Date: Mon, 20 Feb 2012 22:53:10 -0800 From: Doug Hardie <bc979@lafn.org> To: Chip Camden <sterling@camdensoftware.com> Cc: freebsd-questions@freebsd.org Subject: Re: /usr/home vs /home Message-ID: <2585F37E-7B1F-4E76-8925-838B40C0F4DE@lafn.org> In-Reply-To: <20120221062011.GE6294@libertas.local.camdensoftware.com> References: <4F3ECF23.5000706@fisglobal.com> <201202210910.20658.erich@alogreentechnologies.com> <20120221052603.GC6294@libertas.local.camdensoftware.com> <201202211240.53859.erich@alogreentechnologies.com> <20120221062011.GE6294@libertas.local.camdensoftware.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On 20 February 2012, at 22:20, Chip Camden wrote: >>> I believe the 5MB removable were RL01. They also had a 10MB = removable >>> RL02, which we used for software distribution. We resold them to = our >>> customers at $170 each. >>=20 >> yes, this sound familiar. The RL02 came later. >>=20 >> I think that tapes were much more common for software distribution = those days. >>=20 >> I still remember the responsiveness of RSX-11 even compared to = FreeBSD under all circumstances. Real time is real time. >>=20 >> Erich >>>=20 >=20 > Oh man -- we wrote process control software in Fortran-77 on RSX-11M = to > automate our software distribution processes. That was the best! = DECNET > to communicate between systems. RSX-11D was slicker than greased lightning. Used it for a number of = systems. The first 30 pages of the kernel source were the = documentation. The description of every table and the values for every = field. What each module did was documented at the top of the module. I = made numerous improvements to the kernel most of which were adopted by = DEC. However, it was nowhere near a fully featured OS. It was quite = bare bones. Great for real-time requirements. There was a guaranteed = maximum time that interrupts were disabled and it was very small. We = interfaced a number of instruments to it and none of them ever saw any = delays. Most of them automatically fed data to the computer. There was = no triggering of that. The instruments just pushed the data. The RK05 had one removable platter in a plastic housing. It used a = voice coil movement mechanism that had to be aligned every week or you = would lose your data. It didn't hold much and was quite slow. We used = those at first but the system couldn't quite meet its performance = requirements. I still have one of those platters on my wall at home. = Departure present from the unit. That particular platter had a head = crash so the remaining oxide had to be sanded off to sanitize it. The = timing side is out with lettering on it now. We used 4 RK05s in one = rack and each was mounted as a separate disk. The controller was single = threaded so you couldn't get any performance improvement with creative = disk assignments. We switched to 5 platter drives RP04s which were extremely reliable and = didn't need frequent maintenance. They also ran much faster than the = RK05s and held more than 10 times the data.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?2585F37E-7B1F-4E76-8925-838B40C0F4DE>