Upgrade: Changed the title and the end part related with hash tags.

Two weeks ago, my friend Iván Pulido and I, started doing some experiments about open publication for the web. The idea was to put a static blog site, made in Nikola with IPython support, inside the web2py powered website of our local hackerspace. We share some raw documentation in several places (in Spanish).

Because we used portable, self-contained and easy to learn technologies for this experiment (web2py, fossil, IPython, Nikola), this can be put inside a USB thumb drive or a low cost rasperry pi alike computer. And it is easy to imagine some not so distant scenarios, where the people will be able to upload, using a webform, some IPython file to a web2py + nikola + fossil powered site, that they can take with them to work off-line/on-line and write together in a distributed fashion. And with the upcoming support for real time multiuser writing inside IPython, is even possible to imagine a full spectrum of collaboration on data narratives that covers single/multi user + on-line/off-line writing, computation and visualization on several subjects, from academic publishing to citizen science or data journalism.

Here, we're scratching our own itch a using some self-made incomplete solution and I would add dirty code, in the sense of not having yet good programming practices. At some point, we started to think with Iván, which would be the hashtag to share the process and partial outputs. We came with the ugly and kind of silly (when you try to pronounce in Spanish) tag #OpenPubWeb. But yesterday I was reading and writing for my PhD middle term exam about the Indie Web, and I thought that what we have done had this kind of indie spirit in the sense of selfdogfooding and also owning your data and your infrastructure for publishing and using free open source software and formats for that.

In fact this idea of low cost portable technology is important in the Global South, where Internet connectivity and computing resources has much to be improved in terms of quality, quantity and availability. And despite all the hipe on Big Data, I think that there is a lot of important work to be done here on Slow Data that can be supported by this kind of solutions. This "low tech" solution is also able of running in your garage hardware or in your neighborhood. Of course, big S Science, doesn't like to be a garage phenomena, in contrast with, lets say rock music, that shows some kind of proud when becomes indie garage rock. But for me is important giving the tools and places to your neigbor to get involved and for that, if we want to build a future open science with more participants beyond the universitary knowledge elitism, low tech portable infrastructures and open spaces like hackerspaces are important.

Some people said that we need some kind of GitHub for science. I disagree with them. Places like that tend to build monocultures (for example around git as a DVCS). I think that we need more like a BitTorrent for science where different implementations like the ones explored/proposed here, can talk with the ones wich are more visible in the discourse of Mozilla Science Lab or Software Carpintery. For that, protocols and metadata will be more important in enabling interoperation between the different approaches, but, following the advice of the Indie Web movement:

• UX and design is more important than protocols.
• Use visible data for humans first, machines second.
• Build tools for yourself, not for all of your friends.
• Build for the long web
• Have fun.

so we will start with the human experience part of this experiment for ourselves (including having fun!) and then we will build the more robust and technically oriented part of this.

Finally I thought that #IndieSci could be a good hashtag for this kind of approaches, that also try to cover another narratives beyond what the "Science", with the big "S", is occupied with, but after using that hash tag I see that is also related with the bad CammelCased tag #IndieSci-Fi for indie sci fiction, and in fact indie science is broader that the thing I'm talking here, that's why I corrected the title so I prefer and prefer for now #IndieWebScience (or #IndieWebSci). May be #IndieDataNarratives could be a better hashtag to cover things like data journalism, that despite of being centered on facts, is not Science in the traditional sense, or maybe #IndieWebDataStories is short enough and conveys the message.

What do you think about all this?