draw-something

There isn’t a web front end for it yet, but draw-something is making a drawing a day at –

http://robmyers.wpengine.com/draw-something/drawings/

Paintr

Paintr is at http://robmyers.wpengine.com/paintr/

Why Net Art Software Should Be AGPL-Licenced

Restricting the study, production, display, preservation or other uses of artworks removes the freedom of those involved in art and thereby damages the cultural, social and economic value of art. Where restrictions take the form of copyright, copyleft licences are a good way of restoring peoples freedom. The freedom of curators, critics and academics, collectors, audience, and artists to use software is part of their freedom to use software-based net art as art.

For media-based net art the Creative Commons Attribution Share-Alike licence is the best copyleft licence. For software based net art a different licence is required (and Creative Commons explicitly state that their licences should not be used for software).

The GNU GPL is the best copyleft licence for software that people use on their own computers, where it is “propagated” to them from elsewhere by downloading it or installing it from DVD. Software delivered to galleries or collections, or to other artists, counts as being propagated under the GPL, so the GPL is the best copyleft licence for software that will actually be delivered to its users.

Software accessed remotely on a server online does not count as being propagated, even if it is used as it would be locally but through a web interface. To handle this a variant of the GPL called the Affero GPL (AGPL) was created. When you use software over a network, for example through a web browser, the AGPL requires that you be able to acquire the source code of that software just as if you were using it locally under the GPL. The AGPL is therefore the best copyleft licence for software used over a network. This includes software-based net art.

The average piece of software-based net art will use a free operating system, and a free software scripting language, web server and web browser. It may use a free software database and many additional free software libraries of code as well. The difficulty of the artwork’s conception or production does not provide an excuse for making it non-free any more than the difficulties of creating the far greater body of work that it build on did.

It is much easier to install and maintain software that is not restricted by its licence and that provides its source code. Art that takes the form of software must be installed and maintained to curate and preserve it. Critics, artists, students and audience can benefit from studying the source code of net art. Even if they don’t fix bugs they can learn from it and maybe even appreciate it. And if the server goes down and you don’t have a backup, someone else may and will be able to give you a copy back. These freedoms are all protected by the AGPL, giving a strong practical benefit to using it. This fact should be borne in mind when discussing the curation, archiving and preservation of net art as well as when discussing its production.

The support of people’s freedom and the practical benefits to artists
from supporting the curation, preservation and scholarship of their
work provide strong reasons for making net art free software. Net artists can and should protect the freedom of the users of their software using the AGPL. See here for details of how to apply the AGPL to your work.

!source

I’m currently working on a free software microblog bot licenced under the AGPL, which raises two questions. Firstly, who is a user of it? The AGPL only applies to users. And secondly, how can it provide its source code? The AGPL requries that source code be provided to users.

Reading a microblog post (a dent or tweet) does not make you a user of the software that generated it any more than reading a document produced using a word processor makes you a user of that word processor. But sending a message to the bot and it responding probably does count as a user interacting with software over a network.

Microblog users have user pages that could be used to provide details of how to get the source code for a bot. But that wouldn’t be a way of providing those details specifically to users who are interacting with the software.

The solution is to allow people to send a message to the bot that contains a command which will cause the bot to respond to that message with details of how to get its source code. This both makes sure that the bot has a clear user relationship with the person requesting the code and provides a mechanism for that user to request the code. It’s a minimal use relationship, as the use and the provision of source are the same. But it is use, it does trigger the AGPL, and it does satisfy the requirement that source be made available.

The message is a simple one – !source

The response is also simple, a short URL pointing to the source repository for the version of the code the bot is running (modified versions have to supply their own URL).

Send !source to a microblog bot and it will tell you where you can get its source code. I’ll be implementing this in the microblog bot project that raised this question and I’ll be implementing it in my art microblog bots as well. I recommend it for any microblog software that users can interact with, and as a way of ensuring that users can interact with the software and get its source code even if they otherwise would not.

(With thanks to Matt Lee and David Bausola.)

Paintr

I came up with the idea for paintr one Friday morning in 2005 while thinking about Harold Cohen’s arguments regarding computer art in his essays and while thinking about the work of Pall Thayer. Paintr’s tag line was “art in the age of network services”, or “art as a network service”. By lunchtime I had something working, and by late afternoon on Saturday it was feature complete. A few weeks later I exhibited it at my show “Howto” in Belgrade.

Artists don’t make art by sitting around waiting for flashes of abstract conceptual or aesthetic inspiration then realizing it in visual form, but paintr does. The original version did so purely using Web 2.0-style web services; colr.org for colour palettes, flickr for (copylefted) photographs, and an online version of autotrace to convert the photographs to drawings. These paradigmatic web services were glued together with the paradigmatic web scripting programming language PHP.

Many of my projects take a linguistic (verbal or visual language) description of art or reality and drive open the cracks in it by taking it literally to making something ironic and unstable. They are disproofs of theories, illustrations of mistakes, but they have a remainder that has its own meaning or effect. Paintr is a good example of this. It’s an analogue to art or artistic activity, the realisation of a popular misconception of how art is made. It’s an exploit on the idea of art or on the misunderstanding of it.

The relationship that paintr has to Web 2.0 hype is similarly ironic. Web 2.0 makes it easy to create new software by gluing together the public APIs of web services, but you are limited in what you can ultimately do by the affordances that those services provide. Human socialisation can be planned, effected and recorded online in great detail and with great reach through social networking sites, but it is reified and channeled through normatising affordances. Art isn’t something that should be created and vended as a web service like weather data or news tickers, but if that’s the case what is special about art as a human activity that isn’t about human activity in general?

Paintr makes something that isn’t art. It’s easy to say why it isn’t art but it’s less easy to see why it isn’t art, unless contemporary art of the housepaint-on-aluminium school also isn’t art. This entanglement makes paintr about something more than itself artistically as well as socially. Art computing is usually dismissed out of hand by mainstream art critics because of its perceived lack of psychological content, subjectivity, interiority, or affect. Dismissing paintr on that basis is trivial because it isn’t even trying to express something. But the intentional fallacy starts to seep through the cracks, and entanglement means that this leads to collateral damage for more critically acceptable forms of art.

Aesthetics is resistant to corporate information culture because quantifying it doesn’t capture its value. We can chain back from this obvious example to the more general case of human experience. The supernaturalism of qualia isn’t necessary for aesthetics to have an experientially irreducible core. But paintr itself cannot experience this core. It weaves human affect and activity into its activity (colour palettes and images posted to social networking sites) but it is inhuman, beyond even death-of-the-author, a representative of corporate information culture and its exploitative cultural asset-stripping of “cool”. It loops back, conceptually. The remainder of this loop is its artistic value.

The latest version of paintr has a back end written in Lisp and runs autotrace locally. It now has an RSS feed, always part of the plan, although it doesn’t have an API yet. It’s going to expand to start from expressing emotions rather than from abstract aesthetic inspiration. It will probably use Wordnet to map more creatively from its initial tags to the colours and images it searches for. It is becoming increasingly an example of social-network-based collective intelligence programming and increasingly an example of how this reifies human experience. And it looks good while doing so and in order to do so.

Art Loves Wikipedia

Furtherfield have published an essay by me on how digital artists can work to help improve Wikipedia’s representation of digital art by becoming editors and participating in Wikipedia. There’s been some good discussion of the article on mailing lists, and hopefully the article has demystified Wikipedia’s editing process a bit for artists and encouraged people to get involved. Click here to read it –

An Artists’ Guide For Editing Wikipedia

Wikipedians will notice a curious omission from the article – I don’t mention NPOV. I did this deliberately as I didn’t want to distract from discussion of the ideas that I have seen people on mailing lists having practical problems with. Which is why the article focusses on notability, sources and deletion. Hopefully it explains some of the rationale behind them and how to work with them to make better articles for Wikipedia, for society and for digital art.