Chapter 6: Developments

Gateway States

Late July – Early August, 1982

Back at Uni for the start of second semester, it was a mix of déjà vu (the introductory lectures for Computing I and the programming clinic) brand new things (the three subjects I was taking this term), and more of the same (back into the thesis project). This was now the third semester I had given the introductory Computing I lecture; I had it all down pretty pat. We had a large number of students wanting to do the Programming Clinic; I am sure most of them heard from the second year students about the benefits they had got from attending the clinics. We had to find a slightly larger lecture room for the clinic; during the mid-year break Michelle had made the necessary arrangements for us to use the other lecture room on the fourth floor of the building. In fact, it was right next door to the room we had originally intended to use.

The three post-grad subjects I was enrolled in looked interesting; Communication Theory was a pretty esoteric subject; intended for those people undertaking research work in communications engineering. Digital Communications would give me some good theoretical background to help me talk with engineers from Telecom, and elsewhere in the ARPANET environment. Microprocessor Systems was an attempt to try to stay abreast of what was becoming a rapid developing area of the information technology field.

With the thesis, the immediate task was to determine why the packet throughput using the EGP software was much worse that we had predicted. I suspected the problem was in how I had implemented the state machine protocol that was defined in the RFP; I looked through the state transitions and actions that had been defined, and it appeared I had implemented the design correctly. But something wasn’t working correctly, and the only way I could determine what was going on was to add another level of diagnostic logging into the router code; one that would record every state transition, including the details of the event that triggered that transition. I added those changes to my routing software, and lined up a time to test the software again with Patrick.

The performance of the gateway was much as before; despite using a different model of the LSI-11 with faster memory, there was no improvement in packet throughput. I started to go through the logging information that the test run produced; and I was surprised at the number of state transitions from the “up” state to the “unknown” state, which was a state between the “up” and “down” state”. I looked at the events that were triggering the state changes, it appeared to be that because I was using a digital X .21 line, rather than a traditional analogue V .24 or V .35 line, there wasn’t a direct one-to-one correlation between the analogue control lines, like DCD and DSR and the equivalent digital control signals. But, I couldn’t still see the point behind the ‘unknown’ state in the specifications, all it seemed to do was force an unnecessary neighbour reachability / acquisition process. Time for me to check with the people who came up with the RFC.

I spent a day working out just what I would put in my email to the two designers of the protocol, Eric Rosen and David Mills. I didn’t want to come across as some ignorant amateur who had no idea of what I was talking about, or being arrogant criticising the design from these two very experienced and knowledgeable network engineers. I ran my draft email past David and Claude, making a few changes before I showed it to Doctor Dunworth. With some trepidation, I sent the email off to them, copying Patrick as well.

I was pleasantly surprised when the next morning I had extensive replies from both of them. Eric apologised for the state diagram still having the ‘unknown’ state in it; apparently that was a carryover from their initial draft, and they had decided that there was no need for that state. Eric was also very interested in the work I was doing, he was working on something similar using the LSI-11 as the platform. Eric sent me a revised RFC, it was due to be released in October, once it had gone through a few more reviews, but he assured me that the state diagram and processes in this would be the final version.

It took a few days to update my EGP software to implement the new state design; but it appeared to work much better; the packet throughput exceeded what I had expected. I sent details of my results back to Eric Rosen and David Mills; they confirmed that I had successfully implemented the exterior gateway protocol. David also provided me with information on the ‘Network Time Protocol’, a method that I could use to synchronise the hardware clocks on all of the connected computers back to a reference atomic clock.

With the EGP software written and tested, it was time to turn my attention back to writing the management configuration software. I had worked out a programming model for all of the different configuration and management tasks to share the same framework. I would be using the ‘curses’ programming library for managing the menus and data entry forms, each configuration function would update a particular section of the main configuration data file, and would then call a program that would read that section of the configuration file, and set values and restart some of the daemons or other services that implemented particular implementation functions. Those same programs would be run as part of the router start up process.

I also maintained regular correspondence with David Mills; not just about the EGP gateway routing software, but on our individual router design projects. He was doing something similar to me; using a LSI-11 as a dedicated router, but rather than using a stripped down version of the UNIX operating system, he had developed his own compact operating system, called ‘Fuzzball’. He was impressed with some of the design decisions that I had made, and I looked at some of the things he had implemented in his system, too. What was interesting is that we were both working to solve the same set of problems, and there were quite a few similarities in our approaches. We both felt that we could learn a lot from each other. Having the ARPANET connection, with email and FTP available made it so much easier to communicate and share information with him, and some others in the networking community.

We had a phone call from Fifa, saying that George had the final version of our album ready, and wanted us to come around to listen to it before it went to the pressing process. She also wanted to show us the proposed album artwork, and said they were thinking of changing the album name slightly to “Capriciousness”, so the cover artwork would better fit that name. Since none of us had strong feelings about the name we had proposed; we didn’t have a problem with the minor change. One night near the end of July, we all went around to the Alberts studio; George played the master tape. It sounded very good, I wish I knew just how they did their post-production ‘magic’, but they were very good at it.

We loved the album cover; I don’t know who did the actual drawings that were an integral part of it, but it was like a childish cartoon of us on stage; all doing silly pranks. Me pulling off Allison’s skirt, Traci was pouring a cup of water over Fiona; Andrew was poking Phil in the back. Every time we would look closer at the picture, we would see some other funny scene. We all thought it was a great cover, we gave them the go-ahead to get the album released. They would plan for the release party at Chequers on August 23rd. That week, being the mid-semester break, we would be doing gigs every night, supplementing our usual Thursday and Friday night gigs.

We had also been asked to appear in several gigs at the Hordern Pavilion that 2SM was promoting for their ‘Rocktober’ festivities. It was suggested we appear with AC/DC, Cold Chisel, Divinyls, and maybe another Australian group over several nights. It all sounded pretty good, and we were keen to get involved.

After everyone had a chance to look at the mock-up of the new album cover, I pulled Fifa to one side.

“Just a quick question, the advance that Ted organized to finance my house purchase. How am I going paying that back?” I asked her.

“Oh, don’t worry about that, with the success of the last few albums, plus the tours, that should be paid back pretty quickly,” she said. “I can get you a statement if you really want, but you really don’t need to worry about that at all.”

“Well, that’s all good, but I never quite understood why you did that for me,” I said.

“Oh, it’s pretty simple,” she said. “We wanted to make sure you, I mean you as the group, still had access to that great studio you have above your garage. Not only does it make it easy for us to record your albums; as you know we are very tight for recording studio space here in Boomerang House, but you have that great area for rehearsing and developing your songs. And besides, we felt that after everything you’d been through over the last few years, getting that house as your own would be a good thing for you.”

“Well, I am very appreciative, it was a great thing that you did for me.”

“No, it’s just a way for us to return the favour to you, to show you just how much we appreciate everything you’ve done for us.”

To be honest, I really still didn’t understand why Alberts were so generous towards me; but I wasn’t going to look the gift horse in the mouth. If they wanted to give me an ‘advance’, and somehow have that reduced even month, I wasn’t going to complain.

Lisa and I continued to make progress on the garden beds in the yard; where we had dug the cow manure in the soil was looking good. The worm farm was starting to produce a good supply of ‘castings’, once we had a decent amount we would dig that into the vegetable garden. Some of the tomato seeds had started to produce seedlings, it was still too early to plant them out, but they were looking promising. One Saturday morning, we drove out to Dural and called into some of the nurseries, and bought a bunch of shrubs and plants, wholesale. We also got several trays of flowering annuals that would be planted along the front path.

Now that we were both back at our studies, our life together was starting to fit into a pattern. We would have breakfast together every morning; and I would call her several times during the day and we would talk briefly. She would be home most days when I got back from Uni; with the exception of Wednesdays, which was her night for lectures at uni. There was something very comforting about getting home from work, walking into a house that I owned, and hugging and kissing Lisa. We would talk about our respective days, the things we did, problems we had to sort out, and things we wanted to do that evening. After dinner, we would generally go downstairs to the study and work on whatever homework or projects we had to do.

Lisa’s subjects for this semester including Introduction to Accounting, and another on an introduction to the Australian legal system. Mary Beth offered to help her with the accounting subject; the course that Lisa was doing was pretty similar to the undergraduate degree that Mary Beth did; expect that Lisa planned on concentrating on industrial relations and HR management, while Mary Beth’s degree majored on accounting and financial management. But the core subjects they did were pretty much the same; even though they were at different universities, they still covered much the same material.

Considering all of the people we had in the house and the education each had done, we had two technical writers (Murph and Wendy), a network engineer (Patrick), an Australian historian (Fiona), a business manager (Mary Beth) and a computer programmer / network designer (myself). Looking further afield to the other people in the group, we had an architectural draftsman, two carpenters / cabinet makers, a pharmacist, two teachers, an accountant, a nurse, and two computer programmers. We certainly had a diverse set of skills and education among us.

During August a long stretch of wet weather hit us; Lisa and I weren’t able to get more work done outside in the yard, but at least I thought the rain would be good for the ground, and would really soak in. Instead, we turned our attention to the rooms downstairs; I constructed some shelves to contain the wine bottles that we had purchased on our Newcastle trip. In the utility room, between the wall to the darkroom and where the pabx had been mounted on the wall I stacked up some besser blocks, using some epoxy resin to glue them together. I had them aligned so that the holes in each block faced each other horizontally, and ran lengths of two by four between the two stacks of blocks. That provided shelves to rest the bottles on; and it worked out so that the gap between the two lengths of two by four would allow the ‘shoulder’ of the bottles to rest in, making sure the bottles were angled so that the corks were kept wet.

Not that we were big wine drinkers; but it was convenient to have a variety of bottles available, if someone suggested that a wine would be a good accompaniment to dinner, it was easy to go downstairs and select a suitable bottle. At the rate we drank wine, what we had would last a year, we were lucky if we had more than a bottle a week. But at least I could say ‘let me get a bottle from the wine cellar’. Pretentious, sure; but it was fun.

I wasn’t completely happy with how the utility room looked; along two walls were the air conditioning unit, the hot water heater, the main breaker cabinet and the pabx, pus my wine shelves. Sure, that’s what the utility room was meant to be used for, but it wasn’t all that attractive. I had gone up to the Ikea store at Gordon to get some storage cupboards, at least that provided a place to store things like the vacuum cleaner, broom, mops and other cleaning utensils, but there had to be a way to improve things.

Wendy suggested putting up floor to ceiling curtains that could be drawn across to hide the utilities; the A/C, water heater, and all that. It seemed like the ideal solution; I didn’t want to have to put up more internal walls. We measured up for the fabric, she got to work with her sewing machine while I attached a set of tracks to the ceiling. Once they were hung, it did improve the room quite a bit.

The next room to work on was the family room; initially we had Patrick and Wendy’s lounge room furniture – a sofa and two single seat chairs – in there, along with a small cabinet with their stereo system. Lisa had some ideas of how we could decorate the family room; she wanted something that was informal, and a contrast to the rooms upstairs.

“What if we set this up as a room that seems inviting to just sit and relax; maybe read, or listen to some music?” she said. “I was thinking of a coffee table, and some side tables, a bookcase, and some simple pictures on the walls. Eventually, when Patrick and Wendy move out and take their furniture, we can get a sofa and some other chairs, but until then, we can start on the accessories.”

She had found some items that she liked at a second-hand store; we had gone looking around places over at Rozelle and Balmain; the style she said she wanted was ‘mid-century modern’, these were typically Scandinavian designs, characterised by simplicity and functionality, with clean lines. I had no idea of what she was talking about, to me furniture was just that, something to either sit on, or put things on. But I liked the items that she had selected; she said that most of the Scandinavian designs used light coloured timbers, like ash, but she wanted something darker.

We found a suitable coffee table, two matching end tables, a magazine rack and sideboard in a darker timber, the sales assistant said it was Tasmanian Blackwood, a nice golden brown colour. She also picked up some lamps; a floor standing lamp with two adjustable fixtures, and two matching lamps that would go on the end tables. At another store she found a bookcase that matched the timber used in the other items; so with the trailer loaded up, we headed home.

Once she had finished reorganizing the family room, it did look good; comfortable and welcoming. Together with Patrick we moved his stereo system and record collection from the old cabinet he had into the new sideboard, Lisa placed the lights around the room, and we put some books in the bookcase. It was now a very comfortable and relaxing place to sit, listen to music, and read.

A few days later when I came home, she told me to come down to the family room. She had got two large framed prints of old racing yachts, taken by an English photographer, Frank Beken. These had been taken at the end of the last century, and were quite large, at least three feet by two and a half.

“Do you want these up on the wall?” she asked. “If you don’t like them, I can take them back.”

“No, they are fantastic! I’m amazed at all the details, look at all the people on this one!”

“Certainly more things on there than with ‘Koolong’,” she said. “I thought they were interesting photos.”

I got a step stool, some picture hooks and my hammer, and soon had them up on the wall, either side of one of the fireplaces. I liked how it finished off the room; sure, the existing sofa and chairs didn’t really match the style of furniture that Lisa had bought, but that would only be a short-term situation. When that time came, we could get some lounge chairs in that same style to replace the Tarpey’s furniture.

Time is Getting Tighter

August, 1982

I was reviewing my emails one morning, and I noticed one from David Mills. That wasn’t anything out of the ordinary, we communicated every few days, I would give him updates on the router project and he would offer suggestions. This email was suggesting that I include a “Network Time Protocol” server on my local network here. He had been working on the problem of synchronising the clocks for all of the computers on a network, using atomic clocks run by the US National Bureau of Standards as the main reference source.

He had sent what he was proposing as the initial implementation of this protocol, and he had given me access to the source code repository for the UNIX client and server software. David suggested that I set up a dedicated machine as the local NTP Server, and for it to synchronise to the servers that he had set up in Boulder, Colorado. I would set up the NTP client software on all the host systems here at the university, having them sync their clocks to my NTP server. He had tested the algorithms on trans-Atlantic links, and they worked, he wanted to try even longer connections, hence the request for me to set up a server here.

I would build the client into the router software, and add the configuration tasks to the management software. Once the main AARNET had been rolled out across the country, I would envisage each institution having its own NTP server, which could link back to a main (or several main) Australian reference server(s), which would in turn be synced to the master reference servers in the US. It took me a few days to set up the software on the proposed NTP server; I used a clone of the base router software, but removed most of the ‘router’ components, installing the NTP server software, modifying the initialization scripts, and setting the server configuration file to point to the IP addresses of the master NTP servers in Colorado.

When we had everything set up, it was time to test it. I set the hardware clock of the server to about a minute behind the local time; I decided that the Telecom ‘talking clock’ was good enough of a local reference. While the main accuracy of their time keeping system wasn’t anywhere near as accurate as an atomic clock, the system in the Telecom Research Labs were good enough for what we wanted to do. We dialled 1194, and put the phone in my office on speaker.

“At the third stroke, it will be ten forty five and thirty seconds ... beep beep beep.”

All good; we were about seventy seconds behind Eastern Standard Time. David had told me that the synchronization software as currently written could handle an initial time error of no more than ninety seconds. I started up the NTP server, and had it set to show its log to the console screen. We saw it establish the connection with the server in Colorado, and the message exchange as the time synchronization process took place.

“At the third stroke, it will be ten forty six and fifty seconds ... beep beep beep.”

I entered the ‘date’ command, and we compared the output of the server’s system clock to the local time standard – we were within a second of the actual time.

“That’s very impressive,” Doctor Dunworth said. “Had I not seen that myself, I wouldn’t have thought it possible.”

“The next step is to install the client software on all of our servers here, and have them sync to this server. I’ll install a second server here, as a backup, and they can tic tack with each other, too.”

“What do you think, once you’ve go all the clients installed, of writing up a brief paper for the Journal of the Australian Computer Society?” he asked. “Not a full blown research paper report, but they have in each issue some short articles, as an introductory report on some ongoing project. We would have to float it past David Mills, of course, so he didn’t think we were trying to steal his thunder, but what you’ve done here will be of interest to a lot of people. Maybe we can call the paper ‘Managing System Time on a Network of Computers’, or something like that. I’ll get you a draft format for the paper, and you can start working on that one.”

That did sound interesting; and I spent the rest of the day configuring the client software on the various PDP-11 system in the Computer Science Department. That worked just as well as the main server software did; and by the end of the day, every computer on the Electrical Engineering portion of the University network was synced to my time servers. Before I left for the day, I sent an email to David, telling him how successful the experiment was.

Over the rest of the week, I installed the NTP client software on the remainder of the University’s PDP-11 computers and the various routers on our network; the only systems that weren’t synced in were the Control Data Cyber 72 and the IBM 360 in the main computer centre. The University’s system programmers were unimpressed with some ‘upstart’ trying to tell them how to do their job of managing their systems, I suspected, plus there wasn’t a suitable client at the moment for those systems.

Doctor Dunworth had contacted David Mills regarding me writing a paper on our work for the ACS Journal. He had no problem with that, provided we gave him the appropriate recognition for the development of the software and the algorithms that were used to account for network propagation delays. I looked at the suggested format for the publication report; writing a two or three page paper wouldn’t be too difficult a task. I could call on Wendy, Chris and Mary Beth to help with the editing and proof-reading; they were all experienced writers.

I prepared a set of instructions for how some of the other system administrators on the fledgling AARNET could implement NTP on their hosts; they would sync to the servers I had set up at UNSW. That would reduce the network traffic on the satellite link to the US, and cut down the number of requests to the main tier one servers. Within a week, all of the systems on our state network had their times synchronised, first through my servers here, and through them to the master time reference source in Colorado.

I was pleasantly surprised at how easy it all was to set the time synchronisation up; everything worked exactly as David has said it would. He was happy that the algorithms on calculating the network propagation delay worked correctly over the Pacific satellite link; he didn’t need to change any of the underlying formulas. I had sent him a copy of the installation instructions that I had written for the other local sysadmins, and he suggested I work on the user documentation for the client, and the server, that he would include in the final product. I told him I would work on both of those documents; and said I would try to have something for him close to the end of the month.

So, as well as writing some user installation, configuration and operations documentation, I also had to write the short paper for the ACS Journal. Writing technical papers wasn’t my strong point, but I did share the house with two technical writers in Murph and Wendy, plus I could always get Patrick’s input on it as well.

The mid-semester break came around quite quickly; as well as a week’s break from lectures, it was also the scheduled release of our latest album. As before, this was being held at Chequers, in the city, and like all of the other release events, it was an evening of glad-handing industry people, other musicians, and various hangers-on. All very tedious, but a necessary evil. At least we weren’t like some people, who treated everyone with disdain; you couldn’t really bite the hand that feeds you, not if you wanted to stay around in this business.

After the usual speeches, we performed a few of the songs from the album, then mingled with the crowd, spoke about some of the songs on it, and mentioned where we would be doing gigs in the next few weeks. At least I was able to introduce Lisa to some of the people I knew in the industry; disk jockeys like Bob Hudson and Ian McRae. Eventually, the evening fizzled out, we packed up, and headed off home.

“That was a fun evening,” Lisa said as we cuddled in bed. “I love the way you introduced me to others as your girlfriend, you make me feel all warm inside. Sometimes I feel like I’m the luckiest girl in the world, I know that sounds so corny, but that’s how I feel.”

“I’m the one who feels lucky,” I said. “It’s been a long time since I have felt this happy, this content with my life. I owe it all to you.”

“Let me show you just how happy you make me,” she said, swinging her body up and over my face, positioning her pussy over my mouth.

She lowered herself so that her moist lips were touching my mouth, I licked and sucked at her, making her come twice. Once she was wet enough, she moved lower down the bed, took my prick in a hand and guided it into her. She rode me slowly and gently to another orgasm before I came inside her, my hands holding her hips and pulling her body down hard onto me. When I had finished, she rolled off my body, curled up next to me and we drifted off to sleep.

“Can I ask you a rather personal question?” Lisa asked me.

We had finished the week of promotional concerts with a big night at Selina’s; early sales figures on the album were starting to look promising; orders from the major record and music stores were higher than what Fifa had predicted. We had managed to sell several hundred copies at gigs during the week, as well.

“Sure, you know you can ask me anything.”

“Well, I can remember you telling me how back in your first year at Uni, you and Lori started going out together, seriously, after you were run over by Janelle’s mother. When you moved into your first place together, how did Lori get you to agree to have Megan move in so you were a threesome?”

“Hmm, there’s no quick, simple answer to that,” I said. “I guess it goes back to the end of forth form, when I first started going out with Cathy. It quickly changed from just Cathy and me as a couple, to including Lori and Janelle as a foursome. Sure, things chopped and changed around over time, but I guess that was my first exposure to the concept that being in love with someone doesn’t imply exclusivity or ownership.

“Then after the end of high school, when some people were starting to move away, and I was in denial about the fact that Cathy and I would probably break up when she moved to Canberra, there were several others that wanted to have sex with me before they left – including Janelle. Lori spoke to me, saying I should do it; and gave an interesting interpretation of how multiple partners would be okay.

“But the events you’re talking about, with Megan moving in, that came just after we had been living together for a few weeks. Janelle was going to come down for a week and stay with us; I misinterpreted what Lori was intending, and said that she would be able to sleep on the sofa. Lori got all upset, and we had a good talk, and she told me that she intended for Janelle to sleep with the two of us. She told me that I kept confusing love and sex, and how having sex with someone else wasn’t really cheating. And from there, we had Megan move in with us, and then things just got wilder; Jillian, Allison, others.

“I’m not sure that it was a good decision; but we were both young, and inexperienced, and I certainly didn’t know as much as I do now about making relationships work. But I suspect Lori wanted to have the freedom to sleep around with other guys as well as various women, and that way she was able to experiment, and do a bunch of stuff with other people.”

“So, did she have sex with other guys as well?” Lisa asked.

“Yes, at one stage she asked me if I would be okay with her doing that, which was fine by me. I mean, how could I have sex with other women, but not allow her the same freedom? I didn’t ask her who she was sleeping with, sometimes it was another woman, and there would be the three of us; or four with Megan, but what she did was her business. As I said, I think it was all a way for her to experience a lot of things that otherwise she wouldn’t have; she would often come up with some pretty wild things for us to do.”