[FoRK] the fix is in? <rant> lang, os, etc. continued </rant>
Ken Ganshirt @ Yahoo
ken_ganshirt at yahoo.ca
Sat Dec 12 10:29:08 PST 2009
--- On Sat, 12/12/09, Sean Conner <sean at conman.org> wrote:
> It was thus said that the Great Jeff
> Bone once stated:
> > ...[snip]...
> > Random thoughts...
> Random thoughts indeed.
> Here are a few of mine. ...
Thank you for that, Sean. I can comprehend that. If it's at all related to Bone's Babblings, it helps. I grok assembly language and *nix. I mostly get VAX and AmigaOS. All stuff, or similar to, that I've worked with. My serious developer days are so far behind me that I generally have to infer about 100.3% of what Jeff goes on about.
Of course maybe that's him, not me....
My affair with assembly language predates VAXen a little bit. I did a lot of development on PDP-11s running RSX (a realtime operating system), Macro-Assembler and DecNet. That combination didn't give you anything quite like QNX but it was sure easy to do interprocess communication both on and between nodes. Initially we were using it to build SCADA systems, process monitoring and some early crude network management for telco networks. But we even built a major order entry system with it. First mechanized order entry system in the world for telco service order entry, tracking and scheduling.
Doesn't seem like much today, but it was sure the cat's ass in '75. We did demos for visitors from telcos in the US and Europe and even a delegation from Japan. We were the first DEC customer in North America with an actual fully-implemented fully-networked non-SCADA production system on DECNet. (Doing networked SCADA systems on DECNet was trivial.)
We were also incredibly naive. We didn't learn until years later that it's impossible to structure and reuse assembly code. So we just did it. I guess our extensive Macro-Assembler code library was just a figment of our imaginations. DEC's Macro made it so easy to use library code that, after we built the first SCADA system, all subsequent ones were more like playing with Lego than programming.
RSX Macro-Assembler had similar capabilities to the ones you described for CallG on VAXen and the message passing on AmigaOS (actually it took a combination of RSX Macro and DECNet to get that). Not quite as simple but one did not have to make a large distinction (mentally) between making function calls and message passing. Back in the day we mostly just refered to it all as interprocess communication (yeah, I know that's gross and inaccurate but it worked for talking about it in such an early un-networked environment as it was.).. didn't much matter what device the "other" process was on, either.
For communicating with/controlling non-computer devices, when you have to write all your own device drivers it's pretty easy to make them more homogeneous, and appear more like just another process, than if individual vendors do their independent proprietary thing. Which makes treating it as simply function calls or message passing much easier, too.
And, when you also have to roll your own file system, that all makes for a very nice homogeneous environment to play in. So when I first saw technical decriptions of the AmigaOS, it was sort of, What took you so long? Yeah, ours was primitive by comparison but, logically, I saw AmigaOS as evolutionary, not revolutionary.
Sorry about the trip down memory lane. It really is sort of relevant, at least as far as the deja vu component of these discussions go. *nix is still a central and significant, or at least useful, OS today. I am still reasonably functional with it. But that's not where I learned it. In my SCADA development days I also worked with another DEC operating system, RT-11. After becoming fluent with it and RSX, learning *nix -- and CP-M and MS-DOS -- was really just refresher courses. Something old, a little bit new, something borrowed, something.... Well, you get the picture.
Just wanted to thank you for using "language" I could understand and relate to. If it's even fractionally related to what Jeff's rambling on about, it's nice to have confirmation that some of the headaches I've got from reading and inferencing him has not been entirely in vain.
Looking for the perfect gift? Give the gift of Flickr!
More information about the FoRK