.

Educational

Font packages currently included: ttf-essays1743, ttf-junicode, ttf-levien-typoscript, ttf-linex, ttf-marvosym, ttf-oflb-asana-math, ttf-oflb-euterpe, ttf-sil-andika, ttf-ancient-fonts, ttf-inconsolata, otf-stix

ttf-sil-andika

Upstream Homepage: http://scripts.sil.org/Andika

Andika (“Write!” in Swahili) is a sans serif, Unicode-compliant font designed especially for literacy use, taking into account the needs of  beginning readers. The focus is on clear, easy-to-perceive letterforms that  will not be easily confused with one another. A sans serif font is preferred by some literacy personnel for teaching  people to read. Its forms are simpler and less cluttered than some serif fonts can be. For years, literacy workers have had to make do with fonts that were available but not really suitable for beginning readers and writers. In some cases, literacy specialists have had to tediously cobble together letters from a variety of fonts in order to get the all of characters they need for their particular language project, resulting in confusing and unattractive publications. Andika addresses those issues.

ttf-linex

Upstream Homepage: http://gata.linex.org/trac/browser/ttf-linex/

A collection of fonts including hand-writing simulation typographies, ancient Greek and Roman typographies, institutional fonts from the Extremadura regional government and other elegant fonts.

otf-stix

Upstream Homepage: http://www.stixfonts.org

Substitutes for popular Non-Free fonts

Font packages currently included: ttf-liberation, ttf-century-catalogue, ttf-mgopen, ttf-beteckna, ttf-droid, ttf-ecolier-court, ttf-ecolier-lignes-court, ttf-bpg-georgian-fonts, ttf-adf-verana, ttf-goudybookletter, ttf-levien-museum, ttf-linux-libertine, ttf-adf-universalis, ttf-adf-tribun, ttf-adf-switzera, ttf-adf-romande, ttf-adf-oldania, ttf-adf-libris, ttf-adf-irianis, ttf-adf-ikarius, ttf-adf-gillius, ttf-adf-berenis, ttf-adf-baskervald, ttf-adf-accanthis, otf-freefont, ttf-symbol-replacement

ttf-liberation

Upstream Homepage: https://fedorahosted.org/liberation-fonts/

This is one of the most well-known sets of substitution fonts. It’s sponsored by Red Hat and includes a set of fonts that are metrically similar to the Times, Arial and Courier fonts. It’s great for document compatibility and can act as a drop-in replacement without requiring the installation of Microsoft fonts.

ttf-symbol-replacement

This is a replacement for the Symbol font as commonly found on Windows systems. It’s from the Wine project and should work as a drop-in replacement.

Desktop Publishing

ttf-engadget, ttf-okolaks, ttf-opendin, ttf-radisnoir, ttf-rufscript, ttf-sil-gentium, ttf-tomsontalk, ttf-atarismall, ttf-breip, ttf-staypuft, ttf-aenigma, ttf-fifthhorseman-dkg-handwriting, ttf-isabella, ttf-sjfonts, ttf-georgewilliams, ttf-femkeklaver, ttf-adf-mekanus, ttf-dustin

These are all font packages that might be useful for desktop publishing in a school or educational environment. The ttf-aenigma font package alone includes more than 450 thematic fonts that could be used for posters, brochures, etc!

Enhanced Usability or Accessibility

Font packages currently included: ttf-tiresias

ttf-tiresias

Upstream Homepage: http://www.tiresias.org/fonts/

Oh, is that all?

Nope, that’s the beginning. Once we have a good selection of fonts in Edubuntu based on what’s in the archive, we should also extend and find more good fonts to include in the Ubuntu archives.Getting the fonts from the Google Font Directory packaged would be a good next step. If you know of any other sources that we should look into please comment here or on one of the usual Edubuntu communication channels.

What are LTSP Fat Clients anyway‽

Thin clients are a great way to lower initial deployment as well as running costs in classrooms, libraries and similar organisations, but they have some limitations, especially when it comes to graphical and CPU intensive software. Even playing a video on one or more thin clients can often be enough to drag the network performance to crawling speeds. In recent LTSP versions, it became possible to run some applications locally, making it possible to use the local CPU/GPU/memory and drastically improve performance all-round. Since LTSP 5.2.1 (that ships with Ubuntu 10.04), it’s been possible to also run everything locally. This essentially makes a terminal a complete fat client that simply uses the network as a storage device. It combines many of the benefits of thin clients and fat clients, while also requiring a less powerful server since it basically becomes just a file server.

Installation is easy

Step 1: Install the ltsp-server-standalone package

Install it from the command line using apt-get install ltsp-server-standalone or by installing it from the Ubuntu Software Centre. If you’d like to configure a DHCP server seperately, then you should use the ltsp-server package instead.

Step 2: Configure your networking

LTSP will work out of the box if you have an interface configured on the 192.168.0.0/24 range. If you would like to use another IP range, you will have to edit /etc/ltsp/dhcpd.conf and adjust it to your needs. Usually LTSP is run from a separate network interface that connects to the rest of the network. If it’s not yet configured, you can add the configuration for a second interface to the /etc/network/interfaces file. Example:

auto eth1

iface eth1 inet static

address 192.168.0.254

netmask 255.255.255.0

Step 3: Build Fat Client Image

An LTSP fat client image can be used for both thin clients and fat clients alike. This is useful if you have a mixture of newer, more powerful machines as well as very old machines that can be used as thin clients only.

ltsp-build-client --fat-client --fat-client-desktop edubuntu-desktop --arch i386 --skipimage

Edubuntu-desktop can also be replaced with kubuntu-desktop, ubuntu-desktop  or xubuntu-desktop. The Edubuntu one is probably most tested though. It’s recommended that you use the i386 architecture for the thin client environment (you can do this even if your server is amd64), the reason for this is that many machines (such as Intel Atom based thin clients) are not 64bit capable. The skipimage switch will prevent the image from being built now, since we want to install other software on it first. You can get all ltsp-build-client options by entering “ltsp-build-client –extra-help”.

chroot /opt/ltsp/i386 apt-get install htop openarena stellarium vlc

The example above will change-root to the ltsp environment and install htop, openarena, stellarium and vlc. These are just examples, you can choose any other software that you intend to run on your LTSP fat client. Next, sync the ssh keys (used for logins) and update the LTSP image:

ltsp-update-sshkeys

ltsp-update-image

Step 4: Restart Services

Restart networking:

/etc/init.d/networking restart

Restart the DHCP server:

/etc/init.d/dhcp3-server restart

That’s it, your LTSP Fat client server is now installed and ready for use. All you need to do now is connect one or more machines to the configured network and set them to boot from PXE or Etherboot.

Frequently Asked Questions

1. Does this take up a lot of RAM on the workstations? Does it take long for the initial image to download?

The diskless workstation will use the LTSP image in a similar way as a local hard disk and will only read the information over the network as required. It doesn’t download the whole image at boot time, so it doesn’t slow down the boot process by installing more software and it doesn’t use any additional RAM either.

2. How do I set a machine to be a thin client instead of a fat client?

Create a file called /var/lib/tftpboot/ltsp/i386/lts.conf if it doesn’t exist already. lts.conf is the configuration file for LTSP. The [default] section contains settings that apply to all machines on the network. You can specify exceptions to this by specifying the mac address between brackets:

[default]
    LDM_DIRECTX=true

[00:A1:08:EB:43:27]
    LTSP_FATCLIENT=false

LDM_DIREXTX disables SSH encryption of thin client sessions. It is insecure, but boosts performance and might be required for very old machines or of you have limited CPU power on the server. The machine with the MAC address 00:A1:08:EB:43:27 will function as a thin client instead of a fat client. For more settings available in lts.conf, refer to the manual page.

3. Edubuntu integrates with LTSP, why not also DRBL?

DRBL is an alternative to LTSP that does diskless fat clients only. LTSP already matches all the functionality of DRBL, and on top of that supports multiple architectures, mixed thin/fat environments, individual client configuration as well as being properly packaged in Debian and Ubuntu already. Spending more valuable resources on supporting DRBL in Edubuntu wouldn’t bring any benefit to any of our users, although if someone wants it really badly, we’ll be happy to review any packages that they submit through the usual process.

4. Hardware is getting really cheap, is there still a future for LTSP?

The cool kids these days walk around with dual-core ARM based devices with 2GB of RAM in their pockets. With computing becoming so ubiquitous, does it make sense for schools and other organisations to still make use of LTSP? I’m quite sure that the way we use computers over the next few years is going to change drastically, but for now LTSP still offers a great way to cut down on administration and maintenance costs with very little compromise.

5. What effect will Wayland have on LTSP thin clients?

LTSP uses the X Window System to display applications that run on a remote application server. The X Window System is planned to be replaced by Wayland in the near future. Wayland offers better and simpler means to write new software on top of it, which will allow developers to write great software faster and with less trouble than before. In order to achieve this simplicity, Wayland had to drop the network transparency features that is present in the X Window System. This will certainly bring on some required changes on how LTSP is implemented, but at least there are already some options available that the LTSP team will investigate, and there’s no reason why Wayland should mean an end to LTSP.

Want to know more? Just ask!

• ltsp-discuss – General LTSP related help and support mailing list
• Edubuntu ships with an LTSP Live mode that you can try from the Live CD before installing if you’d like to try it out. The Edubuntu documentation page may also be useful.
• edubuntu-users – Edubuntu users mailing list, feel free to use it for LTSP related issues on Edubuntu

I’ll be standing in for Belinda Lopez for the Ubuntu in Education session for Ubuntu Open Week today. I don’t have anything special prepared, but feel free to join in and ask any Ubuntu in Education or even Edubuntu related questions! It’s at 17:00 UTC, Details are on the Ubuntu Wiki.