Safe as Milk

Fluendo releases proprietary codecs

I wanted to congratulate Julien, Christian et al at Fluendo on the release of their codec bundles this week. According to the arstechnica article:

Fluendo’s codec release is bound to stir up controversy and generate criticism within certain segments of the open-source community. A small but vocal minority of Linux users vehemently oppose the commercial sale of proprietary codecs for the Linux platform since such codecs limit user freedom and impede open redistribution. Critics are likely to perceive the sale of codecs as validation of proprietary software business models and a tacit rejection of open-source ideals.

Au contraire, I think it’s great that a company is offering for-money sale of commercial codecs – it’s infinitely preferable to “free” codecs that people are downloading and using – often in non-compliance with the GPL (“but that’s OK, we download them separately, and we’re not redistributing the aggregate work”).

People should realise that proprietary codecs are just that – proprietary. And if they cost money, that’s a great way to realise.

I wish Fluendo all the best with their shop.

Thought of the day

GNOME is not a project which is independent of commercial interests, as much as we fancy ourselves as that.

On the contrary, commercial interests are all around us – several modules change maintainership when the maintainer leaves a company, and most of the committers on many modules come from one or two companies.

The question, then, is not how we go about integrating commercial interests into the project, but how we mould the commercial interests already in the community so that they are community-friendly. The best way to do this is to first recognise the existing commercial forces, and to evolve the project so that GNOME can become a true center of collaboration and communication.

STFU

No particular motivation for this, but it’s probably something everyone in free software (and IT in general) should re-read every 6 months or so: An open letter to other men in the movement by Dan Spalding.

Impressive GIMP plug-in

This is quite possibly the most impressive GIMP plug-in I have ever seen.

Previous favourites gimpressionist, IWarp and GREYCstoration fade into the shadows when compared to this magic stuff.

Help us budget for user groups!

I just sent a request for proposals from user groups to the gugmasters mailing list. The board needs to know how much help user groups need for the year to budget properly, and although we will leave some slack for ad hoc last minute requests, having lots of info up-front will be very useful.

Anyone working with a user group who would like to make a proposal should sign up to the gugmasters list today, and help us out. Thank you!

And then there were 5 (in total)

It is with great pleasure that I present to you the newest member of the Neary family, which, while waiting for a real name to be chosen (once we know whether he’s a he or a she) will be known by the codename “toto”.

We had the first scan today, and everything is perfect. If anyone who hasn’t seen one of these needs directions, ask in the comments, I’ll let you know where all the bits are

Update: It seems some people thought that this was going to be my 5th child – I can confirm that the 5 includes myself and Anne – that’s 3 kids. Isn’t that enough?

Property and investment

One of the most controversial blog posts I’ve written here hasn’t been about software, it’s been about French property prices.

I recently bought some shares (for the first time) and following share prices has caused me to think about why investing in property is such an attractive proposition.

The answer is leverage.

The way leverage works is simple – if I invest €100 of my own money into something, and it goes up in value by 5%, I make €5, or a 5% return on investment. If I invest 10 euros of my own money, borrow €90, it goes up 5%, and I reimburse €92, I’ve made €3, or a 30% return on investment.

Mortgages work the same way – if I buy an apartment for €150,000 and resell it for €200,000, then the more I have borrowed from the bank, the greater (in percentage terms) my return on investment.

Investments which go the other way also magnify the effect – in that initial example, if I put €10 of my own money, and I lose 5%, when I’ve paid back €92, I’m left with only €3, a loss of 70% of my capital. I can also end up in debt if the investment loses any more than 8% of the initial value.

How do people get leverage for investments other than the property market? I have no idea. I would love to find out, though. Anyone have any tips?

SVN migration

To answer the several people wondering abuot why we’re migrating to Subversion for GNOME, I have a few thoughts.

We’re currently using CVS, which ucks for a number of reasons which everyone agrees with. File moves, support for grouping changes to several files into one logical entity, efficient branching and tagging, versioned metadata – these are just some of teh features you expect from a revision control system that CVS doesn’t have. So moving to something else makes sense.

Of all the other version control systems, one (Subversion) has succeeded in fulfilling a number of criteria:

1. Providing and communicating a migration route for CVS users – not just converting a repository, but also migrating your daily use from one to the other.
2. Being included by default in pretty much every distribution.
3. Generating copious documentation, making life very easy for repository admins.
4. Providing an admin-friendly ACL permissions system for repositories.
5. Providing a firewall-friendly way to get source code.

So there are a bunch of good reasons to move to SVN, and the only argument that people have against it is that “everyone knows that distributed’s better”.

For distributed systems, the observation I’d make is that there are several distributed revision control systems (bzr, baz-ng, svk, tla, Mercurial, git, …) so (assuming there’s agreement that’s the way to go) which one do you pick? Obviously there’s some argument about the Right Way to do distributed version control – otherwise, why so many to choose from?

Also, none of those systems has reached a level of maturity and acceptance that you can reasonably expect every developer to have it installed – so you’re adding an entry barrier (compiling the revision control system) for everyone who you want to build your stuff.

In short, the move to Subversion makes a lot of sense for GNOME, and a move to anything else would need a lot of good arguments backing it up before you could even consider it.

Bkor: One small correction – the board did not ask the infrastructure team to move to SVN – Jeff was the release manager at the time he asked, and it was in that capacity that he made the request. The board has never (in my memory) discussed version control systems.

Globe follow-up: Dymaxion projection

After yesterday’s post, I went looking for one of those area-preserving maps I mentioned. Aside from the usual spherical and sinusoidal projections, I found out about the Dymaxion projection.

It’s an invention of Buckminster-Fuller, also famous for designing the Geodesic dome, and the Buckminster Fuller sphere (commonly known as the Bucky ball), which is now the classical geometry used for footballs, and who gave his name to the Carbon allotrope Buckminster-Fullerene (or C60) which has the same structure as the Bucky ball.

The globe is mapped onto an icosahedron, which almost eliminates distortion, and then unfolded to a planar surface. One interesting characteristic of the projection is that it’s possible to unfold the shape so that the continents are almost contiguously connected.

The other interesting thing is that once unfolded, it becomes obvious that Australia and Africa have almost exactly the same shape.

My first globe

Recently, I’ve been seeing a bit more of the world – in recent months, I’ve been to Brazil and Portland, and over the next few months, I’ll be in LA, San José, Brussels and the UK, and probably more.

So now I get to answer questions from my eldest son Thomas like “Is Brazil near school, Daddy?” – I’ve come to realise that Thomas has no sense of the scale of the world.

So I decided to make the boys a globe light.

I bought a cheap paper ball light from Ikea and printed out a decent sized map of the world, drew some pencil guide-lines around the tropics, the equator and the arctic circle, and I was off.

One of the things you learn when drawing a map is how little land is below the equator. The really interesting bits are at the top – so much so I almost hung the world upside-down just to see what it would look like.

Another of the things you learn is that you will always draw the bits of the world you know better than the rest (see Europe). Australasia in particular is a pain to draw – what with Japan, Malaysia, the Philippines and Indonesia, it’s just a bunch of these tiny interlocking islands.

Also of note are some controversial colouring decisions – Turkey is in Europe on my globe, and part of Malaysia is in Australasia, while the rest is in Asia. No need to write in about these – they were simply convenience decisions on the spur of the moment.

Another thing you notice when you try to draw things to scale is how stretched Europe and the US are in our classical world maps. I know that there are map types which preserve area, but I guess that we don’t use them much because people aren’t comfortable with how small Europe is.