Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!swrinde!zaphod.mps.ohio-state.edu!think.com!mintaka!geech.gnu.ai.mit.edu!rjc From: rjc@geech.gnu.ai.mit.edu (Ray Cromwell) Newsgroups: comp.sys.amiga.advocacy Subject: Re: IAC (was Re: Clipboard (was Re: The Amiga's Future)) Message-ID: <1991Jun11.165524.4720@mintaka.lcs.mit.edu> Date: 11 Jun 91 16:55:24 GMT References: <1991Jun9.005806.18799@news.iastate.edu> <4264.tnews@templar.actrix.gen.n <4332.tnews@templar.actrix.gen.nz> Sender: news@mintaka.lcs.mit.edu Organization: The Internet Lines: 86 In article <4332.tnews@templar.actrix.gen.nz> jbickers@templar.actrix.gen.nz (John Bickers) writes: >z> <43@rypty > The most important part of it in my opinion, though I do recognise > that in an ideal environment more of the available resources should > be sharable, such as process space, modems, etc. Don't forget AmigaDOS allows you to transparently mount any type of device. File sharing printers/modems is almost the same as sharing files on the Amiga. For instance, using NET: from software distillery you can netmount a remote Amiga's PRT: to net:prt, then on the local Amiga side you can assign prt: net:prt. Even better would be replacing parallel.device with a version that opened a channel on dnet/parnet which then redirected output to a remote parallel.device. Quite easy to do. Netkeys from software distillery allows you to share the keyboard/mouse on your Amiga with another (you can drag the mouse off your monitor onto the other Amiga's screen and activate/launch programs on it.) The Amiga library, device, and handler system is very powerful. > Er, woop de doo. Why do you want several people messing with the > same picture, or same Microsoft Word file, or whatever? Of the > products that one sees outlined, the project management ones seem > the most reasonable. And that's nothing to get excited about. Source > code management has been available as a file system thing for years, > I'd guess. Multiple people diddling with the same spreadsheet makes > it harder for one of them to cook the figures, so it can't be that > handy for management... :) I agree, I think multiple users editing the same data is a _dumb_ idea. Humans don't intermesh as well as computers, and editing a document is not the same as working on an assembly line. Imagine right in the middle of typing a sentence someone alters the sentence before it screwing up the flow of the paragraph. Not to mention real-time updates (instead of buffering) is very expensive and wasteful of a network. Transmitting single characters at a time adds time to the overall transfer since the same packet header is being sent over and over again instead of once for an entire buffer. Anyone ever been on a chat system that sends what the other users type while you're typing your message? Ever get pissed off when -in the middle of a sentence- a new sentence bursts on the screen interupt yours, scrolling your screen, and interupting your train of thought? That's what this whole thing reminds me of. As you say, the _bulk_ of networking is file-sharing and device-sharing. The next level after that is servers and clients. Future networking will include sharing process-space and ram. Infact one Amiga network can already does this, AmigaNet(tm). Here's a quote from their ad: "Ethernet version 2 compatible, Fully IEEE 802.3 A dnd B compatible" "10 Megabaud transfer rate up to 64k buffer memory" "DMA sequencer with 16 megabyte address range" "DMA addressing directly to chip and fast RAM" "16 bit data path" "Every Amiga on AmigaNet may be a client as well as a server" "Allow inter-computer communications with data and resource sharing (Share hard disks, ram disks, serial ports, parallel ports, printers, etc) " "Send messages to another user" "Run programs on remote system" "Excellent recovery from actual disconnection and crashes on remote Amigas" "Ability to know who is active on a network" "Simple installation" "Cables and connector included" It looks as if AmigaNet has the ability to DMA into other Amiga's RAM, or at the very least, its own. This means AmigaNet takes no time away from the processor(only bus cycles are taken away from the RAM bus it is DMAing into, but it still won't eat 100% processor time, on the A3000 it will probably eat about 5-10% of ram cycles) I'm curious as to whether AppleTalk is DMA driven. I bet it uses polled-I/O. If you want to see a nice network paradigm (object oriented and everything) check out the Plan 9 operating system, it's being developed by some of the original Unix and C hackers (Kernigan, Richie?, I can't remember, I read about it a long time ago). -- / INET:rjc@gnu.ai.mit.edu * // The opinions expressed here do not \ | INET:r_cromwe@upr2.clu.net | \X/ in any way reflect the views of my self.| \ UUCP:uunet!tnc!m0023 * /