3 ms·
> This is why the earliest social experiments in what we would now call “open source” – at SHARE and DECUS – were restricted to individual vendors’ product line
by e12e 13y ago
> This is why the earliest social experiments in what we would now call “open source” – at SHARE and DECUS – were restricted to individual vendors’ product lines and (often) to individual machine types.
Yes, because BASIC and Creative Computing magazine never happened.
Conflating the idea of sharing code, with the idea of sharing system level code doesn't really help this essay IMNHO.
As for the "creation myth" -- I guess that refers to stuff like: http://oreilly.com/openbook/freedom/ch01.html http://oreilly.com/openbook/freedom/ch01.html
I'm not sure how this is supposed to be a "myth". As far as I can tell we went from relatively open systems (if not free) to much more closed systems. And now, after the rise of various standard platforms (x86, sparc, arm) we're in a strange (and a little scary) in-between place where we have oodles of great hardware, half of which we have no real idea how works, except for binary driver blobs.
I guess I just don't understand the motivation for tying together the fact that we've come a long way in terms of system and language design, and that because of that we've been able to get some pretty solid open systems -- and that closed systems aren't (and never was) a bad thing, because, you know *bsd and linux and stuff?
And there's of course stuff like: http://stephane.ducasse.free.fr/FreeBooks/BlueBook/Bluebook.pdf http://stephane.ducasse.free.fr/FreeBooks/BlueBook/Bluebook....
While later than C, in some ways arguably more open than unix (as it was actually published and distributed).
- chalst 13y agoWrt. early social experiments: there clearly was a well-developed code-sharing community by the mid 1960s dealing with scientific computing in Fortran, with textbooks offering incremental refinements of programs to solve problems of shared interest, and there was some portability to such code, within constraints. But I think esr is talking about the emergence of a much more fluent kind of sharing, that needed much deeper portability and automated configuration, and much more direct peer-to-peer sharing of code, of the sort that the emerging connected networks that were to become the internet offered. What you are talking about resembles the early Fortran sharing more than it resembles the Unix-centric sharing. The motivation for the tying you talk of is that it is this kind of code sharing that has really liberated hackers from the grip of proprietary computing. The Fortran and Basic -style of sharing still left hackers with considerable lock-in to the computing environment they were currently using.
- e12e 13y agoWell, I can sort of read that into the article too, but it doesn't actually make that argument coherently nor explicitly...
- al2o3cr 13y ago"> This is why the earliest social experiments in what we would now call “open source” – at SHARE and DECUS – were restricted to individual vendors’ product lines and (often) to individual machine types. Yes, because BASIC and Creative Computing magazine never happened." Claiming late '70s work as "earliest" does sorta miss when SHARE and DECUS were founded - 1955(!) and 1961, respectively.