I have narrowed down my subject options for a masters thesis (New Media at Universiteit van Amsterdam) and landed on a software studyhttp://en.wikipedia.org/wiki/Software_studiesof TeX through the lens of ConTeXt/LuaTeX (the idea being to document the cutting edge). For instance, the many materialities of a TeX document, the fact that it exists as a source file, an evolving memory heap during compilation, a ready-to-print document, and (often) as a print document. Part of my thesis, and a good reason to investigate ConTeXt, is exploring the material aspects of texts generated specifically for on-screen use/how an interlinked PDF differs from a website/etc. Naturally, a materialist analysis leads to discussing liberational potential, and the entire assemblage of TeX will be examined in light of Benkler's 'virtuous software'. (See Benklerhttp://www.nyu.edu/projects/nissenbaum/papers/jopp_235.pdfor mehttp://mastersofmedia.hum.uva.nl/2009/11/01/git-virtue-github-and-commons-ba...for more.) More, I'm sure, will arise. The second half of the thesis is an operational component: setting up an online collaborative web application (with Ruby Waves http://rubywaves.com) that is able to interweave git, reST/markdown/HTML (or a new one of my own that I convert to ConTeXt), pandoc (unless option 2), and ConTeXt to craft a system capable of outputting anything, not the least of which a PDF with layout and typesetting specific to that system (in this case, a new open, post/pure peer review journal of new media studies). I feel this must be possible because I've read that some of you folks are running highly automated typesetting jobs with ConTeXt. However, it occurs to me I should discuss the feasibility of at least the operational component with you folks before I end up trying to develop something impossible come March. The theoretical component relates to the operational component in as much as it seeks to justify it--this part is basically saying 'typesetting matters' but in the sense that a good typeset will a) make the journal feel much more established and allow an easy means to produce an on-demand print version, and b) liberate the information in as much as it is available in both PDF and machine-parsable ASCII, along with everything in between. Hmmm, maybe that _is_ taking a bit on.. Especially sense I don't know much ConTeXt at the moment (for instance, not using them on my final papers this semester). However, I will have significant time to dedicate to this thesis (all day every day starting March), and I've become something of an obsessive about TeX and typesetting, despite not having a lot of free time to dedicate right this second. For instance I find myself reading the Mark IV documentation (perfect for a software studies, btw!) rather than other, perhaps more pressing things. I guess I'm just sending a ping out to see what kind of response you folks have to such a project. I don't feel as if anything is set in stone, per se. The final proposal presentation is yet to happen, so I have time to make changes. Sincerely, John Haltiwanger