Monday, October 18, 2010

On true motives - Some remarks on Bradleys Kuhn's blog post

DISCLAIMER: The following is just my personal opinion on the blog post. It is no official statement by any organisation or something and I admit that I wrote this on a netbook running Kubuntu, which probably makes me biased.
Before I go on, it is probably worthwhile to remind ourselves: Bradley Kuhn is certainly a guy who lives up to his own standards, e.g., working as the director of the Software Freedom Conservancy pro bono for some months and Mark Shuttleworth dumps larger amounts of his own money (estimated 8 million US dollar each year) into a company called Canonical that works on Ubuntu. He does not only give his money, but he also works on Ubuntu himself. So, from the outside I'd say both share a passion for free software and both of them do not want to give their money for the good cause forever. Kuhn will be employed in January and Mark Shuttleworth tries to make his company profitable.
In his blog post Kuhn asserts that it is the admitted goal of Shuttleworth to make Ubuntu an Open Core project. My problem with this is twofold. First, Kuhn's understanding of Open Core is quite different from my definition. For him projects that require some sort of copyright agreement and do not allow outside contributions are Open Core, for me a project is only Open Core if it has an open core and some additional (but very necessary/desireable) proprietary stuff on top. I do not see any evidence that Ubuntu is going for the latter, but I agree that the copyright agreement can be evidence for the first. Second, Shuttleworth does not even say that he wants Ubuntu to go into this direction, let alone admits anything.
I do not know whether Shuttleworth's comparison of the history of GTK and QT is factually right and whether his claims on this topic have any merit, but I guess they don't. So, instead of complaining about this Kuhn talks about Canonical's copyright agreement like in some other blog posts before and jumps to conclusions not found in Shuttleworth's utterances. As Kuhn said "I am reading between the lines". For someone who likes to refer to the golden rule of network protocols "always be conservative in what you emit and liberal in what you accept", this seems not the right thing to do.
I do not claim that the copyright agreement might not be problematic, but I would rather have both of them check the facts and then claim something. Moreover, I would really like people on and elsewhere not to cook up these things. This is just plainly annoying!
Bradley Kuhn has put some clarifications on his blog post (in fact they were there when I was writing my initial remarks from offline country) and also responded to my remarks. Kudos for that.

Tuesday, October 05, 2010

GNU TeXmacs

From time to time I want to present some of the software I use here. Today it is:

GNU TeXmacs

What is it?

GNU TeXmacs is a wysiwyg editor with LaTeX import/export. It handles the typographical needs of those who need to type complex formulas, does automatic referencing and bibliographies without the need to know LaTeX. Moreover it allows you to type mathematics really fast and frictionless due to congenial on-the-fly composing of symbols.

What is real world use?

I use GNU TeXmacs for shorter and longer mathematical writing, i.e., from exercise sheets to complex documents. I have written major parts of my phd thesis with TeXmacs and drafts of articles for mathematical journals. Why only major parts of my thesis? Well, sometimes you have to meet certain standards, e.g., use special LaTeX classes and so on, but TeXmacs is my first choice in those cases as well. You can rapidly prototype your text in TeXmacs and export to LaTeX for fine-tunning and adjusting to the necessities. Daniel Bump wrote an excellent book an Lie groups in the Springer Lecture Notes in Mathematics series in that way.

What to try first?

You should try composing formulas first, just to get an impression, cf. Help -> Manual -> Mathematical Formulas in the menu. Insert a $ and type the following sequence on your keyboard (words in CAPS refer to the corresponding keys on the keyboard and the white spaces are just for readability)


and you will end up with a nicely formatted formula of natural numbers. This looks more complicated than it is, but is much faster than writing down the corresponding LaTeX

(x_n)_{n \in \mathhbb{N} }.

You have to try to believe :-).

But that is by no means all the goodness there is to TeXmacs. For instance you can use interactive sessions like shown below for Shell, maxima, python, R and many more:

There is also a very nice plot mode:

Where is additional information?

GNU TeXmacs has a web site at which is written in TeXmacs as well and exported to html. It is a GNU project and you can find the sources on Savannah

What is the license?

GNU TeXmacs is licensed or GPL version 3.

Any possible caveats?

Be aware that using GNU TeXmacs may be quite addictive and may really spoil the fun of writing direct LaTeX. Additionaly, there is a bug in recent version which breaks the TAB use in Ubuntu/Kubuntu, if you start TeXmacs from the menu. Starting it from a terminal solves this issue as a temporay workaround.

What will TeXmacs be like in the near future?

It will be based on QT. There is a QT port which is almost done, but just is not quite ready for productive use right now. It will look like this: