GNOME OS

There was a very good interview with Jon McCann recently about GNOME 3 and GNOME OS.  Reading public comments on this interview showed a lot of negativity which I think missed the good points.

GNOME OS is unfortunately very loosely defined, but from what I can gather it's essentially controlling the entire stack that GNOME is to make a better experience and make it easier to work on the project.

What I like about this strategy:

• It's focuses on the users that GNOME is targeted at.  We've had a strong direction since GNOME 2 days and the focus on features that these users need through design is the right way of getting there.
• It's dropping old desktop metaphors and moving to new ones.  There are other desktops like XFCE which will continue the Windows 95 desktop metaphor and be successful with it; it's right for GNOME to move on and be more cutting edge.

What I don't like about it:

• It downplays the value of GNOME as a "box of bits".  The drum I'm banging at the moment is about sharing infrastructure.  This is something GNOME has been very successful with in the past and discouraging this cuts off a lot of places where GNOME can get investment from other projects.
• It puts very strong requirements on distributors which they don't want to / can't meet.  GNOME is not like Apple, it can't control the entire stack from hardware to sales.  It needs to work with distributors or have a distribution strategy.  Building a perfect desktop is not enough.

The impression I get is GNOME OS is now effectively the strategy of GNOME and it's generally a good direction.  We need to make sure to flesh it out and ensure that we can have sustained development in GNOME and get wide distribution.

Desktop common ground

There's a common argument that you hear about open source desktops which goes something like "we have less than 1% market share; the other desktops are laughing at us; we should pool together and make a real contender".

And you know what, they're right!  We don't have a significant market share, and we're not at the point where we have a truly amazing desktop experience (but we're getting closer).  Anything we can do to get there faster must be a good thing.

And you know what, they're wrong.  We don't have a finite developer resource.  Open-source is amazing like that - when a project starts that people care about suddenly the community grows.  Trying to mash everyone together into one project wouldn't work and would probably make us even slower.  Putting all our eggs in one basket is a big risk.

How can we share resources to grow that market share without pushing us together into a big compromise?  We need to share infrastructure.  What we need is a POSIX for the 21st century.  We've been slowly building this with things like the Linux, D-Bus, X, GStreamer.  We can do more.

There's been a rise in design thinking in Open-Source which has been really good for everyone.  But I think the pendulum has swung too far.  User experience is not the only factor in deciding what to do.  Infrastructure is expensive.  Every bad API is slowing us down progress on the layers above it.  Every desktop developer that dives into infrastructure is not working on those layers either.

Sharing is hard.  But the cost of not sharing is huge.  Lets make sure the infrastructure we're building for tomorrow works cross-desktop and we can share those costs.

Desktop Summit 2011

Last month I attended the Desktop Summit 2011 in Berlin.  Unfortunately I was only there for the core days because Berlin is an awesome city and the summit is awesome too.

The quality of the talks this year were great, and I only had one or two slots the whole time where there was nothing I wanted to go to.  This summit felt more integrated than the last one and I hope this continues into the future.

Some highlights:

• There was a good response to LightDM.  I felt my talk had a lack of GNOME people present, but I think the GTK4 talk may have absorbed them.
• Lennart did a well researched talk on revamping the login system which sounds very good and left me wanting to know more.
• From what I heard the future of GTK+4 and Clutter looks very promising, but I haven't been able to see the talk as I was doing mine.  Can't wait for the videos to come out so I can find out more.
• Vincent Untz is did a really thoughtful talk on his experiences as a GNOME release manager.
• GNOME Shell seems to be progressing very well and there were a lot of talks on it.
• Plasma/KDE also seems to be doing a lot of innovation.

Some negatives:

• There was basically no mention of Unity.
• There was the usual amount of Canonical bashing and it's not helping anyone.  The GNOME State of the Union had too many cheap jabs and the half hearted laughter shows it's just not funny anymore.
• There was a lack of Canonical people present, and it was commented on numerous times.  I'm personally not surprised, as every year more of my colleagues just don't want to be there.  Andrea Cimitan, who is a great guy, summed it up when he said on Google+ "when I say around here I'm working for "Canonical", people stop smiling :)".
• There was little mention of GNOME OS.  Sometimes we need to be more than just hackers talking about technology and really talk more about planning and strategy.

What I'd like to see at the next summit:

• Increased visibility of other desktops - it still feels very GNOME and KDE centric, I think we can learn a lot from projects like XFCE, LXDE, Elementary, Unity etc.
• Increased collaboration on infrastructure - we need to get freedesktop.org in a better shape so we can pool out resources on the boring stuff and focus more on the user facing component which make us successful.

Broken PDFs in Simple Scan

Since version 2.32 Simple Scan has had a bug where it generates PDF files with invalid cross-reference tables.  The good news is this bug is now fixed, and will work correctly in simple-scan 3.2; thanks to Rafał Mużyło who diagnosed this.  You may not have noticed this bug as a number of PDF readers handle these types of failures and rebuild the table (e.g. Evince).  It was noticed that some versions of Adobe Reader do not handle these failures.

I've added a command line option that can fix existing PDF files that you have generated with Simple Scan.  To use run the following:

simple-scan --fix-pdf ~/Documents/*.pdf

It should be safe to run this on all PDF documents but PLEASE BACKUP FIRST. It will copy the existing document to DocumentName.pdf~ before replacing it with the fixed version so you have those in case anything goes wrong.

If you can't wait for the next simple-scan, you can also run this Python program (i.e. python fixpdf.py broken.pdf > fixed.pdf)

import sys
import re
lines = file (sys.argv[1]).readlines ()
xref_offset = int(lines[-2])
xref_offset = 0
for (n, line) in enumerate (lines):
        # Fix PDF header and binary comment
        if (n == 0 or n == 1) and line.startswith ('%%'):
                xref_offset -= 1
                line = line[1:]
        # Fix xref format
        match = re.match ('(\d\d\d\d\d\d\d\d\d\d) 0000 n\n', line)
        if match != None:
                offset = int (match.groups ()[0])
                line = '%010d 00000 n \n' % (offset + xref_offset)
        # Fix xref offset
        if n == len(lines) - 2:
                line = '%d\n' % (int (line) + xref_offset)
        # Fix EOF marker
        if n == len(lines) - 1 and line.startswith ('%%%%'):
            line = line[2:]
        print line,

LightDM at the Desktop Summit

I'm going to the Desktop Summit and will be doing a talk on LightDM; please come along if you're interested in this project.  It's at the same time as the GTK4 talk unfortunately, but I promise I'll watch that one on video and turn up to mine...

Razing the Bazaar

Currently in GNOME there is some tension as we move into the post 3.0 world about the scope and direction of the project.  I wont go into the details of this, but essentially a number of core developers are pushing for a future in which:

• The scope is widened to include more as part of core GNOME.  This is to allow more control and integration to produce a better user experience.
• The project focus is being narrowed to have tighter requirements.  This is to reduce support overhead and complexity.

This change is being pushed under the "GNOME OS" banner.  While I think these ideas are being pushed for noble reasons (to make GNOME as good as it can be), there are some serious risks I am worried about:

• If we build the perfect OS in GNOME it will not be enough.  History is littered with better products that fail to succeed.  Making an OS successful is as much about the OS design and quality as the ability to deliver that OS to end users.
• If we base all our decision making on "what user visible change does this have?" then we risk losing innovation in our platform.  End-users are only one type of user in an OS and not all changes are relevant to them.  We have to think more in terms of "will this have a bad effect on end-users?" and look at other aspects.
• If we narrow our focus too much we risk losing some of our current community.  The community is an enormous asset of GNOME, and not something we should take for granted.  This is not a company, and is driven by motivated individuals (some of who are then employed by companies).  There is great number of communities out there and GNOME needs to be competitive.
• If we try and control everything then we increase the burden of maintenance onto one project.  There is no funding guaranteed to get us to GNOME 4.  We should always look (within reason) for opportunities to collaborate with other communities.

To abuse the metaphor by Eric S. Raymond it feels like we are razing the bazaar to build the GNOME OS cathedral.  We have a great product in GNOME but to build it faster and better we don't have to clean up our messy edges.  The bazaar around the cathedral is interesting and fun and throws up new ideas.  It's not stopping us from achieving success.

UPDATE:  Changed description of project focus, as it is confusing the point of this post.

And now for some good news

With all the doom and gloom blog posts running around at the moment you may be forgetting all the awesome progress that is being made.  So I just wanted to shout out some things that are happening that I love:

GTK3

GTK+ has been cleaned up and it shows!  GTK is a great toolkit but it had been showing its age.  The tidying up (particularly removing the GDK stuff) has significantly reduced the learning curve.  And more improvements planned for GTK4!

GNOME Shell/Unity

The core user interface is being pulled from the 1990s to the future!  There are real risks and challenges here but it's progress in making GNOME the front-running interface it deserves to be.

GObject Introspection

No more out-of-date language bindings!  With introspection information GNOME developers have huge flexibility in picking languages and all languages are first class citizens.

Vala

A modern language for a modern desktop!  Languages like Java and C# offered a lot of promise, but never seemed to break into GNOME.  A modern language makes us more productive, attracts new experienced developers and gives us an opportunity to escape from the Albatross around our neck (C).

Brainstorm Idea #25877: GNOME System Monitor lacks in-depth information

One of the popular Ubuntu Brainstorm ideas is to improve GNOME System Monitor.  This has been reported to the upstream project, and the project developers agree with the concept.

There is a proposed design: (no attribution as it's not clear who made this)

How can you get involved?

If you have some coding skills, then consider making a patch to fix this!  This is a well-defined feature request, and should be relatively easy to get to work.  Here's how:

1. Comment on the bug that you are interested in working on this.  Ideally the GNOME System Monitor developers will be able to help you, but I am also watching the bug and willing to help out.
2. Get the dependencies required to build the GNOME System Monitor.  On Ubuntu this is as easy as: sudo apt-get build-dep gnome-system-monitor.
3. Check out the upstream: git clone git://git.gnome.org/gnome-system-monitor.
4. Build and test it: ./autogen.sh --prefix=`pwd`/install ; make ; make install ; ./install/bin/gnome-system-monitor.
5. Make some changes and repeat from step 4 until everything works.
6.  When you are done, do a git commit -a; git format-patch origin and attach the patch to the bug report.

Happy coding!

Using the latest SANE drivers in Ubuntu 10.04, 10.10

If you are running Ubuntu 10.04 LTS or 10.10 and your scanner is not supported, then you can try the latest releases of the SANE drivers from my sane-backends PPA.  The following steps from the terminal will enable it.

$ sudo add-apt-repository ppa:robert-ancell/sane-backends
$ sudo apt-get update
$ sudo apt-get upgrade

Please feedback to the SANE project if you continue to have problems!

GNOME3 PPA

In the Ubuntu Desktop team we're currently packaging GNOME 3 components for Ubuntu as per the blueprint decided at the last Ubuntu Developer Summit.

If you're already brave enough to be running Natty, then you can additionally try some new GNOME 3 applications by adding the GNOME3 builds PPA into your sources.  Expect the usual - the packages may not work perfectly, and it's non-trivial to downgrade, so be warned!

We're going to evaluate these packages in the PPA and decide how many are appropriate to include in Natty.

LightDM status update

It must be time to update on how progress is going with LightDM.

• PCMan from LXDE wrote an awesome new greeter that uses GtkBuilder for layout. This allows you to easily theme up new greeters, the default one being the old Industrial theme:
  
  
  
  
  
• There's been some interest in writing a QT based greeter, and I hope we'll be able to show that working soon.
• I've started to write some documentation.
• I non-proposed LightDM for inclusion in GNOME 3 (I don't feel it will be sufficiently ready in time).  This raised the idea of should we switch display managers in the future and got some good feedback.
• We had a UDS session on LightDM which also gave some good feedback.  The outcome is that I plan to make LightDM easily installable in Ubuntu 11.04 as an alternative display manager.

How I Learned to Stop Worrying and Love the Vala

When I first heard of Vala I was not impressed. A domain specific language? That seems like a dead-end; how would we every attract developers to the GNOME platform?

After a while of thinking about it I realised that GNOME already was in this position. GObject+C already is a domain specific language. New developers already have to learn GObject and finding developers who both are proficient and willing to use C can be a struggle.

Recently I have been working on more and more with Vala and porting applications from PyGTK/C to it. Before Vala, PyGTK had a lot of advantaged over C; now the main differences are easy debugging and fast development (Python) vs fast performance and type checking for easier maintenance (Vala).

So, my recommendation is if you have a desktop application that uses GObject APIs* and you are happy with debugging tools like gdb and valgrind then consider using Vala!

* I was working on porting LightDM to Vala but old system APIs were causing difficulty. If you have this case consider wrapping them in C+Gobject first and then interfacing to that.

LightDM UDS Session

If you're interesting in any of the following:

• The Ubuntu desktop login experience


• Display managers


• The LightDM project

Then come along to my session at UDS. It's at 5:10pm EDT and you can join remotely by connecting with IRC to irc.ubuntu.com #ubuntu-uds-bonaire2 and listening in to the audio stream.

Everyone is welcome and I'm particularly interested in gathering requirements for derivative distributions.

GUADEC 2010

I'm at GUADEC all this week so if you have any questions/comments about Ubuntu Desktop, LightDM, Simple Scan or GCalctool feel free to find me!

LightDM

I'd like to announce a side-project I've been working on: The Light Display Manager (LightDM).

Features?

• Fully themeable (easiest with the webkit interface)
• Cross-desktop (greeters can be written in any toolkit)
• Low-complexity codebase
• Standards compliant (PAM, ConsoleKit, etc)

Tell me about themeing...

This is the part I'm most excited about...

You can try the webkit interface online today. My HTML/CSS/Javascript skills are not great so please take that page, copy the lightdm.js include line (simulates the lightdm interface) and write an awesome interface! Please leave a comment here if you make a cool one. Idea - make a greeter that integrates with Facebook.

Got bigger plans? Read the libldmgreeter reference. Idea - make an OpenGL interface that uses Mii avatars.

libldmgreeter not in the right language for you? It supports GObject introspection so you should be able to write a greeter in any introspetion enabled languange (e.g. Python). And the interface to lightdm is D-Bus so you can go more direct if you want.

Wait, what is a display manager again?

The display manager handles the running of the X server (graphical display) on your system. It also provides the login screen (greeter) to authenticate users.

Who is developing this?

This is a personal project developed in my own time. Patches, feedback and help welcome.

What about the other display managers?

LightDM aims to support all use cases of the current display managers, from embedded to desktop systems. LightDM is still in early development and not suitable for use in production systems.

Where can I get it?

Source is in Launchpad (bzr clone lp:lightdm) and tarball releases.

If you are using Ubuntu (Lucid or Maverick) you can install it from a PPA:
$ sudo add-apt-repository ppa:robert-ancell/lightdm
$ sudo apt-get update
$ sudo apt-get install lightdm lightdm-theme-webkit lightdm-theme-gnome

You can test lightdm by running it in a window:
$ sudo apt-get install xserver-xephyr
Create the file lightdm.conf:
[Greeter]
theme=webkit
$ lightdm --test-mode -c lightdm.conf
11/11/2010 - Updated test instructions

MMS internet radio in Rhythmbox

If you listen to MMS internet radio in Rhythmbox and get a "Couldn't start playback" error, there is a workaround. It could be that your stream is actually RTSP - change the mms: prefix to rtsp.

e.g. mms://wms-trn-all.streaming.net.nz/trn-newstalk-zb-akl -> rtsp://wms-trn-all.streaming.net.nz/trn-newstalk-zb-akl

This is correctly handled by GStreamer but Rhythmbox currently does not handle this. Other applications like Totem and Amarok work fine.

PyGI in Maverick

PyGI is now available in Maverick, you can install it with the following:

$ apt-get install python-gi gir1.0-gtk-2.0

and then write applications like this:

from gi.repository import Gtk

def destroy_cb(widget):
    Gtk.main_quit()

w = Gtk.Window()
w.connect('destroy', destroy_cb)

l = Gtk.Label()
l.set_text("Hello World!")
w.add(l)

w.show_all()
Gtk.main()

Have fun!

EDIT: Updated to note you need to install the GI package for the module you are using.

GObject introspection in Ubuntu

At UDS we planned discussed how Ubuntu will support GObject introspection. I am currently working on packaging PyGI for Maverick.

You can read more about this in a blog post from Tomeu Vizoso who came to the session and is an upstream developer for Telepathy, GNOME and Sugar and helped inform us about GObject introspection and PyGI status.

GNOME Games Help

If you've just read Paul Cutler's blog post about Mallard documentation and want to get involved, consider translating some of the GNOME Games help!

The games are a good project to start with because their help is relatively simple.

I've opened bugs against Sudoku, Mines, Mahjongg and Quadrapassel (Tetris), leave a comment if you start working on it.

To get started look at the existing documentation either by running the games or browsing the help in git. Learn about Mallard by doing the ten minute tour.

GCalctool 5.31.1

GCalctool 5.30 was a case of two steps forward, one step backwards as I didn't have the time to complete all the changes I had planned. GCalctool 5.31 has these changes. Note that this is an unstable release at the start of a development cycle; the design will be refined before the 6.0 release (GNOME 3.0).

Introducing basic mode:


The most noticable change is the use of colour thanks to a tip from Chris Lord.

I've also added memory operations and squares and square roots. I'm not sure if this is too much (or too little) for "Basic" - feedback welcome.

Advanced mode shows a lot more functions. Also of note is the ability to see angles in degrees at any time:


Programming and financial have similar changes but need more work. Number bases are easier to work with than in 5.30.

The internal changes that you can't see in a screenshot:

• It now starts fast. Really fast.


• It's a well behaved GTK+ application. This means that keyboard input and accessibility should work without any problems.


• The code is modular. The calculator widget could be split out and used in other applications.

Please try it out! If you are using Ubuntu Lucid you can try it using the GCalctool PPA.

There are a number of design issues that I don't yet have good solutions for, ideas and mockups welcome:

• I'm not sure of a good label for the memory buttons, the current labels don't feel right.


• There's not a good way to delete variables/add new ones. I've been trying to integrate this into the popup menus (GTK+ is getting in my way).


• The colours are done using primary colours and blending them with the GTK+ theme. I tried using Tango colours but they looked worse to me.


• While removing the spacing between the buttons has reduced the optical illusion of the grid it still remains to some degree. I'm not sure if I can get around this without making custom widget (I'd like there to be no space between groups of buttons). Note the buttons do not have to be in a grid so please do a mockup if you have a good idea.


• The programming bit editor is too big.