Freestyle - FLOSS In Design - Transcription 2

Freestyle transcription pt. 2 Willie Le Maitre, Playlist Graham Harwood, The Gimp Eric Dooper, OSSL Kit Blake, Silva Jaromil, Dyne:bolic Artem Baguinski, V2_Jam

Steve Kovats: Ok can everybody come back to these comfy little white seats here, in this wonderful black box, to start the second half of our programme (...) Ok thanks a lot good to see you all back and hopefully the rest will trickle in shortly. We are just about to enter the second half of today's session which is gonna concentrate more on actual projects dealing with the subject matter and for that i'd like to introduce Matthew Fuller, the rec... the lector (lol), I always say rector, at the Media Design Research group at the Piet Zwart Institute. So Matthew you'll be swinging the mic this afternoon.

Matthew Fuller: Ok thanks. What we wanted to do with this session is basically to present six different projects that bring up a number of the issues, again through practice, that we've looked at first in the morning. One of the things I think that has been very good about the morning session it's been the ability to set up a position where cultural practitioners, artist, designers are able to talk with developers, programmers and so on. I think one of the ways which this is able to be done is directly through actually talking about software, talking about the ways it's used, talking about the very kind of material quotes of working practices and the way work is organized, the way, for instance in Roger's presentation about the relation between printers - you know, people that actually run printing machines - and designers and the software which they use to construct their organization between themselves. So I think within all of these projects that have been presented this afternoon there is also this question of work in practice, this question of a culture around software I think that's one of the things we need to pay attention to here in order to maintain a kind of discourse between the projects. All of the projects will present for roughly 20 minutes and there will also be time for discussion at any point so do feel free because, I mean I think we've got a pretty literate audience, so do feel free to interrupt and ask questions and so on, as it goes on because I think the main emphasis is on dialogue rather than anything else. The first of - there are two projects from V2__ here today. We start with one, playlist and we'll end up with another one, V2__ jam. The reason why these was done was I think to pay attention to V2__ as an institution that is organizing open source projects in order to think also about what the space for longer term organization is in relationship to this development. So I want to introduce Willi LeMaitre of Playlist which is an online authoring system for collaborative streaming. Again I think one of the things that Willie will raise is this question of working between developers and users and also the question of open standards. Both issues that also came up this morning. Ok, thanks.

Willi LeMaitre

Willi LeMaitre: I identify myself primarily as a video artist and the part that I consider myself as a new media artist is ... I really identify as located in extending the bounds of video-oriented work, and finding and exploring new situations of collaboration and ways of presenting work that perhaps will throw some existing hegemony of the moving image, you know, some kind of a wrench. And I'm coming here to speak about the Playlist project which happens to be directed to both open source guided principles but as a project in process we've come against to some things that kind of have thrown us into a dilemma of the forms versus the content in the project. The content being the video and the audio project that we are working on collaboratively and in terms the software itself which we've considered from the beginning as a vehicle for perhaps changing the media. So we identify the software as content as well, but the two aspects are in some kind of battle for time, and a backward scheduling within you know, a large project. I'll get into that a little bit later, and first I will give you a little overview of what the project is and of how it came about. It originated by a number of video artists and audio artists that were (...)(since) scattered in different locations who got together and wanted to work on a project where they could carry of the editing and the development of the project, discursively where you know, in sort of a conversational mode, they could interact with one another's work, you know, editing conversationally and not just adding to one another's work's sequentially and chopping other people's work and interrupting but...interrupting in the sense that the conversation is like one in a room where people can break of into tangents and explore different avenues and the resulting work would be visible to viewers as a kind of graphical representation of the structure, the multi threaded structure and could employ some of the methods of making the work, the conversations, the metadata that was available to us to put into the database and visualize the work using some of this techniques or this developments and techniques.

So essentially the Playlist software is an online editing software plan to access a ... allow the artist to access a central database and share each other's media, dragging and dropping their media onto a sort of white board that is live to all of participants and they can sequence media and a library of this structure. This is a, this part of the project is something that the artists have had to... kind of, work in theory that this works and sort of approximate a type of work that would fit into this, when is done...

I can't show you at this point the software as it is in process but I could just, in the background, show you some videos as a kind of example of a kind of content that has evolved with the consideration of this non- linearity, and (is this playing back right? ok), a consideration that a secured structure that has no beginning or end, and a consideration for narrative where the work would inevitavely progress in time but could not predictively amount to something that you know, you would never see again. there is always kind of a recursive quality and you know, in design, the viewer would never see the work quite the same way twice, but they would see it kind of in fragments, this kind of multi-threading was kind of a deliberate avoidance from a user interface that would give the user an option to go to sort out, other ways of searching the work and exploring on that (...) I'll get into that a little bit.

The work was conceived with the idea that the representation of this multi threaded structure would require, or would benefit very much from a dynamic 3 dimensional representation of the growing structure and with the interaction of being able to search it for keywords or maybe represent its scale with different parameters, it could be... We consider it as a work of the artists as much as anything we would do with the audio and video exploration. So that's kind of it in a nutshell, and this question of it being open source is kind of a, it's an aspect to the project but it is not an overriding concern. There is an obvious appeal for to direct work towards this open source emergent scene or world or direction... It appeals, has appealled to everybody involved in the project politically and all things considered it was a, to begin, a priority but we found with the passage of time, at some kind of compromise of finding that we couldn't do certain things and that we needed a player that was a proprietary player that was a proprietary player and even though we could have a code that was open source there was inevitavely something where we had to make a compromise, you know I think this is probably a fairly common for people developing applications in this field that are hardcore you know, directed to the tool itself but are focused on the content as well and this project being a user/developer project where the development of the project is done in a closed consultation with the artist in some ways they are testing the work as it develops just sort of mannered the work in a certain way.

In terms of having the option to develop an application proper as opposed to some sort of stretching the limits of an existing set of tools, the thing that we've really had to decide on was this sort of priority to make a new tool that could possibly advance the medium in some way and in some scope but would perhaps have a much more narrow audience and have a, or be more theoretical, or the other tendency or inclination was to take existing things and stretch them and break them and sort of orient the work towards to some kind of reconstituted avangardism and modernist thing of like crashing and surpassing the limits of something given. So we were in our case were directing our efforts towards something for the Linux platform. But we found that in a rush to deliver a product, for the strange form of funding that we were granted to do this project, we've had to compromise it.

This is like something that is perhaps, you know something to reflect onto with open source tools is this possibility of a... the project was developed and proposed to a sort of industry research and development program in the canadian government, never have they heard of artists doing...their work could be constituted as research proper, so it was something of a kind of an experimental project for them where... to consider that artists themselves could do research, was a bit of a test and this is... there is a favor in the consideration for the project that it would be open source and... I don't have much experience with funding with the sort of industry oriented work, I work almost exclusively by subsidized means or artist typical funding for, available to artists, so I don't know if there is some industry people in the audience this is some kind of maybe there is something to discuss here, with the possibility of artists and designers in those concern being able to pull interesting platforms and applications out of sort of, research funds. So, I'd like to show you, has this been looping? I sort of got a bit distracted.

I wanted to show some examples of some of the other artist's work that are involved in the project. Presently we are 4 here at V2__ doing a workshop residency, Sonia Cillari, Jaanis Garancs and myself and 'evolutie', Brigit Lichtenegger who you probably are familiar with her projects synesthesia, data cloud that are inspiring to us and we identify with their work that we are doing here and that brought us to V2__. So this is work by Sonia Cillari who is involved in the visualization aspect and the interface for the presentation of the work. She works in a sort of dynamic architecture, based on a different concepts of input and representation of those inputs. So, yes fine work. It's not the time really to show you a full demonstration but you can see a very compromised example of her work.

This is another, Jaanis Garancs who is working with Sonia, Brigitte and me doing...he works primarily in VRML, which is quite an interesting open standard form that perhaps will make some kind of re-appearance and reconstitution in the Linux world but... you have to really see the work to really grasp what goes on in it . He works primarily showing his work in stereographic format so he is involved in this aspect where the video work is annotated and notated in a column of the database, mysql, a database and the VRML interface pulls the database for cues to sort of respond and be fed, sample images and reference other's, as the work is the kind of illustration where the interface proper to interact with the work as a kind of net like structure showing the multi threaded work. And I think that's it probably I shouldn't take anymore time and I leave it up to the other fine panelist that are up in coming, looking forward to hearing from them. Are there any questions for...yes?

Calum Selkirk: What kind of video format were you using that became problematic in terms of...?

Willie LeMaitre: Specifically mpeg4 which was a kind of sought after format that we could use in a work flow that was familiar to the environments of the video artists that weren't really that technically oriented and that were using standard tools of mac and pc...

Calum Selkirk: But you said that it was a client-end problem because of the player...

Willie LeMaitre: It was to do with the 'smile', which the synchronisation protocol that we needed to have the editing sort of...

Calum Selkirk: But it was something outside of the actual encoding scheme...

Willie LeMaitre: It was determined that we would in order of having an open source streaming server, Helix real server ( www.helixcommunity.org) we would stream and have the stream in Real and, it's complicated but it mostly has to do with the 'smile' being a kind of thing that was only supported by Real although initially it was supported by QuickTime and things like that...

Matthew Fuller: Ok, so I want to move to Graham Harwood. He is an artist based in Southend-on-Sea in the UK. Graham's had a long working history in relationship to the use of computers for design and art practice. He is the author of the world first computer generated comic, back a long, long time ago. But since then he's also been developed a number of software initiatives including NINE a social software project developed with the Waag in Amsterdam. And has also, here he is going to present work using the Gimp but not using the Gimp as something that is supposed to be a parallel to Photoshop, but as something that actually benefits quite radically from the connection to networks. So if we think here what is the difference in paradigm between the network and the stand alone application I think this make a very good example of cultural work done with networks through an application that is developed not for designers but through the actions of developers to meet their needs and we'll see some interesting results.

Graham Harwood

Graham Harwood: Good i never listen to a thing you say...

Maybe we can begin a little bit, okay, what I wanna talk to you about is the Gimp and the first thing I kind of already want to talk about is the history of the Gimp. These are the two young men that actually made it when they were at college... And we can have a look at the history here (http:// gug.sunsite.dk/docs/Grokking-the-the Gimp-v1.0/node14.html") The important thing about the Gimp is that it has no colour theory built into it. And this is... can be quite a problem if you wanna do something like set your scanner up to have the same qualities as your printer and your machine: which basically means that the whole idea of desktop printing with the Gimp is not useful, at least in the version that I've been running and using. And so, like, at first you would think that if you was gonna do any kind of DeskTop Publishing this is completely useless and it is. But if you...but that would be completely the wrong approach to look at the Gimp from.

From the late 80's I started programming paint systems, and so like, sort of have an internal way of how this thing works and if you look at Photoshop... Photoshop comes from a culture of photo retouching,that's what it is photo retouching in a software package. The Gimp, on the other hand, comes much more from the point of view of like web administrators needing a tool to automatically generate buttons and, if you look on the Gimp, if you look on the actual interface, we may have some here(...) then what you tend to see is that most of the elements on there are filters for generating buttons on the fly as a web administrator.

One of the things that Matthew asked me, ...if I can, let me see if I can show the actual user interface here it is, you can see here, look here, colours!. Matthew said to me, I've looked at your notes, but the thing is can you show, you know, the desktop version of the Gimp?. And I said no, well I've never really used it. And it's not something, even though I've used it, it's not for looking at where tools are or something, it's not really an approach that I've had in my work whatsoever and if we go back...

So I am not gonna talk about the new version of the Gimp. I've not really used it in a way just like Photoshop because I don't think it comes from the same culture, I don't think it's the same kind of software at all. How I've used the Gimp is as a server and what that basically means is that it's a bit like Photoshop but with all its guts hanging out, so all the wires are showing and what you can do is... you can actually start to plug together the different functions. So most people I expect to be more familiar with photoshop but, means that you can automate everything within Photoshop and if you can program everything in Photoshop and with the added advantage that like with the Gimp, the the Gimp is online so on a server you have Photoshop you know serving the net and then any image incoming can be live processed however you want and put back out again that's beginning to see where the Gimp could become, as a design tool it can become, you know it's just fantastic to be able to use it in that way...is really important... let me just remind myself of my first notes...

Okay, a really important consideration in this, because I'm mainly talking about image and me and Matthew had recent discussions about the difference between images that are networked and images that were not. So images in books the way that they work, it might be the same book with different texts. And the way that that relationship works, and then an image that's networked which is in constant flow. There is a really big difference between the two. And then the Gimp can respond to those images that are networked in ways that Photoshop is just not designed to do...

Okay, so a really important thing whenever I go to a new piece of software is trying to understand the cultures from which it is emerging and you saw when I looked at the user desktop here this is where the culture of the Gimp is coming from and there is here, some great examples of fantastic desktops that people have got going. And see you kinda look the cultures of where this code... or where this activity is coming from. This one is fantastic, you know, obviously like, take some heavy mushrooms then get on the Gimp, you know. And so you see you can quickly... ascertain what the hell is this stuff you know, who is thinking about this.

This is the icon of the Gimp here. Now think to yourself... Whoever invented this as an icon of a program really that shows where the program is coming from, you know, you can really... you can just, instinctively understand it. But anyway, I don't mind those things. So understanding the culture, this is a small image, a small part of a large image that I'm gonna show later, this is where, I wrote a piece of software that actually goes out and harvest 2500 images, bring them back and collages them in the Gimp and then outputs something that I don't understand very much. And this is like... So to understand the culture of the code's become very important and then I'm trying to deal with this complete glut of images and trying to understand what sense I can make out of them.

Also another really important feature of this is that Photoshop has defined the 'unblemished'. Photoshop has become a really important tool for us to aestheticize like the lack of, you know the lack of wrinkles in the skin and has set up a mode by which we can all make ourselves handsome and less blemished and has set up this really interesting set of public aesthetics and that's kind of become really important for the discussion I want to have. But in its construction, of these aesthetics it has also allowed the counter, which is, images like the ones that we've seen recently coming out of the Abu Ghraib prison, which is like the opposite of the 'blemish-less' image which is, you know the corpse image. So it's really important how something like the Gimp, how Photoshop, how they construct public aesthetics and then how this aesthetics actually played off each other.

So here is an introduction to the Gimp. This is mine which is like sort of, what i'm gonna show. (shows image) I've been concerned for quite a while, about how you can take an image that is an expression, put it online and that image can find its own context. And by finding its context bring back that context and place it on itself. And so here we have a little introductory script which is like sort of 'use the Gimp' and it says, constantly 'corrupt the culture of the code' which is for me a really important way to approach a piece of software like this. Interrogate where its code is coming from, what culture created it and then see how you can twist that. And here you have things like use the Gimp Foo which is a particular language, a scripting language that allows you to write plug-ins and different filters to the Gimp. So it's a bit like in Photoshop but you can just write your own plug-in and add it in. Then use PDL. PDL is Perl Data Language and a lot of scientists use the Gimp in order to construct visualizations because you can program it, you can program very fast images with it.

Here we are using my library which is Fuck The War!, then use the database insert and is using like, sort of using local time, Greenwich Mean Time which is the time origin of the empire... you have to know these things when you are coding. It keeps you sane to write your own comments. And then how you begin is you do the Gimp init, which is, like, you start up the server and the server sits in the background, you have no interface to it. You have the command line, but basically you have...the code is your interface into this, this contraption. Then you have this thing, the Gimp On Net which I don't know what it is. It creates a call back to something but I don't care it's a piece of special information that it requires always to run. And then the Gimp set trace. All this is really, really interesting in terms of the culture of the code. This is why I've been able to like spew out various uninformative crap about the functions taking place. But absolutely gives you no indication of whatsoever the hell the whole thing will work. And as an example here is the image. This is the main image that is acting on okay?, which is an image from 10 years ago, was in a publication, before Tony Blair became prime minister, of him as a cunt and what the idea was, it was to take an image that 10 years ago was used in some public environment and then see what would happen, what context would it find by itself on the net using the Gimp as a way of rendering out the information. Anyway, so we come back here, so this is, this is an early attempt but what's interesting in this is that... so this is an image that's gone out and it has harvested you know news about the war, images about the war, come back and start to stick them all over itself and try to construct itself, but the thing is that it had a procedural database that failed on line 924 fuck the war, and you can see this is actually a crash, when the crash in the computer usually tends to mean some kind of expression and so I was trying to find out this expression but actually it just couldn't handle what I was trying to do.

Okay, I wanna make kind of a point here about pictures which I think is very important when we are viewing any software, and how that can help us... so I just, so it's really important to understand that pictures are experienced through sets of photoreceptors you know, situated in our eyes and our eye muscles as they relate to the head and the organs of the head and the whole body. We don't experience pictures in the way that we generally think about them they are experienced internally and externally at the same time. And how we do that is that we find convergence in what we see. And that's very important for understanding that pictures are always problematic, they can never be, there are very precise ambiguities irreducible to language so when programs like the Gimp, actually try and process images they are processing sets of information that represent images but those images are different to how people actually imagining them. As an example of that we got Nicholas Berg here and... I'll just put down a couple of notes I thought were really interesting just for this talk about it that it's kind of a sacrifice to tactical media and that's really kind of cutting for me having been involved in tactical media for quite a while that my techniques have been used for such brutal purposes: to, actually rip someone's head off but what is also interesting... I didn't, I couldn't look, I couldn't look at it, the guy having his head cut off, I didn't wanna know, it's like, there's enough right wing agendas running around. But, what I did do... I went to the site and started analyzing what server it was running and it was running Apache which is open source and the video itself was encoded with Microsoft Media Video, which kind of made me feel much better actually...

Okay, (shows image) and it is really important to, within our consideration of images and how we may understand them to see that an image like Nicholas Berg altered the flow of oil and all its economies. And so like how are you gonna have software that can actually treat that or reproduce that or interrogate that becomes very, very core and the culture of the code creating that becomes ever more important. Okay, so here we have a bit more code where we are displaying the image, putting it into normal mode, ah, okay I know now what I'm talking about and here what we are doing... one of the things that really...

I'll come down now onto this. This is that same image of Bush that we are looking at here but what's has happened here is that on the Gimp it's gone out on the net, found all the texts of all the web sites that are about the war, that day, brought them back, brought back the font, size, the color all the information about the text and it's just like sort of... etched them straight across Bush's face in an attempt to try and understand what's going on. I don't know what to do really, it was an experiment. And you can see that like from a designer's point of view, that it would take you a very, very long time to actually try and construct this, and later on I'm gonna show much more significant examples of being able to do things with tools like this that you can't do physically where you can have you know like 20,000 texts, render them in ways that you can't... you can begin to create new aesthetic environments.

Okay bananas, I'm gonna change slightly here. Here, I got this obsession about bananas after going to Southend market and finding out that I could get 7 pounds of bananas for a pound okay that's like an incredible amount of bananas, you just can't move 7 pounds of bananas for a pound from anywhere in the world. And some friends in Jamaica had been talking to me about the problems in their economy around bananas and as I started to look for ...I started to do things like this, which is a (...shows image) so this would be like a survey of banana eating online, okay?. And you can see some various examples of banana eating here and compare that with cherry eating which is much much less, okay?. And I did some calculation on several searches doing some several thousand searches and I think I found out that banana eating accounted for 35%, yes 35.234% of fruit eating online. So, that's a significant relationship between like networks and banana eating and the representation of the images within that.

So here this is a collage of a couple of hundred images of banana eating using the Gimp. this image is about 8000 pixels by 8000 pixels and it just sits there constantly rendering people eating bananas. So the image is forever changing as well which is another, as a design, you know, that you can have an image that is never the same that is constantly moving in directions that are unexpected. Here we have, this is banana eating. It's a bit hard to see it all in this image right?. This is banana eating where it's one large image on the top, small images at the bottom dark on one side, light on the other side, and it's orientated the angles according to the mean value of each image and again this would be constantly changing image around the possibilities of banana eating.

(...) Okay, this is an image where I was again kind of you know obsessed with this kind of war theme and what was going on and so here this is 2.500 images. It is hitting 200 urls and is constantly bringing back any image that it finds and bringing it back to the Gimp, arranging it in this kind of landscape and down here I should have a... This is a large scale version of that which, which kind of indicates or shows ways in which you can begin to perceive images that kind of like you didn't before. Like this is 2500 images that is just to do on a search which is like 'Bush War Blair' and you can see that you begin, in terms of the designer, you begin to, there are tons of ideas in there that you get from just the juxtaposition that the computer is creating for you. There is a fantastic one here that I really love, I've a love of arses. Look at that one!, With Bush it's a great one. And also because it is doing a search you get things, you know, because it's Bush you get other kinds of bush emerging. And so aesthetically even though you know I'm not overly keen on these aesthetics but in terms of like visual ideas they generate very rapidly for the designer a certain interconnectedness between the images. What's interesting is that you get a lot of resistance images coming through as well which when life is shit you can kind of look there and find a lot of help and so on.

So here we have, this is the same thing the 2500 images but with a text rendered over the top and I'm not gonna zoom into this, but you can begin to see that in terms of aesthetic possibilities in programming something like the Gimp or imagining how you can take you know a photo package, just apply it to a network, you know if you can take Illustrator, apply it to a network in a different way that's, you know that's a very different notion for the designer. And here, here we have, this is the Blair one but with the text over, you know over one side of this and this is actually a massive image, and you can see that if you used Quark Express or something like that you'd sit there forever trying to put down that many texts, you'll be there a long time, well you can do this one but not a couple of the other ones I showed.

I'll show a couple of these. I'll show this one, this one is around about the eve of the war. I wrote a piece of code that would go out and amass what the changes were and the stories to do with the war and so it's like a media storm brewing over the Thames estuary (which is where I live) and if I kind of zoom in here you'll see what I mean. This is the media storm brewing as you are moving along and so this is kind of, this is the actual storm. These are the urls here which you can kind of like zoom in on to see this activity and there is probably one more example of that which I could show. This is one where what I was trying to do was well, I just scanned myself and then what I was doing was telling it to go and find images of the war, text from the war and come back and stick them on and here. So the Gimp is kind of generating this montage, then what it is also doing is like tiles the image , tells the server to stick the image on and then in here it's also got the text in a Wiki form that then people can edit as well underneath and so you see you are constantly changing this whole monster as it's kind of emerging, that you are using. But I mean, all this is stuff is still very experimental. I'm gonna finish with that. I just wanna say that if you want to design for a networked environment then you have to think very differently, outside the paradigm of proprietary software. It becomes a very different process of design.

Matthew Fuller: While Eric Dooper's setting up is there anybody with questions any kind of comment or things that you would like to know?

Geert Lovink: Just a technical question, I presume that you installed the program on a Linux box and then you worked on that or you just... And then you exported the images that it produced or? Can you just explain a little bit more about, yeah the technicalities of it...

Calum Selkirk: Actually this machine was installed by me remotely from Amsterdam. From Amsterdam wasn't it?. He put, downloaded the CD, burnt it, put it in the machine, booted from it brought up the network, sent me the IP address and then I connected remotely and installed all the software. Which just leaves you to explain the image...

Graham Harwood: Yeah I mean Calum set up the basic installation of Gentoo Linux on this machine. I really wasn't gonna go near Gentoo on a PowerPC that's beyond me. Then I would install all the Perl programs and then started editing it from there, I mean my chosen language is Perl so that's why I understand it from that point of view and I would code it that way. But then you could use the foo, script foo which is kind of a scheme, which is quite a simple language but it has a lot of problems in it, as a language. If you are a programmer you can get very pissed off very fast. But the real strength comes in when you know when you are piping different processes to the Gimp. So the Gimp, if you view it from the culture that it's emerged from then you think about how you can skew that. So if your program is linking, you know like, network traffic into it and then back out again. So, I think that with a software tool it's always how you are sending processes from one application to another - it's designing at a whole other scale of design. Matthew Fuller: Any further comments? So the next speaker I want to introduce is Eric Dooper. Eric is from Open Source Software Lab, a part of Cyburg an initiative from the Gemeente in Amsterdam, that is looking how civil society and I guess initiatives from electronic culture digital media can mesh with those. OSSL is an advocacy organization that I think is, at least for me when we were researching the seminar, was one of the best sources for information on Open Source, especially in Dutch and one of the best places to go looking for software if you want to try stuff out, so I would recommend visiting their site as well.

Eric Dooper

Eric is going to show three applications that were discussed earlier by Roger. Eric is originally coming from a developers background so you will get a different perspective on these applications.

Eric Dooper: Let me log in here first... Yes my name is Eric Dooper, I am one of the people in the Open Source Software Lab, and amongst other things we maintain OSSL.nl lab (... ) like Matthew said which at at least in our view is geared to those people that don't have information but are interested in FLOSS and want to have more information like: What do I have to do to install it? Is it actually useful? What's out there? Lots of stuff. So Matthew has asked me to take a more in-depth look at three applications: two vector graphics applications and the by now infamous Scribus.

So let me start off with SodiPodi, that's the first one and before I start, I don't have a presentation I do have some steps I want to go through but its a topic that it has, popped up during the whole morning, the comparison and actually the alternative discussion between proprietary software and open source software. Like, the Gimp is the open source for Photoshop, SodiPodi is the open source of Illustrator, Scribus the open source of QuarkExpress. To me, as soon as somebody says that, the expectation is set and people will expect the proprietary thing for free. And be frustrated, and by saying that it takes away the open mindset that you should have and could have, if you look at the open source stuff by itself, on its own merits. So having said that I'll show you what the current state of graphics applications is in Linuxland (and I'm gonna get a sip of water).

This is SodiPodi a scalable vector graphics application, for those of you who don't know what a SVG is. It is global web consortium standard for describing vector graphics. It is in XML format, which means that you can actually work on the source files, the graphics files with a normal text editor and change it. Sso it's an open standard, so you are not tied to one application if you want to edit that. Sorry, I pushed it away.

Let me start off with the user interface. I think it is quite messy I only see buttons, I have to hover above it to see what it actually does. There are funny bottoms, like I can zap it here, and then zap it back and extend it. It has a lot of functions and a lot of functions are in a right click context menu and actually stuff you would use a lot is down here somewhere... So let me start off with just doing some basic things. People who know Illustrator would probably think like this is basic stuff, but by me doing things, I'll quickly run into the good and the bad things of the application. So I've a rectangle here. And now I want to, I want to have it stroked different, so I have a stroke property file here that I can say stroke paint black or whatever, with alpha channel, let's make it red, I can also type in the x digit and make it fully opaque. Now I should zoom in because I have a very thin red border. Now it is getting clearer, I don't want any fill. Now I want to make a circle, so I find out that is down here so you have to get into the context menu then you see here the botton has changed so, if I want to make a rectangle I have to the same exercise again with my context menu to go to the other function so I can do the colour exercise again because I want to have this one in red.

SodiPodi knows the basic Boolean operations so I can combine the objects I can subtract them I can get the intersection, I can group them and later ungroup them. I can change the levels in which, well most of you who work with vector graphics know what the basic operations are with vector graphics, so this are all there. I can also change, go into note editing mode I can change these two objects, to, no let me first combine them. Now they got the color from the other ones, so if I want to change the colour again I have to do this, and this, and no fill. I want to now, go and manipulate the different nodes, so. This is the button that converts the object to a path, and I go into note editing mode, I can select these nodes and then I can make segments, I can convert them to curves, let me think where it was... hold on, I'm lucky that they are down here, because my screen is(...), yeah see? Here they are I was looking for this. I can smooth them and then just do your basic node manipulation things. This is what it is okay, yeah I'm not saying that it's great but is useful. The basic things, also texts. I can type texts and then there is a different text dialogue here and I can increase, the text and also convert this two texts and manipulate, the different letters and, well... you get the idea... So, it is usable. It is, like I said, it's not great , it lacks in my opinion quite some, well not basic, but some functionalities, specially in the import-export area. I can import a PNG or an SVG image, and I can export to an PNG Bitmap, not other formats. But this is quite in the Unix and Linux tradition, because Unix tools are actually, well yeah, Unix tools are designed for one thing and do one thing well and because there are other tools like the the Gimp, out there and you can take those and use the Gimp for example to convert it to something else. So SodiPodi is ok but what it seems me is, is the user interface, and also from the website we have the statement, what is SodiPodi? It's a vector drawing program like Corel draw, and they aim to make it the best vector drawing program from the free software world. And that's it and then it goes on to its technical merits, it has a modern display engine it gives you an antialiased display and it's written in this tool kit. From a user I don't think that I care a lot, and that's the thing I noticed in a lot of these applications and I'm really, I'm a big fan of open source, but the focus is technical it's the origin, somebody has said ok, I'm gonna make this program, the birth was at this particular problem which is solved and then it grew, and the user interface it's been noted multiple times before, also by Roger is coming last for some reason, and that's a big inhibitor for I think, wide acceptance. Let's see if I have... SodiPodi has a fairly active user community and one of the things I find very interesting that they support is a flag collection. There are I think 140 different flags from all different kinds of nationalities made in SodiPodi and distributed with the software. So, I find this content great actually. These flags are made available under this open source license so not only the source code of SodiPodi is free but also the stuff they make with it and since it's SVG it's an open standard and they say, well here it's under the Creative Commons license. It's usable, so instead of buying an expensive clip art collection with nationalities, users of this program have created that. So, I find it quite interesting that the users who are probably also kind of developer types, also go into content and then contribute that, so I can show you some flags. Another thing, where you can see that it comes from is that it comes from the Unix world, SodiPodi is like the Gimp. SodiPodi you can use it from the command line and use it as an engine for something else... so lets see here are the flags, I can - these are all flags made from the people that use SodiPodi - pipe this into another program, to the Gimp as a link in a long chain of things. And this to me is also one of the things that is unique to open source programs in my view.

Well, enough from SodiPodi because the next application, Inkscape, is also an SVG program and actually shares a lot of heritage with SodiPodi .This project is about a year old. It's started by people who are dissatisfied with the pace and the direction the original program had taken so they took the source code went into their own direction so as you can see the user interface has improved, still looks like from last century big icons. But at least things are accessible and I can still easily find them. Also functionality has improved in Inkscape you have things like multi stop gradients, more boolean operations on paths all this kinds of things so you can actually create some nifty stuff with them. So I made, it's quite big, hold on... I wanted to make a shiny button. I see lots of shiny bottoms, I'm not an artist, so I'm strictly a user, I've no creative vibes whatsoever. So but I thought let's try to create this shiny bottom so with Inkscape I prepackaged it (28:40)

So let's try to create (...) with my white thing I thought well ok, it's not bad for a techie. So thesse are actually, in here, are quite some useful things, for example this alter, this ring here has a double gradient , this is a complex shape with a single gradient this is the result of a boolean operation, looks all fairly simple , but I was quite pleased. Also cool stuff in Inkscape is the way it handles nodes. I can do stuff like this, I can then select it and get my stroke dialogue... I do not wanted filled but I do want it stroked. I'll convert it to a path and then, there are lot of nodes here and I don't want them all, so they have the simplify option which maintains the original shape but with less nodes you can also repeat that. But now I want to have a contour basically get an outset, and now I have an outset here, I think, no, how did this work again?. Oh yeah, and this is quite useful because now when I modify the interior shape, the other one goes with it, so I think this is nice. This are all little features that probably they are in Illustrator and on the other packages but to me this little things make it useful, and there are a number of other things that, which shows that this program is getting there. For me, because I work on Linux exclusively, I don't have Windows on my machines anywhere, these applications like SodiPodi, Inkscape and Scribus they are almost a necessity because I have chosen not to work with Illustrator. And, I could buy them and I could go to the Windows site... But I like Linux a lot and actually every time a challenge for me is to think how far I could get with just open source tools so for Inkscape I think this application is getting there, it's not going to be an Illustrator, and like I said before it shouldn't be, because your thinking and also the thinking of the developers is constrained and, I was very impressed by Harward just now, what he thought of doing with the Gimp, that should be an insight into what the effect of Unix culture might be on design.

So now, to a third application. You've heard a lot about it by Roger, Scribus, again, to me, I'm very happy to have something like this on my platform, because there is nothing else that I can do if I want to use this kind of application. If I want to typeset anything, if I want to produce a flyer or a menu. Because, Open Office or any other word editor just doesn't... it's not good enough... So let me walk you through some steps in Scribus. Scribus has a lot of features, one of the most important things, is colour management, so it is there, you can choose. You can even create your own (ICC) profile or use other ones for your equipment and you'll have to have that if you want to produce high quality pdfs. But the installation of Scribus, it's kind of tricky because you have to have another library installed, a little CMS library. If you want to use this feature, it's not very well documented but you have to have it if you want to do really useful stuff with an external printing company. So I figured that out after a while. So let me start just a new document, I did that a bit too fast. There are a few things that I can do here, choose my paper size, printing... like Roger had a problem you couldn't create an arbitrary document size. It would only print A4? I don't know... ah It was it the printer? (Calum: it was the printer). So I can already say here I want to do two column, text frame. I can make already a really, really basic layout on this page, but I don't want to do that. So I'll keep my margins. I want a portrait of number one. It has guides, so I can add guides to help me with my navigation basically... and replace my old ones. There is also a big reference sheet for this shortcut, it's very useful, because all of these mouse things...

The user interface of Scribus is illogical. Roger noticed that already. Things are not where you expect them to be. Chances are that they are there but in a completely different place, like print previews and extras not in a file where I would expect them and any other user by now. And the zooming I can also do here. So you see I've created some guides here, I can also use grid, I can add text frames and in the properties I can tell a whole lot about this text frame, I can tell the exact position, so what you usually would do is just throw the box on your page and then with this just type it in for exact measurements, I can then lock it so I don't accidentally move it... The text options that are operations that are recurring I don't know what it is, (he talks to Roger...) your h and your v? The horizontal and vertical distance... but you can, oh yes, you can, I can show you how... I'll get some text ...Maybe this is what... let me zoom in... I'm curious if this is what you mean... I can go in here, for example and then... Now, this is the entire text and somewhere there is a... I know what you mean and I've done this, I've moved individually the letters, ok, maybe I shouldn't do that now...

Roger Teeuwen: But you can't do it for the whole text?. You can't do measurements for the word distance for the whole text... Can you do it separately for example between the in and the 18 where your cursors are, but you can't automate it.

Eric Dooper: Also not with styles?

Roger Teeuwen: Well, I haven't found it.

Erik Dooper: Ok, you can create styles for your text and in a style you can... (no that's line styles... paragraph styles...)

Roger Teeuwen: No, it's not there.

Erik Dooper: It's not there? You are right. This is one of the things where the documentation, says differently than the actual program, because I've seen it somewhere. So, anyway, small side step... I can add picture frames, specify that the text should be close around it or not. So, it does a whole lot of what a text or DTP program should do, not everything. Because the expert says I'm missing this, this and this... and then it doesn't print alright... which to me is a bummer. Because, usually open source software is designed with a specific problem in mind, like Scribus. Like Jaromil says it's designed by somebody who had a particular problem and usually the core problem is solved well. From there it grows and lots of functionality gets tagged on. But to actually produce something printable, that's the core function of, at least, this application and I think - and from what I read - it should do. I haven't tested it myself. I haven't created full blown pdfs and sent them out to a printer. But I can show you what the possibilities are and if it would work, which I think are quite impressive, I can... According to the documentation Scribus is fully interoperable. The pdf's specs have been published so it is now an open standard, in the sense that everybody can provide input about into a document standard. So, and it has some cool features in pdf creation, again, colour management, on which paper it is going to be printed, presentation effects, I can tell you that fonts are embedded in the pdf so you are sure that your font arrives to the printer...well, maybe spaces won't but... So I think this is an area where this application, is of huge benefit for everybody who works on Linux, because to create high quality pdfs there aren't as far as I know, any other applications that can do stuff like this and usually it takes tweaking. But in communication with the outside world they should tell you what to expect and then you have to map that and make it happen. Usually there is still this state of affairs in open source land and Linuxland and I would have to adapt to the outside world. There is one more thing I'd like to note about Scribus is this... the Gimp can be extended through Perl and Script FOO and Scribus can be extended through Python, another open source language that has gained massive momentum. So there is one script that is delivered with the application and basically it's an entire application inside an application. I can select some fonts that I want to have sampled, and it creates the document for me that has all the fonts in different sizes... I mean it's a fairly simple example but it shows you quickly the power of this. And again, it can be used in a chain of lots of applications. The way they are communicating with other applications make it more possible that there are no monolithic applications that you're working with. So, in that sense, it doesn't have to do anything because, if there are tools there that can complement that, it can work with it because the communication goes both ways. It's been kind of a radical thing how I went through these programs. Is there specific stuff that you would like to see? Like, does it do this, does it do that or, does it crash?.

Matthew Fuller: Anybody have any questions for Eric Dooper?

Anne Nighten: How do you see the possibilities of collaboration between people like Roger this morning, with co-developers, working in these kind of growing applications. Because your last remark I think is very typical, a programmer's remark, and you can add another program and you can... but how do you see the possibilities of a dialogue in a kind of fruitful way, between designers and artists, and co-developers who are interested, to also include and implement this ideas. Do you have an idea on that?.

Erik Dooper: Well, the last thing you said, how to get it implemented?, That's unfortunately that's entirely up to the developers working on this particular application. Or what Roger can do is join the community and in this case it's a mailing list, describe his experiences and ask, what could I have done differently? What are the questions to ask? How can I make sure that my printer understands me? Do I need to know more of his field or does he need to know more of my field? Maybe send the pdf that was generated for inspection by the developer... Basically speak up and provide feedback. If you work with this, it doesn't work and you discard it, like it doesn't work it doesn't get better. Even complaining or providing feedback helps and Scribus has its own page where there is lots of information about how to get involved with the community. And yeah, I know I say 'you can, you can': because there are technical possibilities. But you would have to have some technical knowledge to chain these things together and some artists can do that, Harwood can, others don't want to do that. That's an area I'm not too experienced in.

Calum Selkirk: Well I mean it could be simple, I mean, you can fund software development, because these people have to eat. And some companies even say, well I need, we need an Apache module that would display pdfs and do whatever and they funded the Apache Foundation to develop that. In Scribus's defence I have to say that a friend of mine recently, in terms of printing, a friend of mine recently was making a postcard for an upcoming event. It was a postcard entirely made on Scribus and then the pdf was uploaded to a web site that do this one thousand on demand postcards and then the printing job came back exactly as specified. So I'm not quite sure, what happened there.

Erik Dooper: No, I'm not either, I'm hesitant to say well, his problem is a perhaps a one off and bad luck and for the rest, not the case that it would work perfectly. But I'm pretty confident that is a very very useful program...

Calum Selkirk : Well I mean I think the other problems are probably - the usability ones are much more substantial than what we can understand as a one-off failure...

Kit Blake

Matthew Fuller: Okay, unless there is no other questions for Eric Dooper, let's move on to Kit Blake. He is from a Rotterdam company Infrae. Infrae make a Content Management System called Silva. I think, when we are talking about design in relationship to networks and information...what about when we are talking about the relationship of images with the Gimp on vector graphics like SodiPodi and Inkscape, here we can see some possibilities for networking documents, and one of the things that Silva does extremely well is document management in large scale applications, organizations such as universities where, managing multiple variables in a document it's of primary importance. And I think some of the innovations, I think they have for people working in design in relationship with documents and documents on networks, okay, thanks Kit...

Kit Blake: Here we are... okay, now we have good resolution. I have a company together with a partner here in Rotterdam, it's called Infrae, we make open source software. That was the initial, part of the initial concept of the company, when we we wrote the business plan. One of the things we were aware of was that the traditional software companies that create proprietary software they earn only 20% of their income on licenses, the other 80% come from support and integration, and customization. And we decided to scrap the 20% and time that you have to spend on licenses, lawyers, keys, encryption and contracts and we are investing that same 20% in our software and I think, we've existed for three years, we're successful, we kind of picked the right model at the right time. We have 8 people now, 6 independent contractors. Our market is mostly the academic and research institutions but is now beginning to expand or 'cross the chasm' if you read that book, into commercial companies. We have a number of products, the biggest one is the Silva Content Management System, another one is an editor, a WYSIWYG editor, that works on top of a CMS not just on top of Silva but on others. I'll mention a couple of related projects just in case anybody out there happened to be interested or doing something similar, who'd like to hear about that. I'll talk about the Infrae business model later on, and a few things we've learned perhaps and also about FLOSS communities.

Silva is a content and asset management system, it's focused as a medium for independent publishing and structured content. All the text is stored as XML. The name Silva is a word that comes out of the botanical sciences.

A Silva is the set of trees in a particular region so if we go to the Hoek Van Holland, there is a certain Silva there, the trees that grow in that region, and we use that metaphor for the name and also throughout the user interface.

There is a complete separation in the application between content, business logic and presentation. It's unicode: you can put Chinese and Indian text in the same page. It has versioning, timed publication and a hierarchical workflow.

It began here with Erasmus University in Rotterdam, a rather big university, 8 faculties. They publish in multiple mediums every year. They had a fundamental problem, they came to us and told us about it and it's a problem known to a lot of organizations. Every year March comes around and it's time to start the new publications, the new study guides for next year, and the fundamental problem is, where is the original content?.

Yeah that might be in your colleague's computer or preferably on a file server somewhere, maybe it's in some other medium, maybe it's published on HTML on a web site made with FrontPage or something. HTML is a terrible storage medium. We came up with Silva, it stores everything in XML and it exports to different mediums and overlays the proper presentation on top of that medium.

It's built for large organizations. A strong point is that it works as a delegation system. It matches itself to the organization's hierarchy and it eliminates the classic problems of the webmaster bottleneck. The workflow flows through the publication process, there is a separation of tasks, you have people creating content, you have somebody doing design, you have a developer doing code. They can work simultaneously at the same time within the application and most of the time, nothing goes wrong... it integrates with the authentication mechanism such as LDAP, and other such systems. It has advanced user managements with virtual groups and work group creation. It's pretty scalable, it's written in Python. Zope is a platform underneath it. It does some pretty high performance sites, it works in a ZEO cluster Zope Enterprise Objects cluster with Apache in front of it, there is built in caching for static information, such as the XML but for dynamic information such as you might pull from other databases and such, that's not cached.

In the back-end, the user interface there is also, yeah, scaled you could say, hardly polished, we had a lot of feedback from inexperienced users, they've never seen a command line and we've learned a lot from that. There is an advanced metadata architecture. Users can create their own meta data set, templating system, transparent integration with other information sources, you can put 'high res' images and scale them, and the idea is that in the future it will be exportable to print and in general we try to give a system that has a feature proof content storage.

Let's take a look at it. I have been using it right now, this is my presentation so, you can see, we are about here right now, this is, the screen is (...) And these windows here basically walk you through the work flow somewhat. This is a preview of what it would look like to the public. These are properties where you can adjust metadata. Some of these things here appear in the HTML, other ones are automatically stored. Here you can define access with this area, this brunch of the hierarchy within the site this is the publish screen where you can control the status of things . Let me jump up a level and go into this document. Here you can see the versioning in progress. There is a bit of colour coding, these documents here with a little tree icon have different states, different colours. Red is warning, blue like a stop light, caution, green as a go, published is a normal link and closed is gray. You can also see some icons on the left folder have a couple of trees on it, we also have a special kind of container called a publication which is a small forest icon.

I can go deeper into Silva if there are any questions we can pick them up later, but let me jump along to the next slide. Silva has extensions. These are... yeah they have different functions, one has built an (...) to the open archive's initiative, the harvesting of meta data, it also does news and RSS feeds, both reading and providing them, and then there are other external data sources, some are made by third parties such as ETH in Zurich and special objects for the University in Luton.

The Kupu WYSIWYG editor, is a Javascript editor, its major strongpoint is that it works in both browsers, most of the others out there work in one well and in the other very little. It has an object oriented architecture, CSS driven, (59:30) and saved into the server. I'm going fast, Matthew said if you have questions just shoot, but the technical stuff I'm gonna zoom through that and try to show more visual things.

It's plugable, user-customizable is open source under the OSCOM umbrella. This is the open source content management systems site, we are a member of that and interoperability between different content management systems is something very important, we think.

Here is a quick look at Kupu, if we go into editing mode. I'll go into this index document here and create a new version of it. Sorry, I picked the wrong editor. This is Kupu, it's a... it's not quite fitting on the screen anymore... but is basically What You See Is What You Get. I can jump in here and add some letter content for instance, and I can scroll up and change the style too, so a subheading for instance... this works a lot like most word processors. There are certain certain familiar icons at the top, bold, italic and so forth, there is a different kind of list that if I want to make the authors into one or another list, I can just select them, and then continue. So this is an extremely user friendly environment, there is no HTML, there is no XML, there are no tags. There is a little bit of technical knowledge needed, if I want to add an index item about the mic, then I need to know what an index item is but I can add an index item. Behind the scenes, this is all saved in clean XML with no presentation. So the authors are creating XML, but they have no contact with the underlying content structure.

Again there are more aspects, I can show off things, I can respond, you can see I'm gonna sub like for instance you can see I'm in an index, I don't know if you can see on this but, this actually highlights because I am in an index, If I switch into a link, then the link would highlight. So it basically chooses the last node and turns on what's important. If I jump into a table, then the whole table thing then highlights. And is, so far in our experience it's been very popular. But to be honest this is a bit of the bleeding edge it's still pretty buggy. So, that will improve shortly.

If anybody out there is involved with a CMS, we'd love to hear form you because it's running on top of Midgard from the Perl crew, Wyona, Lenya from the Java group and Interchange from the e-commerce group and a few others. So interoperability is extremely important .

So, my latest project. We are working on a project we call Railroad because it's a very common problem that you have a Content Management System and people say, 'great!' and they start uploading 50 gigabytes of videos and the thing goes down and half an hour later it comes back up. So Railroad is supposed to make that more manageable using WebDat, going around the content management system. Again it's an independent product that is not just for Silva or Lenya but it would work with others. The Palm project is something with OSCOM but we want to exchange the RTF data and the information between different systems, eventually building up ontologies and topic maps and the semantic web and such.

DocMa is a conversion between text tools. For instance if I do an export here, we can actually have a look at the XML, once you have XML, it's extremely transformable...This is the file I just exported. If I close it in here and open the DocMa - theDocMa project is DOCument MAker - here is the XML. So this is yeah, the XML of my presentation. This can also be translated directly to, I won't actually open it, but this a .doc doc, we can translate to a .doc format and back. And the important thing is that we can translate not just one page but a whole hierarchy, following the structure of the XML.

We are also involved in a learning management system which is being made in collaboration with ETH Zurich. Over our business model, as I said at the beginning, we do FLOSS software. Customers pay only for developments and support. We as people, are developers. We really like doing development and that's what we wanna do. The support has to come with it , so we do the support and integration but we really wanna concentrate on development. We don't do design, at least not as a service. I mean the user interface has design but we don't provide it as a service to our customers.

All users within the community enjoy the resources that we maintain. There are various issue trackers, there is an open CVS where various people have check-in rights, we also participate in Sprint. Sprint is a sort of mass programming hacker events. We held a Sprint in this very room in December, which was designed to create the next generation of the Zope platform. The guy here, I can't resist pointing him out is Guido van Rossum, the father, or the benevolent dictator, of the Python programming language. At the Sprint we had 30 people around the table hacking for 5 days, working on the next generation and that was a totally volunteer initiative of everybody who showed up there. It makes it, yeah it makes it fun. I sincerely enjoy the environment, I do question things somewhat.

I like the FLOSS name because I don't like the Free Software name, because it's not Free, at least not when you are working on this scale. So calling it Free Software is not something that I would say to most costumers because it's not!. It's not just a little package if you use something like what we have. It's more of a framework and that requires, investment either internally or by outsourcing it to a company like us. If you are trying to be successful in the FLOSS world it is essential to find development customers and those, again going back to crossing the chasm, they are the visionaries, they are the people who want to take a risk and invest in you, who accept the fact that you're gonna deliver stuff and there's gonna be mistakes, then there is gonna be another delivery and there's gonna be another delivery before it finally runs. Right now in our experience it's popular in the academic world, it's progressing in government maybe some of you have heard about the city of Munich that has gone completely in Linux and open source, that was really a breakthrough on this continent, but it's also very big in France. One of our sister companies, is doing a lot with ministeries and using... using Python and Zope is the platform. We do have a couple of commercial companies, they still... their sales, no, their purchasing departments still try to deal with it like licensed software, so they say how much does the next upgrade cost?. Well, it depends on how complicated it is but, it depends on how much work we have to do for you. And they say, oh, well you make us an offer. And so it's free but we have to spend 2 days working with them to convert the data and get their templates up to date, so the price is 2 days. They are fine, it makes sense to them, they can stamp it, it goes through the mill and the payments come in.

On the middle-long term, when I say we are getting nervous...I'm not sure how long the mill is gonna work. The first three years of our company have gone really well. We've had basically no downtime. We're, we both were busy but, we are noticing we have a code base growing and this is not just a little package this is really a framework of code... The problem that we are seeing not too far away is that customers order features, they don't order code base mantainance, nobody, unless you have still a visionary who's putting money into it. Nobody says I want to order code base mantainance, because what they get back in their eyes is exactly the same thing. Now underneath the code might be much cleaner, might be much easier to develop new features but we are running into a bit of a problem here...am I doing on time?, ok.

This is one of our customers, ETH Zurich, is one of our so-called visionaries. They have 120 Silva sites running and they're planning more on the way . They have funded a lot of developments and they also funded the Kupu editor development. It was their insistence that it run on other CMS platforms, which is interesting. They are a technology university but they are acting like venture capitalists but the return on their investments is the success on the product. Erasmus University I mentioned, Institute for Dutch History, the University of Luton. There are other Silva users which are not customers, the University of Wales, they're using all of our software and the only contact we've had is a few questions on a mailing list, but, they provide good feedback. They find interesting bugs, sometime they find the cause of the bug, so, we're quite friendly with them.

In terms of community there is a brilliant public library. The guy behind this is one of this Silva core developers with check-in rights in the CVS. So this is again a Silva site, it looks completely different that's part of our strategy is delivering some of (the blank slates) so is very easy to put your own or an existing design on top of it and then these are some other organizations that are using it as well. To round off, I wonder how many people in the audience are actively working in an open source project ... ok... well hopefully some of you are considering it because this is more or less the last slide I wanted to... just remark on some things.

Open source packages and products: I believe if there is no community around it, it doesn't have a future. Of course, maybe we are talking about a really little utility that's maintained by one person and that could be maintained for years. But, certainly when we are talking about applications as opposed to infrastructure on the Unix level there really needs to be a community around it, otherwise it won't continue to be developed and that's a problem because community, in my experience, it either happens or it doesn't, you can't force it, you can't make it happen as hard as you try. There are some thing you can do, in the early days, to try to help it along. One of the things that we've noticed and we've looked at this is, in the successful communities there is this always a personality in the beginning who becomes known and connected with the project or the package and that person is somebody who communicates. That person is somebody who answers every single e-mail, no matter how silly or stupid or whatever. It's a real mistake to have this, yeah sort of 'prima donna' attitude, 'this e-mail is just not worth my time' and you see that quite common, quite often. If people contribute stuff it's really not too smart to be fussy about it. People, yeah programmers, they are like artists, they're kind of sensitive, they don't take criticism very well, but is really not so important. The quality of the community is not the quality of the code. There are packages out there that have very active successful communities and the code is full of 'xxxs fix me' , 'fix me' but, it runs, I mean, most users have no idea what's going on and so, it's a real mistake to think that these things, the quality of the code, is terribly important... it's not . But that's a really easy mistake to make, specially with people who are, well, I'll be nice, inspired. And, you have to keep that under control.

The last point, since we are talking about design is I really believe software is like food, it better looks good. It might have the best functionality in the world, it might be absolutely delicious but if it looks bad, it's not going anywhere, so, yeah that wraps my presentation up.

Matthew Fuller: Ok, thanks Kit. Anyone have any questions or comments?

Bjorn Wijers: I was wondering under what kind of a license this is...

Kit Blake: BSD

Bjorn Wijers: sorry?

Kit Blake: BSD

Bjorn Wijers: Ok, thank you

Kit Blake: you're welcome.

Steve Kovats: Kit, I got a question here to you from the chat from somebody at the university of Hillversum, art and media technology department, committed to XML stuff. He is basically saying that in this Content Management System that you're presenting it seems to fit perfectly with XML and a lot of other speakers have also been using it and he is wondering how important the XML standard was when your particular Content Management System was built and if XML would be 'the' standard format for all open source programs and are there real limitations or other reasons not to use XML?

Kit Blake: Ok, I see at least two questions there and if there is more, correct me. When we started XML was absolutely the goal. Because we saw a lot of future in it and, I think that's now really becoming obvious because even proprietary commercial companies like Microsoft are storing their data in XML. It's really becoming a universal interchange format and if you want interoperability, which means you are not locked into one system or another, then you need to be able to convert your data. So that is absolutely essential for any sort of organization dealing with long term data that has a lifespan. If is just a forum discussion, yeah it could be html, it could be anything. But if you want to have your stuff re-usable I still think is essential. To go to the second question is there any reason not to use it? Yeah, the downside is it's not a really efficient data format and it's not fast. We actually do store XML, you saw it on the screen, but we've also done some performances tests and we know, yeah it could be faster if we did something else. The important point is that you can get XML in and out whatever your interior, your internal data format is, maybe that's something else, maybe that's something more efficient. What that would be is a bit of an open question mark. I'd be happy to take it up to guy in Hillversum maybe he has some ideas, we have some too, but that's the one limiting factor I think...

Jaromil

Matthew Fuller: Ok, the next speaker is Jaromil he is a developer of the Linux distribution called Dyne:bolic and he will show it now. Jaromil is currently a resident at Montevideo which is the Netherland's Center for Media Art, based in Amsterdam. He is also active in other initiatives such as the Italian hack labs which is a grass roots network of computer hackers and social activists in Italy also currently in ASCII, an internet cafe in Amsterdam.

Jaromil: Yeah, let's mention that ASCII is 5 years in existence so there is a celebration right now and you are invited, you know that. I have a bit of a problem with the projector because it was disconnected and reconnected so I think I have to restart my computer...no, no, no is no problem, but so I will start talking and then show you things. First of all why I am in this context, I mean, Matthew invited me, I was very happy to come, I am a programmer, I am not really a designer, even though I did some web design for my websites but I started the coding like, yeah, quite early and in general I like the code very much, I like it specially video manipulation. In the early times I tried already in the early days yeah, I tried some 3D and this kind of stuff but then I ended up doing some software, yeah some funny software like ASCIIcam putting letters into images and then I started to focus on software that focusses on communications and I started envisioning this situation on the operating system market or scene or whatever you want to call it, that is...'We have a lot of softwares to listen to but we have few software to speak with'. So the world communication topology shaped by software was oriented in the way that you have a desktop and you tune into some streams, some feeds of information or whatever and then you get it on your screen on your ears and your speakers and so on. So, I've been quite concerned about this and then I wrote an application that is called Nun. I was doing a radio online with some friends and this application is done to do radio on the net, to stream audio and so I'm maintaining back now a more recent software that I'm also developing in Montevideo which is called FreeJay and I will show you how I package it together into this distribution.

This is my desktop, I'm going to show you the development version I have on my computer. This is the command line interface and I'm opening this parenthesis - I asked Matthew to print out this - you can have a look there it's a nice article about how command line can be your friend. We talked about it today and it's a very interesting topic I think. You shouldn't be afraid of the command line, there are a few things that even users used to graphic interface can learn that can speed up very much your work. Just imagine you can do with command line things like, ok, you recorded 40 hours of audio and you have a hiss in your audio, like your camera was doing hiss or your tape recorded was doing this high frequency hertz, something that sometimes happens and then you can, with a command, you can take out the hiss out of the whole 40 recordings, without opening them picking them up and closing is the best script capability for instance. Sorry, one moment...

So that's how does it look like Dyne:bolic. I have to show it to you from my hardisk because my CD, by trying and trying and developing it got a bit broken so I can't really boot it. Yeah, laptops break really easily. I have some copies of this CD and you can take them, they are for free and last day I asked Matthew to copy a few so, you can try it out. I bring them specially for people that cannot download them and so, if you have bandwith, yeah leave it to those who cannot. But I guess here, much people have bandwith. It's a software made for running on old computers specially so, my focus at the beginning was to recycle machines like 200 mghertz, 300 mghertz pentiums that you can just boot a CD in, they don't even have to have a hardisk and they still work and they can work as a browser, as a text editor and with a couple of features that you want to have on your workstation and basically does it. It runs even on a Pentium 100megaherts and the minimum requirement is 64 megabytes. With that you can have, well not all functionalities but almost all. I mean for video and audio you need some machine that is pretty faster or a bit faster. But, ok you have video, a couple of tools for different tasks and audio as well. I'll go back to audio because there is something interesting going on in Linux right now, then there is image, where we have some tools like the Gimp. Ok, we love it. But then there is also Jah Shaka, it's like a very interesting image integrator it's called. But yeah, it was presented here (at v2), I saw it here presented actually yeah you was also here Artem, and yeah is going on pretty fine. I think he would be presenting it also in how is it called? The Wizards of OS? in Berlin? Ok, and the CD also has Blender by Tom Rosenthal, who is also from Amsterdam. So, what else to mention because... you have text, Scribus, I put it into the category text (lol) not image, or not like 'publishing'. But, Ok desktop publisher it's called. Then there is Everyword, than can open and can save .doc files. Restart which is a very nice old editor for rich text files. I have 20 minutes so I'm going very fast through that but just to show you the basic functionalities. It looks old and is running like with a footprint of 2 megabytes on your memory so even if you have like a really like a slow computer you can still use it for editing rich text files. Then for the net, there is a whole lot, there is some browsers, e-mail, with encryption which is very interesting. I will go back also there later on. For file sharing, peer to peer, chat programs, also encrypted, FTP and remote desktop control like VNC lets you have like two miles on the same desktop it can connect you remotely on a desktop and stuff like that

I've also some programs to burn if you have two CD's you can burn a CD. I have a lot of games because games are important and there is also a world atlas which is basically a program to browse the information of the CIA information about the whole world, it's something that gives you data about... oh wait I went to the wrong place. It shows like where are the airports and you can click on one point of the map and then you get like description of the place and it's an interesting database and probably you saw it, is called the Fact Book, which is also free, so I included it. It might be interesting to know sometimes what is written inside besides this it gives you also the possibility to hear the national hymn of places like clicking here... in midi.

And the games are mostly multi-player oriented. And I tried also to pack for audio specially and video, programs to produce, as I said, to come back to this topic, I mean...I would name the topic like 'free software for liberation' or 'the liberation of media consumers' or something like that.

I mean if you want to listen to a Real stream like they are doing it now online our friends, you would have to download a free player. So there is no problem there, it's free. It's gratis I mean, it's gratis. But if you want to produce the stream you have to pay royalties. That's the problem so this brings a topology which is very similar to what we have right now on radios and what we have right now on televisions and being Italian, I tell you that's a real problem.

So, in fact if you really have a buyer that is economical on sending out some content, then you have monopolies on that, very easily or at least you have capitalists of information of knowledge like we have now. People that enter everybody's house with a media that is spread, like television online now, like it could be the computer in the future, it is really huge here in Europe and yeah, they have the monopoly on information there. So it's very important to have softwares then that implement the functionality of communication in different topologies not only talking one to many but also many to many.

Coming back to a question that John Hopkins was putting on the IRC at the beginning... he was asking what is the connection between code and reality then. I rather define code as a meta meta language. Because if we see that in fact a program lets us speak between us. So in a way a program shapes and determines the way we speak between each other, because for instance, encryption, I mentioned encryption earlier. If I know that I send you a mail and I know that only you will read it I will write different things in that. If I know that someone else goes in here I will write in different things. I'm talking about a sentimental level, you know it's really intimate. If I speak in IRC and I know that many people is in that room I will, kind of wear a different personality. So it shapes the way we communicate, but this is software so it's a meta-language. And software itself is written with a language that is code and there is a kind of osmosis that goes through all this process that I really cannot identify because I am in one end, on the other end sometimes, because I don't compile code myself so, anyway, it's interesting to see how it conditions it. So, my concern about doing software for communication was basically not to leave these arbitrary decisions to companies that have only commercial interests.

I don't have any commercial interests in doing what I'm doing. I'm printing out CD's, yeah maybe they ask you a fee for coming here but yeah, I don't really have much interest in selling out the CD's on the street. But companies, yeah they produce a software if they see there is a market for it and they are not simply doing a software, they are deciding the way that we are gonna work on it.

Like the discussion before was like if a programmer wants to change the software, he might also not want and what is his criteria for that? The community around himself?. Most of the time the open source is like that, like enthusiasm but sometimes it's just money and that can be also worked out for a single person, but when a company is a monopoly in that field, that company will be not so easy to convince... to produce things that are maybe not so marketable but still one needs it.

For the programs I wrote I show you some of them. One is Muse, I have this simple interface that August Black once helped me to formulate, in fact I wasn't really able alone to envision it and we need ergonomists, and we need designers into the open source software, free software world.

This is like, yeah you have a playlist, you can add files to it, (oh gosh I closed...). You can open files inside this, now I don't have really a music file into this development... You can open up new channels, mix them together, you can put your voice on and then you click here, it's a bit hard to understand but to...we have a tutorial of 4 pages you can go through it and then you fill this with your data about the server you're going to stream and you can do radio online. This program was very much used for doing radio on IRC channels, like on chat channels, that was its origin, but now is even getting a new interface and I'm working together with Adam Hyde and Eugene a new programmer that joined the project for a scheduler. So we want to do like a radio scheduling program, all integrated. A program that can schedule just radio programs and then you can step in and put your voice on, so basically the idea is to make a kit for people to do a radio.

As a host for this program at the beginning. I did this distribution that was distributed in 16 megabytes of RAM of space, so it was much smaller than a CD and I wanted to have it as a CD that you can boot on a machine and have a radio right away from there. That was very handy and coming out of the experience of many media centers in which I also contributed some work, some activity, like in Florence or Genova or so. As well I have my favorite users which are Palestinians actually I was also doing a project in Palestine. I took Dynabolic with me and I really wanted to see what was the situation there for employing a software like this. You can see the project, I have to go online, oops! Sorry I opened the... This is a particular suite that came out, I'll tell you about...

You can nest Dyne:bolic so you can save your data and your settings on a USB key or on a hard disk inside one file. It's one monolithical file. It's really not invasive. It's a CD you put inside your computer, you take it out and it's out, you still have Windows if you had it before... it works only on PC right now. And you can save your data on a USB key, so you can have basically no laptop but still you have your mails and your important data on a USB key and you have your CD and you are sure that you're finding back everything the next computer you find.

I didn't have a laptop before this. This was a present so actually I really leave it in that situation of not having a easy way to carry my data and to have everything I needed installed right away in the machines I found. Right now it's really handy to me, to myself... I mean, in this way. And then you've network configuration, what I really wanted to do... This are small applications I wrote.

With Glade is a rapid application development system that lets you like drag buttons, put them in place and then you write the code back - but basically it's very good to design interfaces. I just put like a text dialogue here and so on. And if you're interested in using Glade I'm doing a workshop on Friday in the ASCII in the evening.

So I will continue here... ok, sorry for this. So, I can open an alternative browser here like links here for instance... it's a very nice browser. It's a bit different from Mozilla but actually it runs with 4 megabytes of RAM. So if you have a very slow computer it pops up like fast, like you saw it. It does quite good rendering. It lacks just two things...it lacks what was it?. HTTP authentication, when you have like the pop ups from the web sites to authenticate, this is not supported and it lacks also cut and paste which is quite important, of course, but for showing web sites it's very good. It's made by a Hungarian and is developed since two years, I think, it's a really good alternative HTML rendering engine.

So that's the web site of Dyne:bolics where you can download it. I just wanted to show you how also it went far.

Dyne:bolics, I had it as an independent effort, really like I started doing it and I started giving it to friends and basically one of the first places where we employed it was in Florence in the Social Forum and, at the media center of the Social Forum and then it grew and people just started using it so I had it distributed.

Actually in India it was quite popular, the 'Daily Hindu' also put an article and it was distributed by PC Quest , so yeah PC Quest in India and then was distributed in Italy 2 times, the next month it comes out in Greece or in Czech Republic, yeah it's been distributed around. I don't get any money for that, it's just free. I mean, I'm not talking that language anyway. I'm not concerned about that.

I didn't do Dyne:bolics for doing a business so I'm also trying to play with that and try to make it kind of provocative in certain ways because I see many projects in the open source and free software world that they are just looking for business, where to survive, I'm not looking for that, I'm just looking for something else like a different statement and therefore you can see sometimes this flag. I call it Rasta Software, which is software really for the liberation.

Yeah, this is basically my webpage where you have all this software... and, I'm a very spaghetti flow, I hope you cope with that. Ok, this the web site, Muse is included in Debian, and the freeSD and the netBSD and Gentoo... don't get angry Calum. No, because you left the Gentoo development. That could be an interesting topic as well. Yeah, it would take a seminar itself it would raise a lot of issues there, I'm quite concerned about that but I want to underline some issues about it, of free software that I thought about. And they were quite arising so I'm not gonna bring anything new here.

One is interoperability which it was mentioned, before. Interoperability is very interesting on Linux on GNU/Linux and is basically a major point where for instance in the professional audio world GNU/Linux is getting better than proprietary software. Interoperability on GNU/Linux is much higher because there is no competition, on free software there is no competition. Why should I be competitive with another developer, I share code with him. Niels Elburg develops VJ, I develop a program that is called Freejay, they sound similar, they do something similar. They have very different approach but we are really good friends and when he comes over to Amsterdam, we smoke good grass and we have good ideas about our software sometimes. And so, just to let you understand that people once they are interested in efforts that are similar and not really competitive about them, so this way brought, the co-existence...

I'm going far and far on this direction, of a system for audio programs under GNU/Linux, that is called a Jack. This system can run like that in background, you have the daemon, it tells you that your computer is doing well and you have connections. Then you can open audio programs like, you can open a synthetizer, something that you connect to your midi, it's a free software and i put it inside.

(Opens software.) I have a problem... I wonder... I am sorry I have a problem, that's a real time...a real life...yeah I'm showing you the development version so I'm sorry if... this is strange... actually. But is more a hardware problem so I will shut down... ok, you can open the program basically and you will see a popup with the list I showed you before... A list of programs that could cope with Jack and I think that also Artem is going to show you later on... and you see the programs popping up in that list and you can connect it them together.

So basically you can connect in and out of audio programs, all together. So basically what you had, what you said was really right, the Unix philosophy is to do one program, little program, that what it wants do do does it well. You are not a company, you don't want to cover the world field, just do what you need. So in audio it is. You have many audio programs, they do their little thing, there is a synthesizer, there is a sequencer, but when you can connect them all together, like that, then you have a rack then you have a full music studio, full good audio quality actually with Jack. So that's the point in which the professional audio, right now we are seeing it , it's really like getting better under GNU/Linux than on Windows where you have like Cubase USD that doesn't cope with another program like Cool Edit or... yeah you can only save a file but Jack is about piping through, it's real time stuff. So basically that is one of the issues that I'd like to highlight.

Another one is about new options of interaction, you mentioned it also a bit, like about, yeah it changed your aesthetics. Might be good, might be bad but it did. And basically when you have a program you operate it, it decides how you operate that program, isn't it ? So I'm going to show you my desktop. On Linux you've a lot of desktops you also showed them. Was it you? no, was Richard... Graham

So that's an image that can evocate you my emphasis on console. It's a very nice approach sometimes and it leads you to... it leaded to new partings of interaction with certain programs, I can show you for instance a program that I use for mail it's MUD, it's really fast program actually for...oops, for reading mails is just text. It has threading you can select things like, I want only to see messages from people starting with 'T', you can save things around and so on... and is very much compliant it doesn't execute any attachment, it operates completely from the keyboard and it has much more performance than other programs and you can access it from remote.

All Unix geeks use this program, it's really amazing how it's like a standard, it became a standard because is one of the best, when you learn it, is the best. And you don't change it. And it is with editors, that I use EMACS there is a whole world between EMACS and VI, I use both anyway, but I use VI for configuration files and I use EMACS for code (lol) but any... (Calum is commenting something I can't hear) Yeah, I should give you the... but I will not because I like EMACS. And so on, you have programs that got really fast in doing operations and yeah, and they are quite different partings sometimes.

What I wanted to say furthermore is about this philosophy and this why I do it?. I'd like to do two quotes just like, yeah, they are like ideal quotes. One is, just think about it, is a metaphor, I like this book I read in the past by Suskind, I think is Patrick Suskind, 'Le Parfum'. It was very nice and evocating this world, an ancient paradise of perfume makers you know and they have all these kinds of ingredients. They put them together and they made like new brews very beautiful to, yeah you could condition persons, I mean this ends up with an orgy with all around because he formed the best perfume around and people were, really like, yeah... It's a nice book, not because of the orgy, but this makes you an idea of how ingredients are good to have, they are good to exchange, and recipes are very important, so with software and with algorithms and with code we are talking about the same. We are talking about ingredients, and recipes and we are talking about another thing, actually to copy a recipe or to copy an ingredient to pass it over to a friend, I don't want to spend mine I mean I'm not consuming mine that's the new thing. I mean, that's the wonderful thing, I mean imagine if in the 1800 it would have been like that with perfumes, what we would be doing now? I mean, it's beautiful now, it's happening with code, this is what Eben Moglen calls the digital revolution. I really like it, I really enjoy it and I really think it is a revolution. It's a small piece into the anticapitalist revolution I join, in my small being. I have another quote from Andre Breton, 'je ne parle pas logique, je parle generosité'. It's quite a nice simple quote about when somebody asks you, 'Well why you do it that you give it for free?', 'Make money'. I mean this is another language, not only French. So, I think is a small world at the end, the free software world. It's not such a big world. We are not talking about everybody's life and there is a lot of other things and a lot of other problems to be solved beside the proprietary software in this world, but is kind of a small revolution, I mean into this world. This is really something that is an independent effort and started also with a political view, even when, between the lines, even if people were not totally aware and it's really, it's really getting there, even if it's provoking some reaction.

And I leave you with sad news, that today it was the released that the European Parliament, the ministers, they accepted the software patents, so now it's going back to the parliament and is getting discussed. It called some reaction from the business world, it's going on as quite some , yeah some 'telenovelas' , sometimes with Microsoft and SCO and so on but is reaching some moments that the movement, as I see it as a political movement didn't reach it and so I find it really interesting and I'm quite happy to be able to contribute something. Thank you.

Artem Baguinski

Matthew Fuller: Well I think we are running exactly to time so I want to move straight to Artem Baguinski to discuss the V2__Jam project. Also, Anne did you wanna speak?. Ok, V2__ jam I think is interesting again with this idea of interoperability and the integration of media forms, again also because it comes out of cross-disciplinary work between, artists and developers, programmers, and is one of thesse hybrid forms that the seminar has seen lots of... Ok, so we'll head straight over...

Anne Nigten: Ok, I will briefly kind of introduce Artem Baguinski, he is one of the developers of the V2__lab and Artem will present the recent research and development line by V2__ that's called the V2__jam system and as has been mentioned earlier by several speakers today, V2__ jam is based specially on the idea of interoperability. The interoperability between different software applications happens to be a very timely issue especially for those people who are working in the field of digital live performance and who are not exclusively or don't want to be exclusively restricted to a certain type of software which is based in the previous disciplines as we know them in traditional arts. Artem will tell a little bit more what we exactly mean by that, but the ideas of a lot of open source applications are based on existing proprietary software and on the kind of tools which were based on the physical reality. Now we see a lot of people who are using the digital domain as the new area of work which does not have to have necessarily a representative or an equivalent in the physical world. I think all this is a paradigm shift to interdisciplinary or kind of vague disciplines, blurred disciplines, however we could name it, and the work specifically created in the digital domain without the equivalent in the physical domain is the main target we try to address with V2__ jam system.

And also, as I have the microphone right now, I would like to invite everyone here to continue the discussion and I'm interested to see if we can in one of the activities of V2__ later on continue the dialogue between this very interesting group of future users or current users and developers because I think that this is a kind of crucial asset not only to have an online meeting and exchange but also try to figure out how we really can show each other what we mean and how we can improve or foster collaboration in the art and technology scene. So, Artem it's yours...(Hands over mic.)

Artem Baguinski: Thanks, The title of my presentation is V2__ Jam, free as in free jazz and by this title I meant... I would like to try to bring some freedom to the world of digital performing art the freedom that is comparable to the freedom that a musician who has mastered his instrument needs to make free jazz and free improvisational music. First of all I want to tell what is the area that I'm interested in and where I am working in at V2__. It's digital performing art I couldn't find a free definition of this term in the free encyclopedia. I found digital art and performance and tried to combine them: there is all this group of people and computers producing or modifying media before the audience and. The research, V2__ jam research focuses on real time interactive cooperative multimedia tools and I will explain each word in this area. Real time aspect means that we are interested not in tools for producing some objects which then continue to exist as pieces of art, but we are interested in producing a process which is an object of art and this process cannot be exactly reproduced again and again. Every time it's different. Therefore the analogy we see in improvisational music.

The interactive aspect has two side streets, one is interactivity at the performer's side. The performer producing this art piece... interacting with the tools he has with his instruments. And the other side is that the audience can also interact with his piece of art and this interaction is considered by the artist an object of his piece. The cooperative act, again can have different meanings. One is the cooperation between the different tools that the performer, performer users have. Another one is that several performers can act like one whole in producing an art piece and cooperate with each other and they can do it by just some social means like looking at each other and just knowing what the other is going to do, but they can also do it by technical means, sharing some synchronization signals and stuff like that.

Then multimedia, refers to different mediums that are used in this art but I have to say again that it's not these objects like video clips that are objects of art. It is what happens with them during the performances that artists are showing as art. And finally tools the instruments that artists are using to perform.

Then there was a lot of words today about interoperability of software and it's very important for the area of real time digital art, because, for example if you are talking about design tools like Quark Xpress and Photoshop, you are interested in the production of documents, so, you can take your time, work parts of it in Photoshop, collage them together in Quark Xpress and save it into some portable document format and open it in another program that understands that format so, the only interoperability issue there is to have an open and widely supported standard of representation of media. While in real time performance you want software, you want tools that interact with each other, exchange information and control and you want to be able to synchronize their work in real time and there are not many standards and widely supported techniques there. There are some proprietary software and hardware solutions in specific areas like audio processing. There are some open standards in, again in narrow areas, but there is nothing that could help you to produce mixed media art. And that's mostly what we are doing here in the V2__ lab, we are helping artists that are not satisfied with the constrains of the tools that exist. They tried already to create artworks which are documents, with a lot of possibilities, and they feel they are not creative anymore, well they are concerned by this frameworks that they have.

So, one thing you can do to create something new and interesting is to combine some features of several programs, combine several different techniques into some coherent one piece and they don't see the way to do that and they come to the lab and sometimes I get to help them to do what they want, and after they explain what they want I just think and I say, yeah sure it can be done. I just need to do some coding, some scripting (...).

There are technologies available to synchronize different programs running and do whatever they want, but they are hidden. I think, Graham mentioned today that in Mac OS X, there is no bridge between the level of CLI level and graphical user interface level. There are all these Unix technologies that I listed here, like interprocess communication technologies, various scripting languages and lots of libraries to do interesting stuff but it's so hidden from people who are not using command line, not only command line but who are not programming and as, well...

As a little break I will describe a couple of such a projects that I was working on with artists.

First is 'Run Mother Fucker Run', a project by Marnix de Nijs. This is the fake picture that I made to explain what is going on there. There is a treadmill in which a person can run here and there is a sensor that is not shown, that checks where the person is on the treadmill and if you run faster then the treadmill starts to run faster and the video projection that you see here, is also played back faster but not only that. There are also some hidden things, like if you tend to run fast the new movies that you will see will also be faster, and there will be more things happening there than if you are lazy. And you just walk, you will just get boring movies. And there is also a video here showing the person in action. There is somebody walking in the darkness on a treadmill and this is actually the intersection, which is totally fake, is three frames of real movies which are organized by running on either one side or another side of the treadmill. You can turn into this fake intersection and then get into the movie which is real footage, I think that's enough...

I have the same problems with Windows like other people have with open source software... So this is another one of the projects and here is another important aspect of V2__ jam research, I'm not only interested in the so-called 'traditional media' like sound and video, but also in treating the sensory input, as another medium and treating the human body as a medium.

Here interaction of the person running on a treadmill with the system is like a third medium mixed with the video and sounds and this is the other project, featuring me. FloatScanner by Hermen Maat, which will be presenting I think starting tomorrow, in Amsterdam.

It consists of a little boat, inside the boat there is a water bed with green cover in it, and there is a video camera outside the boat and another video camera inside filming. And the person on the water bed and the images of these cameras are mixed together and streamed wirelessly on a shore where other participant can choose to... you see the person inside and you can for example choose some different background from another city, but inside you don't see where you are actually sailing so you would assume that... you just don't know wether what you see it's outside or not. The person inside the boat can interact, he can choose to appear or disappear on top of the background image.

So these are two projects that I made my way, the easier way for me, by coding and scripting and using some different open source softwares that did allow me to mesh them together into a new piece. But this is very far from Free Jazz, because the only thing artists can do is to come to me and ask me to do that magic stuff and he doesn't really have much freedom to experiment and to try different things and to find some interesting and inspiring and create a new piece. So here the accessibility comes in and in V2__ jam research I'm researching what artists want and then in V2__jam development I would like to create a new tool and also new techniques that would allow artists to do, well not everything that a programmer can but, at least much more in the way of combining the softwares together to do some things that they can't do alone and one thing is that artists as well as designers they find graphical representation of everything, easier to understand or to manipulate their text representation and one strange thing about it that part of this problem I find just psychological.

There is a program called Max MSP and a similar open source program, Pure Data which allow you to create different kind of sound synthesis and effects and the only difference between this program and my way of doing things is that it shows all the programming blocks that you can connect together as graphical objects that you can connect using drag and drop and people who use it, they do programming, they learn to do the same things that I do with text but they just see it different because it's graphical, they feel more secure doing something with this and I find important to provide more of such visual cues for them still, which you can do with computer... with different software and specially with Unix-like operating systems where you have a lot of possibilities to combine productions together. The other thing that the graphical programming helps to achieve is hiding irrelevant details like, I don't remember, who said it -I think also Graham - that, 'this piece of code', 'I don't know what it does but it should be here...', and this is one of this things that I as a programmer I also learned to do. There are some magic invocations, that you have to use in your programs and sometimes you don't know what they do, sometimes you do know but you have to repeat them over and over again and they are just irrelevant for artists and they would have distracted him from what he was actually doing. He was creating an artwork but now he is busy with allocating memory and remembering to call free memory and writing it out a hundred of times and that also distracts you from the creative process.

And the most important is... what V2__ jam development would try to achieve is to encapsulate engineer's knowledge into something more manageable by artists, so, yes I know how to pipe one programs input into other program and how to make a long pipeline sort of production working together, but how can I make it obvious that you can do that? I find that the mainstream Windows environment, they don't really help people understand that all programs can be combined because every program has its own window or set of windows and they are all...in your head they are all disconnected like different entities, you don't see. They are all just one big thing running on a processor. Yeah that's what I'm interested in and that's maybe what I would like to get as feedback from designers who are, well-educated I suppose in representing things graphically and pay attention to how we perceive visual information and not just mathematical information.

I, myself treat programs as some mathematical constructs, because of my engineering background but I understand that other people do it differently, but it's difficult to become a different person and start perceiving things differently. So I feel that we programmers, have the ability to create tools that are more usable but we just have troubles with trying to see what's, what people who think different would want and what would be more useful, more convenient for them. I look for inspiration in other softwares and there are some pieces of my presentation that I would skip because they were already mentioned many times like the media exchange, of the open formats and stuff like that and I would go to the Jack audio connection kit as one of the sources of inspiration for me. And that's what Jaromil was trying to show, because this is a windows machine I cannot show it live but I made a couple of screen shots. This is the look of my desktop when I'm trying to play guitar at home... this program has a drum machine, this one is a midi synthesizer and this is a, they call it digital audio work station. All three of them are Jack-compatible and here I show the connections between them so you can see that I connected the synthesizer and drum machine to the audio work station, then the audio workstation to itself so it can mix three tracks, that's my guitar and to generate the signals and then I play it back to my speakers. This is a, I don't like the interface of this interconnecting thing and I cannot come up with something that would be more obvious and so far I'm busy with trying to extend this concept with supports for other sources of media and not so much with interfaces just because I don't have enough experience with making human friendly interfaces.

So, one thing I wanted to try to create is a compatibility with video exchange engines and it should be compatible in a way that you should be able to synchronize the video and audio, but there are some dark details of Jack internals that are not video-friendly so it should be something external and sounds complicated but it makes me excited about trying to implement that.

The other thing is that you don't necessarily have to create different programs as different executables

the solution for many problems can be just a library, but again a library is something that usually users don't see. Half a year ago in Bergen in Norway, me and other video developers, open source video software developers came to the conclusion that it would be useful to create an open standard for video plug-ins and plug-in is a fancy word for a library but with the difference that plug-ins usually have a very specific goal and a very specific interface to how they would interact with their host application. In the proprietary software world plug-ins usually are for some programs like you have some Photoshop plugins or some AfterEffects plug-ins. The difference in the open source world is that we already have a sound plug-in architecture that is common among all the distributions, not all but many...( this was the last slide).

Yeah I still have to mention that we are also doing experiments and developments with all kinds of funny hardware and middleware, by middleware I mean the software that interacts with the hardware and then understands what the hardware tries to say and then passes it to actual software and uses the midi provided by the hardware and yeah, that's about it.