A Long, Incomplete, and Mostly Wrong History of Free Software

November 30, 2009 10:00 PM

In Two Bits (full-book in html) , Christopher M. Kelty investigates the history and cultural significance of Free Software, revealing the people and practices that have transformed not only software, but also music, film, science, and education. The author encourage his readers to modulate the book.

Sadly, Two Bits has not been endorsed by Richard Stallman. There have been reviews, notably in MIT's Technology Review and Inside Higher Ed. It also has provoked interesting reactions from readers, including this story about the early Korean Internet, and this (title-inspiring) comment from Thomas Lord. [via LtU]

posted by Monday, stony Monday (16 comments total)

(I do realize Kelty is an anthropologist and not an historian -- the title refers to this)
posted by Monday, stony Monday at 10:03 PM on November 30, 2009 [2 favorites]

From the MIT article:

"how to create similar communities elsewhere--around "movies, music, science or medicine, civil society, and education"

These things are already open source. They exist in a form that humans understand, made with well understood practices. The difference with software is that it can be distributed in an opaque form. Fanfiction, spinoffs, sequels, the progress of science through history all attest to the fact that these things speak for themselves and can be picked up in some way by another person where they were left off. Compare this to what happens with software - if it is abandoned by the original creator it is either opened up by sharing the source, or it stagnates and dies.

The widespread usage of proprietary file formats in the last three decades is a travesty on a scale that I don't think many people fully realize. Not to mention that the way many arts are now done, the skills of the best artists are dependent on closed software, leaving a huge swath of potential human creativity at the whims of a small handful of corporate entities. Of course Adobe, for example, is not run by some Grinchesque villian who wants to deprive the world of art for its own sake, but there is still something odd in seeing how much of our cultural products are now dependent on that one company.
posted by idiopath at 10:54 PM on November 30, 2009 [2 favorites]

<3
posted by unknowncommand at 10:55 PM on November 30, 2009

idiopath: "They exist in a form that humans understand, made with well understood practices. The difference with software is that it can be distributed in an opaque form."

Actually, there's a difference between the music you get on CD is the opaque form. The studio recordings, or the sheet music which describes the work are better formats for making changes. Which is why it was interesting when Reznor offered Garage Band files for a few songs to be remixed.

If you think this point is pedantic, I can dig up a post about people trying to reverse engineer the opening chord to Hard Days' Night.
posted by pwnguin at 12:22 AM on December 1, 2009

pwnguin: sure, but how hard is it to play a song by ear, and how does that compare to the difficulty of reverse engineering an API? A song is designed to be interpreted by a human listener, while a program is designed to be interpreted by a computer. While the music by itself is not always perfectly transparent, it is still quite a bit more self evident than a binary executable is.
posted by idiopath at 12:36 AM on December 1, 2009 [1 favorite]

If a modern computer can interpret it, it can be interpreted by a human. And done fairly quickly with training and tools. Arguably, it's easier to demonstrate a reverse engineering of an API correct than which instrument was mixed into the opening chord. But if you chase this too far you'll end up discussing things like Turing tests rather than Free Software; if you play a song by ear wrong and nobody but the Beatles can tell, does it matter?
posted by pwnguin at 12:55 AM on December 1, 2009

The LtU post is really interesting.
posted by delmoi at 2:39 AM on December 1, 2009

If a modern computer can interpret it, it can be interpreted by a human. And done fairly quickly with training and tools. Arguably, it's easier to demonstrate a reverse engineering of an API correct than which instrument was mixed into the opening chord.

This is spoken with the brash confidence of someone who has never had to parse the text out of pdf files from several different sources. There is a reason why google scholar is so incomplete.
posted by srboisvert at 4:11 AM on December 1, 2009 [1 favorite]

"how to create similar communities elsewhere--around "movies, music, science or medicine, civil society, and education"

These things are already open source. They exist in a form that humans understand, made with well understood practices.

That may be what "open source" means but it's not what "Free Software" means. You can't alter most movies and a lot of music and you can't get access to the data for a lot of science and medicine.

Free Software doesn't just mean you can see what's inside the box. It means that what's inside the box is just as much yours as what's outside the box. The only restriction is that if you pass the box on to someone else, you can't lock it first.
posted by DU at 4:40 AM on December 1, 2009 [3 favorites]

Free Software doesn't just mean you can see what's inside the box. It means that what's inside the box is just as much yours as what's outside the box.

Love this. I would add that because its yours, you have the opportunity to help change it if you find it lacking.
posted by localhuman at 5:54 AM on December 1, 2009

Love this.

Consider it copylefted.

I would add that because its yours, you have the opportunity to help change it if you find it lacking.

Which you just did!
posted by DU at 6:59 AM on December 1, 2009 [3 favorites]

You could also argue that playing some composition by ear incorrectly is in itself a form of collaborative composition, the principle creative modality of folk music.

Which of course would imply that there's a whole lot more folk music out there than you'd generally think...
posted by tspae at 9:16 AM on December 1, 2009

I'm still reading it; so far, the most interesting bits have been chapter 4, which is mostly about the early days of Unix, and how its source code would circulate despite the wishes of AT&T (which wanted to keep it a "trade secret"), chapter 5, which deals with the fragmentation of Unix in the 1980s, and the advent of TCP/IP (over the OSI protocols -- but that part is short, and I can't help but feel that part of the story is missing), chapter 6, which recounts the early days of GNU, and the reaction of Stallman to the commercialization of Gosmacs, and chapter 7, which is mostly about Linux (there's even a treatment of the BitKeeper issue).
posted by Monday, stony Monday at 8:35 PM on December 1, 2009

Oh, and one of the most interesting aspect is the discussion of the changes in the "intellectual property" of software that was brought by the 1976 and 1980 copyright laws in the US. My understanding of it is this: prior to 1980, software wasn't really copyrightable (although many organisation would put copyright notices on their code). So you had to rely on something else to "protect your IP". In the case of AT&T, it was trade secrets; that's why they didn't want the code to spread too much.

But new copyright laws completely changed that, and it seems it was the issue of using some Gosling copyrighted code in early versions of Unix Emacs that led Stallman to found GNU.
posted by Monday, stony Monday at 8:49 PM on December 1, 2009

Monday, stony Monday: "it seems it was the issue of using some Gosling copyrighted code in early versions of Unix Emacs that led Stallman to found GNU"

Emacs started as a collective pile of scripts to extend teco and make it more usable. Gosling went and copyrighted and sold his version of emacs, based on this collective effort, neither consulting nor crediting the rest of the community that worked on them. This was the motivation for inventing the GPL. Simultaneously, the community was using Gosling code, from before he copyrighted it and he was just asking nicely that people not redistribute it.

It was as much a question of Gosling going and copyrighting their code without permission, and selling it, as it was the community using Gosling copyrighted code.
posted by idiopath at 9:03 PM on December 1, 2009

Yes, but from the book, it really seems like the inclusion of display.c, and ensuing discussion, that were the trigger that got Stallman going.
posted by Monday, stony Monday at 9:50 AM on December 2, 2009