Andrew> So whilst the GNU project was announced in 1983 I've heard Andrew> many examples being cited of the source to software being Andrew> shared prior to this, E.g. IBM's VM up to vN.N. And indeed Andrew> recall reading somewhere that up until a certain point in Andrew> time the source to most if not all software was available. As Andrew> vendors made their money from hardware sales and you were Andrew> unlikely to be able to buy a clone or build a machine Andrew> yourself, and so safe-guarding source code was not a major Andrew> concern. I'm not sure if that's really accurate. To take some examples I know: DEC distributed its PDP-11 software in binary form when possible. Some had to be in source form -- BASIC-PLUS components of RSTS, for example, or source modules that had to be assembled on site to customize things for a particular installation. ("Sysgen") But apart license would cost you a whole lot extra (like 5-10x the binary license price). And in any case, no matter what the form of distribution, everything was subject to restrictive licenses and copyrights. Some 1960s era software is open because it's in the public domain -- apparently not intentionally but because of an error on the part of the company. CDC's COS (Chippewas OS for the 6600) and IBM OS/360 are examples, as I understand it. This happened because the sources were made available and were not marked with a copyright notice. At that time (but not any longer) US copyright law required notice for copyright to exist -- lack of notice meant release into the public domain. But those things happened only occasionally; by and large documents and code were marked with copyright notices. I don't know examples older than that. Another angle would be non-vendor software -- either academic, or third party commercial. Then there are of course the various user groups and their software libraries -- DECUS is a well known example. That's somewhat like open source though not necessarily; for example, contributions might have restrictions on them like "not to be sold, rented or leased for profit" (quoted from COSAP V1 -- in the DECUS library, from Lawrence University). Others might be closer to open as we now understand the term, most likely BSD-style rather than GPL-style. I suspect a lot of software was shared case by case informally, especially in the academic community. X writes something, Y hears about it, asks X for a copy, gets a tape (or box of cards) in the mail. paul

I don't know, but it's not quite as simple as closed/unavailable vs open/available. For example, back in the '80s I was at a research lab in academia that ran VMS. We had the VMS source on microfiche, and indeed I used it to write patches for command-line editing (this was before DCL grew command-line editing of its own) - indeed, my patches grew big enough to break PATCH; I had to split my patch into two patches. But I would never say the source was "open", especially given what that generally means today. /~\ The ASCII der Mouse \ / Ribbon Campaign X Against HTML mouse at rodents-montreal.org / \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B