Saturday, 1 February 2003	Echunga
		Top of page
		next day
		last day

My satellite gateway machine froze up again this morning, after only 20 days uptime. I'm sure it has something to do with the tcpdump I keep running; it seems that the resolver library has a file descriptor leak, and I end up with thousands of UDP connections to the name servers. Even after stopping the tcpdump (which gets rid of the connections, anyway), the system doesn't come back. Probably time to install a more recent version of Linux.

Apart from that, more unproductive work today. It's really frustrating to spend a whole day working and have nothing to show for it in the evening. Work on my Daemon's Advocate article went on; due to circumstances I can't control, I'll have to hold off publishing it for a day or two. Also managed to get reduce my mail to a level which I haven't had for several weeks, and did more preparation for publishing the book.

Sunday, 2 February 2003	Echunga
		Top of page
		previous day
		next day
		last day

Some discussion on the NetBSD mailing lists today about the perceived bad performance of RAIDFrame, and consequent discussion of the possibility of porting Vinum to NetBSD. Checked the things I had to do: finish my Daemon's Advocate article, attend to stuff on the FreeBSD Core team, work on the seminar slides for NOIE, and do another run through my book. Decided to leave all that and work on Vinum instead, which at least left me with a slight sense of satisfaction, though I wasn't really able to produce any sources to base NetBSD on. My own master sources and the stuff in the FreeBSD source tree have diverged quite a bit.

After lunch, relented somewhat and did some work on the other topics. I need something more interesting to do.

Monday, 3 February 2003	Echunga
		Top of page
		previous day
		next day
		last day

Today the other shoe dropped. What we've been discussing and planning for over a week: we removed Matt Dillon from the ranks of the FreeBSD committers. It's not the first time, but this time it looks more final. Hopefully we made the right decision. Anyway, it's interesting that the feedback generated less mail than another thread on the accuracy of the rand() function. Maybe we have our priorities together after all.

In the afternoon, teleconference with NOIE, Constructive, but tiring. The good news, which we really need, is that we have another week to finish our slides. Things are beginning to take shape; I'm no longer so worried.

Apart from that, I was so drained that I couldn't do much else. Tried to read some magazines, but even that was too tiring.

Tuesday, 4 February 2003	Echunga
		Top of page
		previous day
		next day
		last day

It had to happen. Somebody forwarded the story of Matt Dillon's demise to Slashdot. Not surprisingly, they got the wrong end of the stick. I wonder if we (core) are doing the right thing by not releasing the whole story to the world, but we tend to agree that that's not fair to Matt, and it would certainly draw out this whole incident beyond what it already is. Wrote a followup explaining what I could, and correspondingly got flamed again by people who had absolutely no idea what was going on. It's not surprising that the majority of the slashdotters are on Matt's side: most of them are even more difficult socially than he is.

Also finished off the Daemon's Advocate article, which was on the same subject and thus had to wait until the matter was public. Didn't do much else, but fortunately there's not much else to do apart from catch up on some important mail messages.

Wednesday, 5 February 2003	Echunga
		Top of page
		previous day
		next day
		last day

What a pain these NOIE slides are! Spent most of the day working on them, and still didn't come to anything I was happy with. To add insult to injury, fired up sydney with the copy of Microsoft “Windows” 98 that Dell insisted on delivering when I bought it, in order to check out how the Microsoft version appeared, and was promptly hit by some Microsoft exploit which caused it to send off scans on port 137:

Spent some time trying to work out what was causing that. People suggested a “virus scan”, but the version that was delivered with my machine didn't find any problems. You'd think a ps would show something, but of course Microsoft doesn't have stuff like that. People shouldn't be allowed to connect such toys to the Internet. Put a complete block on exploit's address on both firewalls, so it can't do any harm outside, and there's no other toy box inside the firewall.

Between that and the Dillon affair, didn't get anything very constructive done. Tried to work on Vinum, but wasn't able to concentrate. Frustrating day.

Thursday, 6 February 2003	Echunga
		Top of page
		previous day
		next day
		last day

It's just not letting up! I haven't done any real work this week, and I'm flat out in catch-up mode on my mail. The Dillon affair is gradually dying down, but it's pretty gradual. The NOIE slides kept me busy all morning, and at the end I spend a bit of time considering how to do them quickly in groff. I suspect that we would have had them ready a long time ago if we hadn't been bogged down in toy presentation packages.

In the afternoon, the first programme committee meeting for AUUG 2003. That looks like it might be off to a good start.

Friday, 7 February 2003	Echunga	Images for 7 February 2003
		Top of page
		previous day
		next day
		last day

And the work carries on. Admittedly, no work on the NOIE slides today, but the work had left me with a big mail backlog, and the Dillon affair kept us going. Round lunch time things were beginning to look better, and then David Newall sent a message asking me if I would help him debug some wireless networking problems. He came out in mid-afternoon with a Linksys access point and network card, which he had purchased very cheaply. He had been trying to get it to run on Microsoft, and of course it was impossible to tell, so I plugged the card into sydney, which recognized it immediately, and found the network, but we weren't able to communicate. Spent quite some time both with that and the access point. Tried installing the card under “Windows” 98 on sydney, which claimed that the firmware was out of date and updated it:

After some time David was able to get the card to talk to air-gw. It really won't work if you set it to “infrastructure mode” (BSS mode). You have to set “ad-hoc” mode (IBSS mode). That really puzzles me, since under FreeBSD there's no difference in the settings. I wonder what we're missing.

In the meantime, Yvonne had a phone call from a distraught Essey Deayton, who was having a domestic quarrel with her friend Doug. Over en masse to see how we could help, and in the end brought Doug back to spend the night with us. Looks like there's going to be some change there in the near future.

What a week!

Saturday, 8 February 2003	Echunga
		Top of page
		previous day
		next day
		last day

Gradually things are getting back to normal. Yana off to the University of Adelaide Orientation Camp in the morning, and Yvonne first to do Yana's babysitting for her, then brought Doug back home and “guarded” Essey while he packed up and headed off North.

Not much to do on the work front. Caught up with a bit of mail and did some work in the garden. It'll still take me a while to wind down.

Sunday, 9 February 2003	Echunga
		Top of page
		previous day
		next day
		last day

Another quiet day. We're not out of the woods yet, but no NOIE slides to worry about, though it looks like I might end up with other things to do while I'm in Canberra. No fighting neighbours either. Instead, spent some time working out contents for an AUUG CD of OpenOffice. Yes, I hate it, but that doesn't mean it's not popular. We'll give CDs away at the seminar. It's a pity that the OpenOffice people didn't go to the trouble of writing a Microsoft installer for the thing. That will make it more difficult to install than Microsoft users expect, which won't help to sell it.

Apart from that, did some catchup work, though plenty more is needed. Also did some more work in the garden. Another week or two like this and I might almost be back to normal.

Monday, 10 February 2003	Echunga
		Top of page
		previous day
		next day
		last day

Just when I thought most of the problems were over, they got worse. Today was the day for the final delivery of the slides to NOIE, though I'm still not too happy with them. I decided that we had too many speakers, and so delegated my part of the talk (only 4 slides) to one of the others. I'm still really concerned that we'll come across as rank amateurs in comparison with the others there.

In addition was the question of business cards. Con doesn't have AUUG business cards, so had to get some artwork together for that. Since we print four cards per sheet, we can have three others as well. Chose Adrian Close (programme chair for AUUG 2003), David Bullock (secretary) and David Purdue (immediate past president, though he chose the title “Éminence grise” as being more long-lasting).

Apart from that, we also had the issue of the OpenOffice CD-R, which proved to be more work for me that I thought. Finally got the artwork done, sent it off to the publishers, and they managed to change the "fi" ligature into something completely different. sigh.

Tuesday, 11 February 2003	Echunga
		Top of page
		previous day
		next day
		last day

Finally things are coming back to some semblance of normality. Still had tidy-up stuff to do for AUUG, but also had time to get back to my book, which I haven't looked at for at least three weeks.

One of the things that was painfully missing was information on how to set up ghostscript to print PostScript on non-PostScript printers. That's really important: the ghostscript documentation is terrible, and it took me several hours to write up some kind of documentation, despite a relatively intimate knowledge of the subject. By the end of the day, I still discovered I had undocumented command options in my print filter. Hopefully the information won't change too soon. I wouldn't know where to find out.

A couple of people in the NetBSD community have been thinking of porting Vinum to NetBSD. Sent them some best guess sources to play around with. It would be nice if we could get something working.

Wednesday, 12 February 2003	Echunga
		Top of page
		previous day
		next day
		last day

Things are taking their time winding down, but at least I'm not completely tied up with AUUG events any more. One reason is that we got a message today from AusCERT about their conference this year—exactly the same week as we were planning our security symposium. There's no way we can do that, so we're going to have to postpone our symposium. sigh.

In the afternoon, finally got my section on printing with ghostscript finished, and carried on with the preface. Gradually it's beginning to look like a book. I wonder when I'll get the hardcopy back from the printers.

In the evening to the AUUG SA chapter monthly meeting, only the second time I've been there in 6 months. An interesting talk about performance by Allan Packer of Sun. Seems he works for Sun USA out of Adelaide. Now where have I heard of that before?

Thursday, 13 February 2003	Echunga
		Top of page
		previous day
		next day
		last day

Still more work on the NOIE slides today. I'm not too happy with the results, but then, I'm not exactly the target audience either. Also started setting up an Open Source Support Directory for the AUUG web site; NOIE wanted to reference it on their site. As usual, got nothing else done.

Friday, 14 February 2003	Echunga
		Top of page
		previous day
		next day
		last day

It's been three weeks now since I've got back from Perth, and I seem to have done nothing but AUUG work during that time. Things continue to get better, but not as fast as I would like. Today we ended up with more updates to the Open Computing Directory, which took a while, especially since Liz had added them with Netscape, which had completely destroyed the format. There must be an easier way.

In the meantime, building a new kernel on zaphod, which continues to lock up. It's an Abit BP6 dual processor motherboard, and there have been suggestions out about overheating, but I haven't seen any evidence that that's the cause. I'll have to do some more investigation. Also, the other test machine (850 MHz Athlon) seems to have some serious hardware problems, which I think were one of the reasons my Linux testing last April went so badly. I wonder what I can do with it.

Saturday, 15 February 2003	Echunga
		Top of page
		previous day
		next day
		last day

Another quiet day. Finally no AUUG work to do, so pondered the claim I had received from Atul Kabra in India that he had been able to compile Vinum under NetBSD in a matter of days, and decided that I should prepare for testing it here, especially since my Duron machine seemed to have problems with FreeBSD and Linux, and it would be interesting to see if it also happened on NetBSD.

Installing NetBSD 1.6 was easy enough, but of course it's a pretty bare-bones system. Decided that I really needed an install script local to my system for installing all the things I like to have automatically, and spent a bit of time on that. Also upgraded the NetBSD source tree to the latest -CURRENT, and pondered how to build that. It's easy enough to build a system, though it used to be easier, but I'm still puzzled about how to replace the code. The NetBSD people I've asked haven't been too helpful: “Well, don't do that, then”.

Sunday, 16 February 2003	Echunga
		Top of page
		previous day
		next day
		last day

Finally got round to doing some more work on the book today, cosmetic tidy-ups which were interrupted by a lot of other work.

One of those tasks was more work on getting NetBSD built today, mainly in the background. Finally found a description of how to do it in the FAQ; it was only a little wrong (you need a DESTDIR when building the distribution). Then watched while it created successive files in the rescue/obj directory and then couldn't understand them:

It's obviously some kind of NFS problem, though it's not clear at which end: on the NFS server machine there was nothing wrong with the directories. Left that for later and started a local CVS checkout, which was still running when I left for the evening.

Round 16:30 discovered we were disconnected from the satellite network again. Almost without thinking, rebooted sat-gw, but only then discovered that the real problem was lack of signal. Even after it came back, though, I still didn't get any connection, and finally spent half an hour waiting for the help line to be told that the problem was "storms all over Australia", even after I pointed out that signal quality was normal. Under those circumstances it was hardly worth mentioning the fact that I could reach other networks on the downlink, but not my own. sigh.

Monday, 17 February 2003	Echunga –> Canberra
		Top of page
		previous day
		next day
		last day

As I feared, the satellite link was still down when I got into the office this morning. Called up the help desk again this time and explained the situation to somebody who had obviously never heard of routing, and who wanted to know what happened when I logged on to the Internet. When I asked him what he meant by that, he explained that he meant trying to connect to a web site with a browser. I told him that there was little point in trying that while I wasn't being routed, so he asked me what version of "Windows" I was running. Asked to be connected to somebody more senior, and during the interminable wait, tried re-tuning the satellite dish. Bingo! It worked. I can only guess that if you start the system when there's no signal, something goes wrong when it comes back and it's not recognized. You have to start the module when there's a signal.

Off to the airport and thence to Canberra, where Ben Elliston picked me up and took me to the hotel. Had quite a discussion there, then off with Gordon and Liz to dinner in one of the restaurants in Kingston that I frequented during my time at IBM. Back to the hotel to find that Con had been waiting there all the time: he had got an earlier flight, but he didn't have our phone numbers. To the bar to discuss the plan of action, being joined by Jonathon Coombes in the process. Late to bed.

Tuesday, 18 February 2003	Canberra –> Echunga
		Top of page
		previous day
		next day
		last day

Up relatively early this morning to be in good time for the seminar. Despite a couple of problems, managed that, and was somewhat surprised by the meeting room, which was hidden in the basement.

First up was Mary Ann Fisher of IBM, who presented a paper which looked so much like ours that you could almost have thought that we had collaborated. In fact, it was no longer as similar as our first attempts. I have to concede to Con and Gordon that my objections to many of the slides weren't as relevant as I thought, since Mary Ann presented much of the material that I had asked to have removed. Bottom line was "open source is good because it helps to implement open standards", pretty much the same as what we said later.

Next came Maggie Wilderrotter of Microsoft, who put up a good presentation saying not very much of substance except that Microsoft was committed to open standards.

Next were Gordon and Con, who put up our position pretty well. It's interesting to see how people see things differently. Con used Gordon's laptop to demonstrate the "desktop" environment, and in the process had some minor difficulties. While he did it, I looked at the attendees and decided that there wasn't enough interest to repeat this part of the talk next time. Later I was talking to Hugh Blemings, and he volunteered his appreciation for the demonstration and the way everybody watched with spellbound interest. Obviously my aversion to "desktops" is showing.

During lunch, spoke to Maggie Wilderotter and Steve Vamos, the latter Managing Director of Microsoft Australia. I commended them on their intention to pursue open standards (which we had defined in our talk as being maintained by industry consensus and available from more than one vendor), and asked when they proposed to start. Steve said "Ah, but we have started. Do you know about .NET?". sigh. I told him that we weren't too happy with .NET (in fact, it was in our presentation as an example of a non-standard system, but Steve had apparently not noticed that). At this point, Maggie disappeared and I carried on talking to Steve about it for a while. I was left with the distinct impression that Microsoft don't completely understand why we don't consider .NET to be an open standard.

After lunch, a presentation by Peter Gigliotti of the Bureau of Meteorology, also an AUUG member, about what they're doing with Open Source. It's interesting, but not encouraging: much of what they're doing has been replacing proprietary UNIX, so the open standards issue is not touched, and they're still planning to migrate to Microsoft, though it's possible that might change.

Robin Simpson from Gartner came on next with a lively discussion full of numbers and probabilities, but without hard copies of the slides. One of the most interesting things he said was that, with 0.8 probability, the GPL will prove to be too restrictive by the end of 2004. I'll be interested to see if he's right there. He made some other more unlikely claims, including that forked projects can rejoin at a later date as long as license issues don't intervene. That sounds very unlikely to me. Also, he claims that there isn't enough information about TCO over a five year period to be certain that open source software would be cheaper than Microsoft. That sounds rather unlikely as well.

Then another talk about a government project, this time by Tony Ablong of the Department of Veterans' Affairs, describing how they propose to replace their current IBM infrastructure with a zSeries 8000 running Linux under VM in LPARs. It's still vapourware, and I was left with the feeling I'd been in an IBM sales pitch.

Then the panel discussion, in which I took part. This was also the only part where the press was allowed in. The questions were interesting in the sense that they showed where the participants were coming from, and also because they came mainly from non-Government people. The summary, which was also printed in The Australian, was the uncertainty about TCO, quite a shame really. I don't see much change happening as a result of this seminar; if there's more uptake of open source, probably IBM will be the real motivator.

Drinks after that and had a talk with a number of people, including Mary Ann Fisher of IBM and Duncan Bennet of Sun Microsystems. Talking to Mary Ann wasn't that interesting, simply because we seem already to be on the same wavelength. Duncan was interesting, though; we still need to hear more from Sun about what they're doing with Open Source.

To the airport with a whole swag of people, pondering a press release from Senator Kate Lundy, whom we had been invited to visit tomorrow, but which we had to cancel for lack of time. She hadn't been at the seminar, and her press release, which called on the government to accelerate the acceptance of open source in the government, didn't really address the same issues, Still, that might be one of its strengths. I'm not really happy with a certain feeling that the government agencies will throw out Open Source with a tenuous lack of proof of better TCO.

On the same plane with Con as far as Melbourne, engaged in a lively discussion. I suppose the next few days will show whether we've achieved anything.

Wednesday, 19 February 2003	Echunga
		Top of page
		previous day
		next day
		last day

Finally the pressure's off! Not much to do today except to read the 2,000 accrued mail messages (not that much, really, since I normally get over 1,000 a day) and write up what happened yesterday.

A few mentions in the press: The Australian got the details wrong, with Gordon as Chairman and myself (with name misspelt) as President. ZDnet did better, but managed to misinterpret something I said: during the panel session, somebody asked if there would be more closed source software released as open source, as in the case of Netscape becoming Mozilla, and StarOffice begetting OpenOffice. I replied that we hadn't seen many cases, and that the results weren't conclusive yet. The article quoted me as if I were saying that the whole concept of Open Source was not yet proven. Grrr.

In between that, trying to build NetBSD 1.6, and tripped over a number of things because of my lack of experience, but also had some significant problems which I can't see are my fault. If there's one area where BSD hasn't really progressed in the last ten years, it's in the matter of building the system from source.

Thursday, 20 February 2003	Echunga	Images for 20 February 2003
		Top of page
		previous day
		next day
		last day

It looks like autumn is coming early this year. It rained all night, which provided perfect conditions for Telstra's telephone lines to fail again, and they did. As if things weren't bad enough with the satellite problems of a couple of days ago.

Called up Telstra round 10:15 am and spoke to a Kylie, who confirmed that the first of my 5 telephone lines had a line fault reported against it, and gave me the fault number S604036302, without any guarantee of repair time. Considering it took over a week in November 1998, I'm less than happy with that.

Off down the road to look for evidence of work on the line. I didn't find anybody there, but there was evidence of unfinished work on the line:

Image title: repair 1          Dimensions:          1536 x 2048, 896 kB Make a single page with this image Hide this image Make this image a thumbnail Make thumbnails of all images on this page Make this image small again Display small version of all images on this page All images taken on Thursday, 20 February 2003, thumbnails          All images taken on Thursday, 20 February 2003, small Diary entry for Thursday, 20 February 2003 Complete exposure details

 

It's a reasonable assumption that the problems we're having are related to this work. Even if it proves not to be the case, you'd expect linesmen to be out there looking at the problem.

Back home to try to file a complaint. At first the receptionist (Joanne, whom I called at 12:50 pm) was not very interested, but when I mentioned the TIO and the press, she at least offered to connect me to her supervisor. She wasn't available, but promised to call back within 15 minutes.

I really did get a call back, from Rhonda Rankin, whose position is “team leader”. I tried to explain that based on the ongoing work I would at least expect somebody to be attending to the problem, but she wanted to send somebody to our house to investigate things locally. I told her that it would be better to send them to fix the line fault. She said “no line fault has been reported”. I told her that Kylie had volunteered this information hours previously, but Rhonda couldn't find anything in her system. I am dumbfounded. No wonder the service is so terrible if that's the level of communication. She checked a different system with the aid of the fault number I gave her, and found a line fault entered against our fax number only (the third of five lines). What a crock. She did enter a complaint for me, but that would only be handled within 48 hours. While trying to get her to send somebody out to address the problem, the mobile phone connection dropped, and she didn't call back for an hour, by which time she had discovered that yes, indeed, there were line problems in the region, and that the issue was being “fast-tracked”, which meant they should have somebody out by tomorrow morning. Grr.

I had plenty to do without a network connection, but somehow I didn't feel motivated. Spent the afternoon writing up this and related incidents and catching up on some reading.

Friday, 21 February 2003	Echunga
		Top of page
		previous day
		next day
		last day

Pulled out of bed this morning by Dave from Telstra, who wanted instructions on how to get to our place to check out our phones. I asked him if he had fixed the cable fault, but he didn't know anything about that. I suggested he go to the exchange and confirm that a whole lot of phones were out, which he did.

Called up Rhonda on her private number. At least she remembered who I am; maybe not too many people are as persistent as I am. She explained that they have reported two separate faults, because two of my numbers (on a pair gain system, no less) are not in the same cable. I am flabbergasted. How can they possibly get such completely incorrect information?

Down to the exchange, through fog which restricted visibility to 100 metres in places. There I found Dave and another Telstra person, the latter installing a phone. I told them the story of what had happened so far this morning, including the incorrect information about the cable. The other Telstra person, whose name I didn't get, said: “Knowing Telstra, does that surprise you?”. It's interesting, in fact: most of the technical people are pretty good, though there are notable exceptions there too, but the whole customer interface infrastructure is unbelievably bad. Dave confirmed that he had only picked up the fault this morning, nearly 24 hours after it occurred. As Dave also confirmed, they call this approach “fast track”.

It turns out that Dave is a linesman, so he was able to go and take a look at what I had seen yesterday. I didn't hear anything from him before leaving for lunch in Adelaide, so popped in to the exchange on the way and found them both there, along with a third person who was on the phone. They had found the problem, but they weren't allowed to repair it: that has been contracted out to independent contractors, presumably the one to whom the line fault had been reported. They're obviously in no hurry, even though it's a serious failure: Dave tells me that it's a 74 pair cable (strange number), and that he estimates at least 50 subscribers are affected by the outage. I'd put the number higher: not everybody has five lines, and we know there are pair gains in there.

To lunch with the ADUUG in town, at the Lion Hotel in North Adelaide, on the way getting a phone call from Dave confirming what he had told me at the exchange, and that he didn't have a date when things would be up again, but that they would probably work over the weekend. In the hotel, discovered somebody else not in a hurry: it took well over an hour after ordering before we got our food. Liz and Gordon also came to that dinner, making it the largest I can recall, with 17 people.

During that time I got another phone call from Norm at Telstra, probably really a contractor, saying that things would be fixed this afternoon or tomorrow morning. Took a look on the way home, but didn't see any activity, and of course the phones were still down.

Spent the afternoon preparing for tomorrow's board meeting; it wasn't as hectic as last time. The phones were still down in the evening.

Saturday, 22 February 2003	Echunga	Images for 22 February 2003
		Top of page
		previous day
		next day
		last day

Checked the phones first thing this morning, but of course they were still down. Off to Adelaide for the board meeting, which went relatively well. I'm at least getting to the point where we get through the material in the time allotted. Nothing much of interest to report that won't be in the minutes. Here are a couple of photos I took at the “Oostende” pub after the meeting:

Back home, still no evidence of any work being done on the phone lines, and of course still no phones. Spent some time reading Telstra's “Customer Service Guarantee” in the phone book, and note that it's almost useless: they guarantee to repair things in two business days, and if not, they refund $20 per day after that. In addition, it doesn't apply to data lines. If water or electricity were like that, there would be all hell to pay. As it is, most people have mobile phones, so the failure of the voice lines is not serious, but being off the net for days on end is a serious problem.

Sunday, 23 February 2003	Echunga
		Top of page
		previous day
		next day
		last day

Still no phones today, of course. This is really getting on my nerves. Spent some time writing articles for AUUGN, including one about service guarantees for data links. I wonder if that's worth pursuing.

Also more work on the book. It's about time to get the final draft out.

Monday, 24 February 2003	Echunga
		Top of page
		previous day
		next day
		last day

First thing this morning, I went down to see if anybody was working on the cable fault. Of course they weren't. Back home, tried to call Rhonda, but there was no answer. The fault line (132203) requires you first to type in the complete phone number. It then performs a lookup or a test and confirms that there's something wrong with the line. Only after you listen to all that information do you get connected to a wait queue which can take up to 10 minutes of mobile phone battery life.

At 9:30 I retried the call. I still couldn't get any response from Rhonda, and 132203 failed several times due to mobile link problems. Once I fought my way through the menus into the wait queue, only to get a busy signal. The bottom line: “due to circumstances beyond Telstra's control, no time is available for restoration of service.”

Finally got on to Kylie, who couldn't give me any more information and also couldn't connect me to her supervisor: it seems there's no supervisor available on Monday mornings. She tried to get on to the linesmen and find out what the problem is, and while waiting for that I was disconnected.

Called the Telecommunications Industry Ombudsman, who put me on to a higher-level Telstra complaints service. There I spoke to Deena, who made it clear that we're not going to get much help from Telstra. She noted that I should already have been contacted by Kim or Emma, apparently as the result of the complaint filed on Thursday by Rhonda. She considered the fact that I had only had one fault every two years to be good (she thinks “bad” is once a month, and she's seen plenty of those). She also determined that the cable had been damaged by third party influence, and so Telstra is not liable for the service. I pointed out that I had been up the area with Dave on Friday, and that there was no evidence of any damage, and she backed off.

She also referenced the temporary work of which I had taken a photo on Thursday; it seems it's called an E71 (and indeed, that's written on the junction). It seems that somebody thought that could be the problem, rather contradicting her earlier claims.

Finally, it became clear that we were getting nowhere when she claimed that it was impossible to contact the linesmen, in direct contradiction with what Kylie had done earlier. I was left with the distinct impression that she just didn't care. I did manage to get the complaint number 1347683 out of her, but she didn't volunteer it.

At 2:30 pm, Kylie still had not called back, so I tried again. This time I got another supervisor, Paula, who took my phone and fault numbers and promised to call back. In the meantime I called back to the fault number, this time giving my main office phone number. They don't seem to know about any problems with that. Got put on hold again. In the meantime checked the number with a second mobile phone and discovered that the first and second number had no fault reported against them, the third and fifth were down due to a cable fault, and the fourth was down due to a "service disruption". Only 40% correct.

When my call was finally answered, spoke to Hazel. The number I had to laboriously enter didn't show up on her screen. Gave her the first of the five numbers. Diagnosis: Exchange fault. Urgent alarms for the battery. With some difficulty I managed to get her to look at another number (number 3) and confirm that they did know that there was a cable fault. The last known update was at 11:27, and stated that contractors were working on the problem, and that there was no estimated repair time. She promised to follow up and call me back. It didn't happen.

I went down the road to take a look, and there's still no evidence that anybody has been on site. Then over to Diane Saunders' place with echunga to at least pick up my mail before it expired.

While that was going on, at about 5 pm, got a voice mail message telling me that this was a major outage, and that they would have it finished by the end of the week. That would make 9 days down time. This was followed up by a phone call from a Diane at the Telstra Help Desk, who explained that the first contractor wasn't able to do the work because it involved digging up a road, and that they were looking for somebody with the earthmoving equipment necessary. Of course she wasn't able to explain why the second contractor hadn't even been contacted until long after the expiry of the guaranteed repair times. She did, however, confirm that we now only have one fault, instead of the three that we had up to a few hours ago. She also explained why nobody had called back about my complaint: it had disappeared from their computer system. Deena had confirmed earlier that it existed, so this suggests that they have still more computer problems. What a mess.

Tuesday, 25 February 2003	Echunga
		Top of page
		previous day
		next day
		last day

Still no phones, of course. At least Telstra's date of Friday has managed to silence me. Got a call from Diane of last night confirming that they had sent off an interim phone, which should arrive tomorrow, and also that yes, indeed, my complaint had been recorded last Thursday. That poses more questions than it answers, of course: why didn't she find it yesterday, and why hasn't there been any response?

Spent most of the morning modifying sydney's mail configuration to be an effective mail relay. That involved more work than I expected:

• Copy the PPP configuration from echunga to sydney. That was interesting: echunga runs pppd on /dev/cuaa1, the second serial port. sydney only has one serial port, and it's running FreeBSD 5.0 with devfs, so the node /dev/cuaa1 doesn't exist. The error message is less than obvious:

  pppd: unrecognized option '/dev/cuaa1'
  

• Copy the name server config from the primary name server, which happens to be echunga.
• Create a higher-order MX mail configuration. I thought that echunga already had this, but it proved not to be the case.
• Modify the mail configuration on the other systems to go via sydney. The problem there was that sydney was on a wireless net, and when connected via wired Ethernet it used DHCP. Gave it a new name (kondoparinga) and pointed the other machines' mail config at that.

Over to Diane Saunders' and started things up again, and discovered in the process that I had forgotten a couple of things:

• The relay parameters weren't set up correctly, so a few messages were bounced because of the "no relay" rule.
• I had only just added kondoparinga to the DNS configuration, so of course it wasn't known in the outside world. As a result, a number of systems refused to accept mail. Fixed this one in the short term by changing the name to echunga.

Left the machine sucking in mail and down the road to see if anybody was working on the line fault. They were! As my documentation shows, just about everything I had been told was wrong. It turned out that the two lines connected by a pair-gain system were working again, though there was no way to find out except by trial and error. Connected the PPP link to the second phone number (normally my office number) and came back to life again. Ended up with about 3000 messages coming in via satellite in the space of 15 minutes, much faster than via the PPP connection. That's not only because of the higher bandwidth: it looks as if the higher-order MXs work on a message-by-message basis, so each one has to time out on the primary MX first.

Wednesday, 26 February 2003	Echunga	Images for 26 February 2003
		Top of page
		previous day
		next day
		last day

Down to the line fault site before breakfast to take some photos. While I was there, one of the linesmen told me that they were having significant problems because sounding out the lines was working badly from both sides, which suggests that there are additional problems with the cable on both sides of the current fault. And still they insist on replacing cables piecemeal.

4,000 mail messages to read! That kept me busy all day. It's really amazing how much time this line outage has taken up. Also spent some time rewriting the documentation of the line failure, but didn't finish.

There had been a voice mail message from George, a Telstra complaints person on my mobile phone this morning, but I didn't call back until I had the documentation in place. Spoke to him for nearly half an hour without coming up with very much I didn't know already. The bottom line was that it confirmed that they considered the Customer Service Guarantee as a guideline, and they don't have enough people to respond promptly to line faults (“we don't have people waiting around for lines to fail”). It was interesting, though, to note that the ACA does have the power to fine telcos if they don't meet service requirements. He stated that if there were three faults in a month, they would do something about replacing the entire cable.

In the evening, by coincidence, read an article in The Australian giving fault statistics. On the whole, they didn't look too bad. The problem is not so much the frequency as the duration of the faults, and they didn't address that issue. I note also that the three fault limit to which George referred had a duration of 60 days, not a month. It looks like they see it as a political hot potato, though, to judge by other articles in yesterday's print edition. It'll be interesting to see how things progress.

Thursday, 27 February 2003	Echunga
		Top of page
		previous day
		next day
		last day

Gradually back to normal, but the line fault isn't over yet. My PPP line went down again, so I called Claire at Telstra faults. While talking to her, another line (coincidentally where I had migrated the PPP) line went down too. They reported another fault, number S603027151, and coincidentally told me that the date for completion of the cable repair is 12 March. As long as I have a phone service, I don't care too much how long it takes them. Maybe they'll consider repairing it properly this time.

More work on the book. It's amazing how much backlog this line fault has caused.

Friday, 28 February 2003	Echunga
		Top of page
		previous day
		next day
		last day

The PPP line was still down this morning, but at least they were working on it, and in the course of the day it came up. They also worked on the line I had migrated to, testing it while I had a PPP connection up and running:

Feb 28 08:34:52 echunga pppd[34029]: Connection terminated, connected for 1204 minutes
Feb 28 09:14:56 echunga pppd[34029]: Connection terminated, connected for 37 minutes
Feb 28 09:24:33 echunga pppd[34029]: Connection terminated, connected for 9 minutes
Feb 28 12:22:47 echunga pppd[34029]: Connection terminated, connected for 176 minutes

Still, people were active. I received an "interim telephone unit" from Telstra in the post, a little late to be of any use. It's a phone-like device which connects to normal telephones but uses the mobile net. I'll hold on to it until we're sure things are fixed. In addition, I got a lot of phone calls asking if my lines were still up. One was from a linesman I hadn't seen before, Jim Downing, who confirmed that the whole cable needed replacement. He's been around here for a while, and was apparently the person who installed a phone for us in Hazelmere nearly 6 years ago. Now how do I convince Telstra to Do The Right Thing?

Spent quite some time preparing articles for AUUGN, which also involved putting up our NOIE slides on the AUUG web site. Converting Microsoft “PowerPoint” to PDF proves to be almost impossible; at any rate, I spent a couple of hours cursing and swearing at this horrible Microsoft user interface and what it produced (4 MB of PostScript, mainly repeated instructions, for 34 slides). Finally tried OpenOffice, which at least did that right, though I had spent plenty of time cursing it as well earlier.

Do you have a comment about something I have written? This is a diary, not a “blog”, and there is deliberately no provision for directly adding comments. It's also not a vehicle for third-party content. But I welcome feedback and try to reply to all messages I receive. See the diary overview for more details. If you do send me a message relating to something I have written, please indicate whether you'd prefer me not to mention your name. Otherwise I'll assume that it's OK to do so.