|
| Greg's diary |
| May 2005 |
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Sunday, 1 May 2005 | Echunga | |
| Top of page | ||
| next day | ||
| last day |
May already, and still no autumn rains! This must have been the driest April on record, and even the “showers” promised turned out to be just that, rather than the more typical deluge.
Today was another day where I didn't reach my objectives. Decided to give up on both the projects I was working on yesterday: I don't have time for either, and eucla already has a disk with a functional system on it, albeit old, so put that back in.
Out in the afternoon for a ride, though I wasn't feeling much like it. But it's ridiculous to have a horse to ride only once every month or two, so I've decided that wherever possible I'll go riding on a Sunday. The next three Sundays are bad for that, since I'll be travelling, so I really didn't have an excuse today.
In the afternoon, tried to turn my mind to other work, but somehow didn't make it. I've shed a lot of my workload over the last year, but it doesn't seem to have helped much.
| Monday, 2 May 2005 | Echunga | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Back to work today, an unspecified document to try to explain the design constraints to which I have been referring in the past week. Got quite a bit put together, but the depressing thing is that none of it is new. Why should this attempt to explain my view work better than previous ones? Not happy.
| Tuesday, 3 May 2005 | Echunga | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
More work on design documents today. What a pain!
| Wednesday, 4 May 2005 | Echunga | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
It's early May, and we still have no rain! The monthly rainfall bulletin for last month says:
Mt. Lofty Ranges (23C)
Rain (mm) Rain Days Decile
Obs | Ave Obs | Ave
ECHUNGA 64.0 8
MEADOWS 5.2 69.5 2 10 1
As so often, it doesn't specify any rainfall for Echunga—that's not because there was none, but because they don't seem to have reported it. Still, we're half way between Meadows and Echunga, and this is showing less than 10% of the normal April rainfall. We've also had almost none in May. Things are looking dangerous, and there's still no rainfall on the horizon.
More work on my design review. It had been meant to only take a little while, but another day went into it.
| Thursday, 5 May 2005 | Echunga | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
As usual, it's taking me far too long to get my tutorial notes updated; the work of the last few days took precedence. Today finally got time to look at them, and once again ran into a problem with creating slides that I have had a couple of times, but which I didn't get round to documenting.
The problem is that ghostscript doesn't offer an easy 4x3 format for slides intended for display. For some time I've been working around this by patching ghostscript/version/lib/gs_statd.ps:
--- gs_statd.ps 2004/07/20 04:53:54 1.1
+++ gs_statd.ps 2005/05/05 02:18:13
@@ -38,6 +38,7 @@
/.setpagesize { /statusdict .systemvar begin .setpagesize end } bind def
userdict begin
% Page sizes defined by Adobe documentation
+ /vga4x3 {480 360 //.setpagesize exec} bind def % output on standard VGA screen (4x3)
/11x17 {792 1224 //.setpagesize exec} bind def % 11x17 portrait
/a3 {842 1191 //.setpagesize exec} bind def
/a4 {595 842 //.setpagesize exec} bind def
@@ -126,7 +127,7 @@
% The values for all but letter and note are arbitrary.
/.pagetypenames
{ /letter /note %do not change this line, needed by 'setpagetype'
- /legal /lettersmall
+ /legal /lettersmall /vga4x3
/11x17 /ledger
/postcard /dbl_postcard /Executive /jenv_you4 /jenv_you2 /jenv_kaku2
/a4small /a3 /a4
I then convert PostScript to PDF with a command (generated via the Makefile, of
course) like:
=== grog@wantadilla (/dev/ttypu) /home/Book/Debug-tutorial 104 -> make slides.pdf
gs -q -dNOPAUSE -dBATCH -sDEVICE=pdfwrite -sPAPERSIZE=vga4x3 -sOutputFile=slides.pdf -c save pop -f slides.ps
Starting with a recent update of ghostscript, this no longer worked properly. The slides didn't get the same size. I traced it down to this entry in the PostScript:
%%BeginFeature: *PageSize Default << /PageSize [ 612 792 ] /ImagingBBox null >> setpagedevice %%EndFeature
I currently remove this with an Emacs LISP function that also does some other reshaping of the PostScript:
;;;; remove page size feature
(defun unsize ()
(save-excursion
(beginning-of-buffer)
(re-search-forward "^%%BeginFeature:.*PageSize" nil t)
(beginning-of-line)
(kill-line 3) ) )
Clearly that's not the correct way, of course; I need to find out how to suppress it in the first place.
| Friday, 6 May 2005 | Echunga | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Finally got round to actually updating my course notes for the BSDCan conference. It's funny how the work itself is relatively simple after I get the framework in place. Managed to get most stuff done, and now there are only the slides to go. I just need to make sure that I have the right software on one of my laptops, and then I'll be able to do it en route.
Yana back in the evening to stay a couple of nights.
| Saturday, 7 May 2005 | Echunga | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Spent most of the day planning my coming trip, for which I had more than enough time.
I still didn't have a hotel reservation for tomorrow night (it was supposed to be made from the USA), so I called the Marriott reservations line. “Rick” told me that their computer system did not check for double reservations, and that Marriott could thus not be held responsible for them. Spent 30 minutes on the phone with “Brett” and finally got a special agreement that they wouldn't impose a penalty if there is a double booking. That shouldn't need to be special.
In the afternoon got a phone call:
Caller: Hi, I'm Rebecca Mornay from Southern Cross mumble
Me: Sorry, I don't understand. Southern Cross what?
Caller: Southern Cross tlco. Can I have a minute of your time?
Me: Sorry, I don't deal with telcos on the phone.
Caller: Can I have a minute of your time?
Me: Sorry, I don't deal with telcos on the phone.
Caller: <click>
At least telemarketeers used to be polite.
| Sunday, 8 May 2005 | Echunga → Los Angeles → Dallas | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Up early this morning and off to the airport, where I discovered that my class upgrade from Sydney to Los Angeles had not come through, and that the plane was absolutely full.
Is it just that I'm getting older, or is cattle class becoming more and more intolerable? The cabin was dirty (looks like it hadn't been dusted for months), yet again the backrest kept slipping back—this seems to happen almost on every other flight—they didn't have the advertised food, and I was wedged between three people who seemed determined to keep me awake: the bloke behind me kept kneeing me in the back, the one to my left overflowed over the armrest into my seat and kept poking me in the ribs all “night”, and the girl in front used my feet as a rubbish area, occasionally kicking for good measure. I really need to reconsider these flights.
Los Angeles was the usual letdown: it seems that All American Airlines no longer includes food in the flight price, so I had to buy a “croissant” for lots of money. Given the quality of food on Qantas domestic flights, I fear they're trying to make the American option attractive.
On to Dallas and met up with Geoff Nairn. Went to dinner at a “brewery”, which did at least have some beers made somewhere else. Not very clever staff: the waitress warned Geoff that one of the beers was very bitter, and when I asked how many IBUs, she didn't know what the term meant. Like the Wig and Pen, their beers are undercarbonated, including a Weißbier that was also much cloudier than made sense, but which otherwise appears to have been brewed with the Weihenstephaner yeast and could have tasted quite good.
| Monday, 9 May 2005 | Dallas | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Todd Viegut arrived this morning and had breakfast with Geoff and myself, then off to the office, which is just round the corner. Interesting discussion: as in many such cases, it looks as if we're in violent agreement about a number of things. Didn't have much time to talk to Todd at the office: his day is a series of phone calls, one of them a conference call including people from Israel. Moved into the office of Jose Juniel, who arrived later. Jose is one of those few people who—until now—had exactly one hit on Google, and he wanted to stay that way, so I was left with no choice but to mention him here.
In the evening out to dinner, and late to bed.
| Tuesday, 10 May 2005 | Dallas → Chicago → Ottawa | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Todd in this morning and took me to the airport. Checking in was strange:
Officiel: Do you return to the USA?
Me: No.
Officiel: Are you a permanent resident?
Me: No.
Officiel: Do you return to the USA?
Me: No.
This began an endless loop, and she explained that she had problems because I answered “no” to both questions. That left two people who didn't understand, but that changed when I went through security: I was given yet another intensive security check. Seems that aliens leaving the country and not coming back are particular security risks. Interestingly, the bloke who searched my laptop bag didn't discover the back compartment, where I had a book. I suppose plenty of explosives would have fitted there too:
|
|||||||||||||||||||||||
Then Murphy struck: the flight was delayed. The original schedule had exactly 32 minutes stopover in Chicago, so almost any delay would have meant that I would miss my onward flight to Ottawa. In this case, they needed a change of equipment and a delay of about 45 minutes, and they put me on another American Airlines flight leaving 7 hours later. I was able to rebook on Air Canada, but for that I had to leave the security zone. In addition, I discovered that there would be issues with my baggage.
Of course, returning to the departure gate meant another third-degree security check (which again didn't find my “secret” compartment), with the result that the flight (and my bag) left without me. What kind of security is that?
Fortunately, the next flight to Chicago was just about to close, so there wasn't much delay. In Chicago, lots of red tape getting a new flight, including problems deciding where baggage was and—of course—another third-degree security check (which of course still didn't find my “secret” compartment). Down to the gate, where I discovered:
I would arrive at a time when Dan couldn't pick me up.
The flight was overbooked, and they were looking for people to volunteer to take a later flight.
As incentive, they were offering a voucher for a free round trip flight anywhere in the contiguous 48 US states.
The later flight was a better choice for Dan to pick me up.
As a result, volunteered to take the later flight, and had to go to another checkin counter to get my boarding pass. It seems that the first flight was run by United Airlines, though the flight number was Air Canada; the latter flight was a real Air Canada flight. The Air Canada personnel were particularly unhelpful: they refused to see me for another 45 minutes (“sorry, we're between flights”).
Off to get something to drink, bumping into Scott Long on the way, and then discovered an INS checkout machine, apparently intended to replace the old system of green forms stapled to the passport. Entered my passport, which it recognized without problems, but after that it wanted to read my fingerprints. It timed out and displayed a message “Please call attendant”. The second one did the same and then went out of order. Found a security guard supervisor who said that he wasn't an attendant (and he didn't know where there was one), but I would have to find a machine that works or I wouldn't be able to leave. Went to the Air Canada counter again, where they were at least able to check me in, and they said that there was no problem with leaving my card at the gate as in the past. Tried to get information about the location of my baggage, but they weren't prepared to do anything about that until boarding, by which time it would be too late and I might risk not being allow to board. Not a good impression.
The impression continued: the plane proved to be a Canadair CRJ, a small 50-seater. Getting there looked like going down a back alley: dirty staircase, cluttered and dirty way to the plane, dirty plane: the toilets had obviously not been cleaned in some time. If this is typical of Air Canada, they're not going to see me again any time soon.
Arrived in Ottawa a little ahead of schedule and answered a barrage of questions from a surprisingly aggressive immigration officer. The only other place I've been asked so many questions was Israel, and at least there they were polite. After that, spent 20 minutes waiting for my baggage, which didn't arrive. Went to the baggage services and found that it had arrived with the earlier flight, the one that I gave up. That's understandable, but:
As a result of the complete unhelpfulness of Air Canada, I wasted 20 minutes of mine and Dan's time.
This means that my baggage took completely different flights from Dallas to here. I thought that was one of the big no-nos of airline security, but even on this flight I wasn't the only one. Admittedly, when I unpacked it it looked as if somebody had gone through it looking for anything dangerous, and paying no attention to keep things tidy:
|
|||||||||||||||||||||||
Off with Dan into town, and finally got something to eat (there was nothing on the plane, not even at a price) round midnight at the local pub, where a number of the usual suspects had congregated. Surprisingly warm for the time of evening—I'd guess about 23°. You'd almost never see that in Adelaide.
| Wednesday, 11 May 2005 | Ottawa | Images for 11 May 2005 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Dan picked us up early this morning and marched us off campus to “Cora's”, a restaurant specializing in large, sweet breakfasts. Nice enough place, but we rather overloaded their facilities, and arrived late at the kernel summit as a result.
The kernel summit was unstructured, which meant that nothing was planned. Some people split off and talked about various topics, none of which I found too interesting, so spent the time updating my tutorial slides for tomorrow. In the middle of that, discovered that my mail system was clogged up: some idiot was connecting with an invalid server name (something like myhost.local.host) and being rejected. He didn't take no for an answer, though: I had something like 35,000 bounce messages backed up. SpamAssassin is really bad at handling this situation: it takes on average 30 seconds per message for various lookups, so the queue would have taken days to clear. Stopped SpamAssassin for a while and let it run through. That worked, but what a mess!
In the evening off to the “Lone Star Grill”, and had the Texas barbecue meat that I had wanted to have two days ago. This certainly isn't Texas, as was confirmed on the way back: the temperatures must have dropped 20° since yesterday.
| Thursday, 12 May 2005 | Ottawa | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Yesterday's mail congestion made itself visible:
Date: Fri, 13 May 2005 01:01:28 +0930 (CST) From: grog@lemis.com (Cron Daemon) To: grog@lemis.com Subject: Cron <grog@wantadilla> /home/grog/bin/mailvolume On 12 May 2005 you received 71091 mail messages and sent 17 mail messages.
Decided that getting breakfast at Cora's was too much trouble, or at least would take too long, so located an on-campus café across the way from the residence hall. They made it very clear why Dan preferred Cora's.
Managed to eat it anyway, then to the lecture hall where I held the tutorial. It was barely big enough: I had 44 participants, which might be the largest crowd yet, and they started arriving over 30 minutes early. Even Mike Karels (ex CSRG) showed up; his interest were the new debugging techniques that have developed under FreeBSD. On the other hand, he also has access to a lot of tools that the BSD/OS people developed, and he has the time and commercial motivation to import them into FreeBSD, and to fix the current breakage in FreeBSD as well. Very positive.
I'm beginning to think that the Royal Oak is the real centre of the conference. We seem to be there all the time; today had both lunch and dinner there. I'm gradually getting through their impressive array of beers, many of which are unfortunately less impressive per se.
Moved into an apartment with Dan in the evening. He's all excited about a new security breach with Intel HTT processors that Colin Percival will reveal at the conference tomorrow.
| Friday, 13 May 2005 | Ottawa | Images for 13 May 2005 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
First day of BSDCan conference today, and first off to Cora's again. Again spent an inordinate amount of time getting breakfast. There must be a better way.
First talk was Colin Percival's talk about HTT dangers. It wasn't what any of us thought: he was utilizing differences in cache access times to decide what another program was doing. That doesn't sound very dangerous until you consider that the other program could be an encryption program that you spawn. He showed some interesting graphs showing cache access time against time. They clearly showed various steps in the OpenSSL RSA algorithm; a clear case where open availability of the source is a liability. It seems that there's a fairly straightforward workaround: add hooks to the scheduler to disable scheduling the second “CPU” for certain critical sections.
Then book-signing along with the other authors present. Kirk McKusick was the busiest, signing both his latest book and some copies of the older ones.
After the break Robert Watson did a talk entitled “The FreeBSD SMPng Network Stack”, but really going over the entire tradeoffs of the SMP implementation. I would have liked more concrete examples.
Lunch was a catastrophe. The pub across the road had space for us, but not the personnel: they hadn't been warned that we were coming. Didn't get back on time, and so missed the next session and went back to my room to work on a presentation on beer, but didn't get far enough to do anything useful.
Back for more book signing, and then listened to a talk by der Mouse on “live network backup”, which turned out to be more interesting and relevant than I expected: he was talking about network mirroring. Some of the details, notably authentication, need work, but it seems to work, and he's addressed a whole lot of the issues that made Tandems' RDF a problem.
Also did a bit of planning to perpetuate the “Real Tom Rhodes“ myth, and came up with some amusing ideas.
Last year Tom Rhodes had difficulty identifying himself, and so both Robert Watson and Wes Peters had signs made saying “I am the real Tom Rhodes“.
This time we got Dan Langille to make six name badges (the number on a sheet) entitled “The Real Tom Rhodes”. We're planning to do something during tomorrow's key signing.
Then a BoF on beer (“The foundation of modern civilization”, about which I'll write a presentation some time), somewhat hampered by lack of equipment and network connectivity. Ended up borrowing a laptop from Henning Brauer (nomen est omen?), and managed a reasonable off-the-cuff talk.
In the evening to dinner at the Hard Rock Cafe, not a place I would normally go to, not indeed go to again. It wasn't as extreme as I feared, but it also doesn't have much of interest about it. I'm also getting heartily sick of a diet of hamburgers and sandwiches.
After that, off with Michael Lucas, Mike “silby” Silbersack and Brad Davis round the periphery of the town centre and ended up in a pub called, incongruously, the Elephant and Castle, where we had to wait to be seated, but got a reasonable Fullers “London Pride“ beer, this time clearly tasting of the ESB yeast (Wyeast 1968) that I have been using for my recent brews. Also came up with the idea of a web site www.therealtomrhodes.com.
Back to the residence to discover Kirk McKusick looking for me to borrow a corkscrew, which proved to be for an interesting 1998 California Zinfandel, not a wine I know much of (Kirk confirmed that it's as good as unknown in Australia), but very good. The room had a singular concentration of authors: Kirk, Mike Karels, Sam Leffler, George Neville-Neil, Michael Lucas and myself. About the only other people present were Jake Burkholder and Dan Langille, the latter also thinking of writing a book.
| Saturday, 14 May 2005 | Ottawa | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Looking for a better alternative to Cora's for breakfast, joined up with Massimiliano Stucchi and Matteo Riondato, the FreeSBIE people, who led me into the CBD, which was much closer than we thought. There we found a little restaurant with good and cheap food; only on the way back did we find a Japanese and Korean restaurant that also offered breakfast. That's one for tomorrow.
Second day of the conference today, and first to Reyk Floeter's (or is that Flöter?) talk about free firmware for the new generation of wireless cards with downloadable firmware. The background is that, unlike FreeBSD and NetBSD, the OpenBSD project won't accept proprietary binaries in the source tree. We've had a certain disdain for this attitude, in particular because of concerns about regulatory bodies, but Reyk did a good job of putting things into perspective, and it now sounds like a good idea. I must confess that I'm glad somebody else is doing it. The extreme stance reminds me somewhat of rms, and it's interesting to notice that he actually presented Theo with an award relating to this work earlier this year in Brussels. Poul-Henning Kamp tried to throw a spanner in the works by claiming that Reyk was laying himself open to prosecution. Reyk was somewhat taken aback, but they have had legal advice, and I'm left thinking that Poul-Henning was just trying to stir things up. He seems to be about the only one of that mentality left.
After the break another OpenBSD presentation: Bob Beck on spam reduction with the OpenBSD spamd, not to be confused with the Spamassassin daemon of the same name. Very useful stuff. It may not have been the most interesting talk of the conference, but I'd give it my vote for the most useful, and the first thing I try when I get back home.
To lunch with Michael Lucas, Dru Lavigne and Rob, her Linux-loving boyfriend. Ate Shawarma for the first time. I had tried to make it last year without really knowing what it was. I did know that it was like Döner or Giros, but I had thought that there was some trick in preparing the meat. There isn't: it's all in the serving, which makes things a lot easier. After lunch went to a nearby bookshop looking for our books. Only Michael was successful, and that was just a copy of “Cisco Routers for the Desperate“.
After lunch listened to an interesting talk about jail(8), which was educative, then a bit of Jacques Vidrine talking about VuXML, an XML-based way of maintaining a vulnerability database. I suppose it makes sense, but it seems like a lot of infrastructure to achieve it.
In the final session finally managed to present ourselves as The Real Tom Rhodes, not as spectacularly as I had hoped. Lots of photos were taken, and I'll update this entry when I find them.
Dinner in the evening was at Yet Another Pub, so we again decided to go elsewhere, this time to a nearby Indian restaurant, where the food was better than it was typical. Joined the others in the pub—they wanted to seat us again—and stayed until the noise level got too high. Back to the residence and decided to drive to Montréal tomorrow: that's where I'm leaving from, and I'll have most of the day to get there. Dead tired to bed.
| Sunday, 15 May 2005 | Ottawa → Montréal → | Images for 15 May 2005 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Up this morning and off into the CBD to have breakfast at the Korean place and pick up my hire car. The first fell flat: there's a big billboard outside, but the restaurant was shut. So was the restaurant where the GUFI people and I had breakfast yesterday. Off to pick up the car, which went surprisingly quickly: once the agent had returned from some parking activity, it couldn't have taken more than 3 minutes. The whole positive experience was somewhat marred by lack of maps: all they had was Ottawa.
Off to Cora's, where the others were still having breakfast, and I got mine much faster than I expected, and then to the residence to pack up and leave. When I did, I found an “Infraction Notice” claiming that I had parked without a permit and demanding $50. Took another look at where I had parked: there was no sign anywhere near that there was any kind of requirement. The people at reception told me that I should put in an appeal, and told me where to go. Went there, investigating as I go. For further reference, when parking at the University of Ottawa:
The “permit required” sign is written in tiny font at the entrance to the parking areas. Unless you get out of the car and approach on foot, you can't read them:
|
|||||||||||||||||||||||
It's not obvious how to get the permits themselves. The answer: there are various machines around marked “P&D”:
|
|||||||||||||||||||||||
This stands (at least in English) for “Park and Display”, or possibly “Pay and Display”. I can't think what it could mean in French; indeed, I had to be told what it meant in English, since the machines have nothing whatsoever to identify them as parking ticket machines, and on Friday I had investigated one to decide whether it was an ATM:
|
|||||||||||||||||||||||
I suppose it's clear enough to people who know the system, but it's completely unfair to visitors like myself, for whom it's almost impossible to know that a ticket is required, or where to get one. Under those circumstances it's good to know that they tend to be lenient in such cases.
Then off to Montréal along the highway 148, skirting the north side of the Outaouais River. I'm still not clear how this relates to the St.-Laurent, but it appears to be a tributary. Tried to find something to eat along the way, without much success. To the credit of the Québecois, there are almost no American fast food shops along the way—the alternative appears to be a casse-croute, though I've seen a number of signs marked “Dépanneur” that appear also to be food-related. I wonder how that happened. Apart from such places, the restaurant culture appears to be similar to France, and I didn't really feel like a full meal at lunch.
On to Montréal, discovering just how bad both my maps and the signposts were. Fond the town centre on the third attempt more by good luck than any help from the maps, and found that even on a Sunday I needed to pay exorbitant rates ($3 per half hour) for a parking space. Went to the tourist information place in the middle of town, which was a complete disappointment: I didn't even get a map of the town. Walked and then drove around a bit, but it seems that I need more preparation. What I saw looks nice, but I didn't have time to investigate it thoroughly. One thing I did decide: the Québecois have redefined the words for the meals of the day:
|
|||||||||||||||||||||||
My best guess is:
| English | French | Québecois |
| breakfast | petit déjeuner | déjeuner |
| lunch | déjeuner | dîner |
| dinner | dîner | souper |
I'm left wondering whether it's related to the difference in time zones.
Off to the airport, which was a bit of an astonishment. The signs were incredibly bad, and the check-in clerk seemed remarkably obtuse. I was flying to Bergen via London and Oslo, and I asked her to check in my baggage all the way. She said that that wasn't possible, since I was staying 3 days in Oslo. It took me a couple of minutes to make it clear to her that she was looking at the wrong fields (I'll be returning from Bergen to Oslo on the 19th).
Through security—at least that's sane again now we're out of the USA—and looked for the British Airways Lounge. No signs, and nobody knew where it was. Finally discovered that it was before security (what a clever idea), so I effectively had to re-enter Canada to get there. That took about 20 minutes.
Finally off in the typical packed-full and badly maintained British Airways flight. Despite my frequent flyer preference requests (window seat, as far forward as possible) they put me almost completely at the back of the plane. Last time I couldn't turn the overhead light off; this time I couldn't turn it on. And of course pressing the call button for the attendant went unanswered; at some later time they just turned the light off. This time, perhaps out of deference to the Québecois, the food wasn't anything as prosaic as bangers and mash, but it was still pretty terrible.
Sat next to a girl from Co. Limerick, Anne Conlan, with whom I had a lively discussion. On a 6 hour flight there's not much chance of sleep at the best, but this time I got almost none at all.
| Monday, 16 May 2005 | → London → Oslo → Bergen | Images for 16 May 2005 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Arrived in Heathrow a few minutes ahead of time, and once again I was reminded how little regard British Airways has for frequent flyers: elsewhere in the world they go through the priority queue, but here I had to go through the normal queue, even though the priority queue was empty. Doesn't British Airways want more frequent flyers?
Back to the same club that I complained about last November. To their credit they now serve sandwiches, which meant I didn't have to go outside to get something to eat. But their information monitors were wrong: they had my flight to Oslo going to Basel instead. How can that even happen? It seems that things are entered manually. That could lead to really serious problems.
Finally got on the plane, sat down and pulled the flight magazine out of the pocket in front of me. With it came a full sick bag, which emptied itself on the floor and my feet, pants and shoes:
|
|||||||||||||||||||||||
(The photo shows a white cloth that the staff gave me so that I could clean up the mess; they didn't offer to help me). The cleaners really need to pull their socks up. Until they do, it's really not a good idea to fly British Airways.
Several hours wait in Oslo airport, where wireless access appeared to be free:
I was wrong. There is free access to the local web servers, but for the rest they charge like a wounded bull. I would have had to pay about 100 Kroner for my stay, about $20 Australian.
On to Bergen, where I was picked up by Jakob Grimstveit. Out to have a pizza; it's difficult to realize how late it is here, when the sun goes down at about 9 pm, and it still wasn't properly dark by the time I went to bed at 11 pm.
| Tuesday, 17 May 2005 | Bergen | Images for 17 May 2005 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Today was Constitution Day in Norway (and the reason why I came here right now). Into town by bus to the Bryggen area to see a parade of the local associations and clubs, somewhat marred by what is apparently typical Bergen weather: alternating sunshine and showers, in this case mixed with hail. When that got too strong, we went into a local pub (Sjøboden if I recall correctly) to have a couple of beers, during which time our numbers swelled to about 8. Then round to see Peter Hansteen's office, one of the most unusually located I have seen: it's in a World Heritage listed timber warehouse:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Walked around for a while, Peter telling me some of the history of the town, and then up a funicular railway for an overview of town:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Then halfway down again to Peter's house, which does not have car access: there are at least 200 steps in between. Certainly the view from the house is spectacular, though.
Norwegian specialties for lunch, then spent all afternoon and evening talking, drinking and designing RFCs (Peter spearheaded the first proof of concept of RFC 1149, IP datagrams over avian carriers, and at present we're thinking of a follow-up RFC about gateways involving simian interchange). It's unnerving how long it stays light here; I discover it is as far north as I have ever been, just a little further north than Helsinki, but it's also much further west, while remaining in the same time zone. As a result, when we went to a nearby lookout at 2300 to watch the fireworks, it was only just getting seriously dark. Back home with Jakob, and of course got to bed very late.
| Wednesday, 18 May 2005 | Bergen → Voss → Gudvangen → Flåm → Myrdal → Bergen | Images for 18 May 2005 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Up early yet again and down to the station, where I met up with Peter and a couple of the others whom I met yesterday, and off by train to Voss, where we caught a bus to Gudvangen, on the Nærøyfjorden, stopping for a spectacular view from Stalheim on the way. By boat down the Nærøyfjorden to the junction with the Aurlandsfjorden, and up that to Flåm. Very pretty scenery, and a stark contrast to the river Murray last month. Interestingly, the distance we travelled was about the same, but this time with the faster boat it only took 2 hours, while last month it took all day.
From Flåm by a very steep railway (apparently the steepest conventional railway in the world) through some very pretty scenery to Myrdal, then back to Bergen, not helped by the fact that they had attached the wrong kind of wagon to the train, and our reserved seats simply didn't exist.
The pace didn't let up back in Bergen: first had dinner at the local pasta place, then to the university for an extraordinary meeting of BLUG, where I talked—again—about beer, and then did the talk “Why BSD is better than Linux” for an encore. Finished at 2200, and off for a couple of beers. Amusing incident at the bar: the barkeeper, who like most people here speaks excellent English, asked me what my native language was. I suppose it happens quite frequently here that non-native speakers speak English to each other because it's the best language they have in common.
Once again, despite the requirement of getting up early tomorrow morning, didn't make it to bed until 1 am. Norway is not good for your sleep.
| Thursday, 19 May 2005 | Bergen → Oslo → London → Oxford → Seer Green → London → | Images for 19 May 2005 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Up this morning not quite at the crack of dawn (that must have been about 2 am), but still early enough to feel decidedly dopey. At the airport had an incredible queue: one person badly handling all checkins, while the priority checkin was empty. Went to the latter and was accepted at once, including checking in my baggage all the way to Adelaide (which I fortunately noticed and had changed to Singapore).
In Oslo, things seemed smooth until I looked for the British Airways lounge. Since Oslo is their only Norwegian destination, you'd expect it to be in the international departures section, but in fact it's in domestic, and well hidden at that. Ended up with yet another trek around a large airport.
The lounge didn't have basic facilities like telephones (I wanted to book a car in London), but the personnel there were the most helpful I've seen in any lounge anywhere, including placing calls on their only phone to car hire companies in England (who didn't have time to talk to me). I also discovered I could have come in here on my way to Bergen: you don't have to be flying British Airways to be allowed to use the lounge.
On to London, where I managed to get through formalities pretty quickly and found a car (with Budget, for only £29, as opposed Avis' best quote of £45 for a preferred customer). Within an hour of touchdown I was in roadworks on the M25.
Down to Oxford, where, apart from a brief drive-through in early 1997, I haven't been since about 1991. About the only thing that has changed is that you can no longer drive through the middle of town, though I didn't have any difficulty finding a parking space. Walked round the middle of town reminiscing—my Oxford time was in the time frame 1970 to 1972—and then back to Seer Green, not far from Beaconsfield, where Catherine (once Cathy) and Nigel Powell have been living for nearly 20 years. I hadn't seen them for about 15 years either. Spent 2½ hours briefly reminiscing old times, took some photos, then off to the airport. Pretty full day.
| Friday, 20 May 2005 | → Singapore | Images for 20 May 2005 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Got a relatively good “night”'s sleep, and didn't wake up until 3 pm. Landed early at Singapore, where everything ran smoothly: through immigration in no time, barely had time to approach the baggage carousel before my bag arrived, didn't even see a customs officer on the way out, booked a hotel in about 5 minutes (two hotels were booked out, which I suppose counts as a slight minus), and within 30 minutes of touchdown and 20 minutes after scheduled arrival, I was on the East Coast Parkway (unlike yesterday, not in roadworks) headed for town.
The hotel they booked for me is the Perak Lodge, conveniently placed just off Sungei Road (better known by the name of the other side, Rochor Canal Road). Pleasant little place; I may come back. Out to dinner with Lim Kim Yong and Jeffrey Lim at the Tien Jin restaurant, specializing in food from that area.
After that back to near the hotel; it seems that Kim Yong grew up in Dunlop St, just round the corner in a little Chinese enclave between the Indians and the Malays. Now it's full of backpacker places, and we had a beer at one of the more unusual ones. They have three “stray” cats there, for whom they're trying to find homes, and Kim Yong is sorely tempted to take one. One looks surprisingly like Jakob Grimstveit's ocelot cat, though without the typical ocelot markings. All three have straight tails, which makes it seem very unlikely that they're real strays: almost without exception, all Singapore cats have deformed tails.
| Saturday, 21 May 2005 | Singapore → | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Up late this morning; the room proved to be singularly noisy, and I actually slept worse than on the plane the night before. Discovered that I had to check out by 1 pm at the latest, and so off to look, once again, for a nasi lemak.
Nasi Lemak is almost the national breakfast of Singapore. Why is it so hard to find? First to Sim Lim Square, where I confirmed that the basement food court did not have any, then back across towards Serangoon Road, passing a food court on Sungei Road which had it—with green rice!
On to my place in Serangoon road. Last time it was closed; this time it was closed down and replaced by something else. Back to the place in Sungei Road and had nasi padang instead; at least then the rice was white.
Then up to Mohammed Mustafa Samsuddin, where I didn't really have much to buy. Dealt with that, back to the hotel (passing a sign saying “Nonya nasi lemak” on the way; to be investigated next time) to check out, and discovered I had nothing to do for 8 hours. Spent some time looking through the tourist guides, but I'm not a tourist, and it seems that even the national museum on the corner of Fort Canning Road has closed down. In any case, for some reason I was feeling a little weary, so off early to the airport, where I was able to hijack an Ethernet connection in the Qantas Club and handle about 3000 mail messages. Off without event towards Australia.
| Sunday, 22 May 2005 | → Darwin → Adelaide | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Woken up at 3:30 am to be told we were landing in Darwin. Once again we had to get out and go through a complete security check, including removing laptops. You'd think they did it to annoy you. Spent it writing a letter on the subject to my MP, who also happens to be Australia's Minister for Foreign Affairs.
Arriving in Adelaide wasn't that smooth, either. It seems that they have new passport scanning software that is a lot more finicky than previously. It detected an imperfection in the attachment of my passport photo, one that has been there since I got it, and as a result I had to wait for a complete confirmation. I was told that about 10% of all passports now fail the scan, and that I should get a new passport. I wonder who will pay for that.
As if that wasn't enough, the real security screening of my baggage showed an unexpected item, requiring complete checking of my bags. The customs officer suspects it was a package of 100 empty CD-Rs that I bought yesterday. It's not likely that I'll bring more in in the near future, but I suppose hand-carry would be an option.
So I'm finally home after probably the longest trip in my life, over 42,000 km:
| From | To | km | ||
| ADL | SYD | 1160 | ||
| SYD | LAX | 12100 | ||
| LAX | DFW | 1920 | ||
| DFW | ORD | 1290 | ||
| ORD | YOW | 1050 | ||
| YUL | LHR | 5210 | ||
| LHR | OSL | 1170 | ||
| OSL | BGO | 309 | ||
| BGO | OSL | 309 | ||
| OSL | LHR | 1170 | ||
| LHR | SIN | 10900 | ||
| SIN | DRW | 3330 | ||
| DRW | ADL | 2620 | ||
| Total: | 42538 |
Spent most of the day processing my photos. In the evening, the air conditioner failed again. Looks like the same problem as last time. At least it happened when it wasn't too cold, and after I got home: remote debugging over the phone is always a royal pain.
| Monday, 23 May 2005 | Echunga | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Back to the old slog again today. Going round the world in15^W14 days doesn't seem to have changed home very much. It's still dry, and there's still little sign of rain. Spent most of the day catching up with things that happened while I was away.
| Tuesday, 24 May 2005 | Echunga | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Phil came along to look at the air conditioner today. Things aren't good: the compressor has a short to ground, and one of the coil fan motors has fallen out of its bracket and burnt out. The repair will probably cost more than it's worth. Phil wasn't able to give me a quote, and the suppliers said that they no longer serviced the system. I wasn't able to find a supplier for the parts. Looks like time for a new air conditioner.
When will it finally rain? The Bureau of Meteorology has been predicting far too little rain for weeks now, and most of that hasn't fallen. So it was today. The forecast wrote:
Tuesday : Cloudy periods with a light shower or two, chiefly afternoon. Mild with moderate to fresh north to northwest winds, turning moderate southwest in the afternoon.
In fact, we had cloudless skies. Things are looking bad.
To Belair Conservation park this afternoon to buy some plants for the garden. Ben seems to think that now's the time to plant; I wonder how they're going to survive in this drought.
| Wednesday, 25 May 2005 | Echunga | Images for 25 May 2005 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Ben Henderson in today, amongst other things to plant the plants we bought yesterday. It still hasn't rained worth mentioning, and I'm concerned that the plants will die. Still, they're in there now. For future reference:
|
|||||||||||||||||||||||
South-East bed. Existing geranium in the foreground. Behind the rock, a single Grevillea Thelmanniana baby, an Acacia Maxwellii, a Grevillea Humifusa, another Acacia Maxwellii, then three Grevillea Thelmanniana babies in a triangle. To the extreme left in the background, two red Anigozanthos “Bush inferno”.
|
|||||||||||||||||||||||
South-East bed from the other direction.
|
|||||||||||||||||||||||
3 Lavender plants of different sizes in the foreground, mint on right, Hardenbergia Violacea “Flat White” and “Carpet Royal” (violet) at the bamboo sticks. Behind that, close to the grass, a number of Bearded Iris that we got from Diane Saunders. We'll let ourselves be surprised by the colours..
When we built the extension to the house in 1999, we installed a pressure pump for the sprinklers. It turned out that it wouldn't stop running: we must have had a leak somewhere, but it wasn't big enough to be easy to find, so we installed a switch and only turned on the pump when we needed it. In the last few months, the leak has become big enough, and Ben identified it:
|
|||||||||||||||||||||||
The connector in the middle is leaking, possibly damaged. The puzzling thing is the unconnected pipe underneath the small one in front. I wonder where it goes.
Also did some work investigating performance. To save on a division in a tight loop, we've been using a table lookup. Instead of this:
if ((index % CONST) == 1)we've been writing:
if (table [index])
table is 16 kB in size, which means that we're filling L1 cache every time we run the loop. Tests suggest that there's not much improvement in the speed: cache misses can be more expensive than divisions. More investigation needed.
| Thursday, 26 May 2005 | Adelaide → Sydney | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
More work on yesterday's optimization, and got completely different results from yesterday. On the one hand they showed that a division is still a lot slower than a table lookup—it slowed the program down by an order of about 5:1—but it also confirmed that, not surprisingly, a third method was faster than the table lookup:
if ((index & MASK) == 1)
MASK is the log base 2 of CONST, and clearly it can only work if CONST is a power of 2. Still, work to be done.
In the evening off to Sydney for an AUUG board meeting on Virgin Blue. How I hate that airline! Maybe it's just that they get on my nerves with their silly talk and their irritating “music”. Today I had other problems as well. I managed to get a place in an emergency exit row, normally something they charge extra for. I've decided that, at least on Virgin Blue, this is not a good idea. Because of the extra leg room, the rows are always full, and the window seats are missing an arm rest. In addition, the window seat in an exit row seems to be particularly cold (maybe missing insulation in the exit door itself), and I was freezing by the time I got to Sydney. For some reason airlines always put particularly inconsiderate people in the middle seat. This one, though not as large as I have seen, did his best to overflow to both sides.
Then to the Rydges Hotel in North Sydney, where they had given me a suite instead of a normal room. That would have been nice of them had they not decided to charge me $50 more on top of the already high $158 for it, without notice. For that price I would have expected doors that shut properly, and a somewhat more modern bathroom.
The misunderstanding also made me more sensitive to other additional charges, and I found plenty: of course the Internet connection cost money (I had to authenticate to find out how much), I found a bottle of wine (coincidentally a Rosemount wine that we drink at home, and for which we pay about $10) for $17.50, along with a label saying “drink me for $5.50 when you eat in the restaurant”. In the restaurant, the waitress confirmed that I would have to pay both prices if I wanted to drink it there. In the room there was also a bottle of mineral water for $3.50 (no other water in the room). The whole impression is that they're trying to pull a fast one.
The Rydges is not ideally situated for finding something to eat at 21:30. I need to consider staying somewhere else in Sydney when I come for these meetings. Ended up eating a very expensive pasta dish in their restaurant.
| Friday, 27 May 2005 | Sydney → Adelaide | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Down to breakfast this morning and got a surprisingly good cooked breakfast for less than I had expected to pay, only about $15.50 instead of the $25 I had been quoted. This was somewhat marred by the bill I was presented when I checked out: they had still charged extra for the room, and the waitress had added $3 gratuity—very high by Australian standards—to the bill. Both were taken off again, but I was left with the feeling that this is a hotel where you need to check your bill carefully.
Most hotels supply feedback forms in the rooms, and I usually fill them out. Most contain general questions (how clean was the room? Condition? Quality of food?), but the one I found today only left space for text. I can't write too well by hand, so just left a URL pointing to this diary (which is also one of the reasons I've gone into so much detail).
Board meeting as normal. Had a demonstration of some web content management software based on Zope and Plone; quite impressive, though I don't really understand enough to be sure. In any case, it's beginning to look like we'll be using it.
Computerized air travel checkin seems to be coming quickly. Today I watched Michael Still check in at the Qantas club, and then had my own experience at the Virgin Blue terminal. Both work relatively well, and they allow a better overview of seating than the checkin clerks do, but they don't eliminate waiting, in particular at Virgin Blue, where I had to wait up to 30 seconds for a seating plan. And of course there were these stupid Virgin Blue people keeping an eye on things. Well, I don't think they're naturally stupid, they're just required to toe the company line:
VB Girl: “How are you today? Everything wonderful?”
Me: “Well, it works, but it's very slow”.
VB Girl: “Wonderful!”
Me: “I don't think that long waits are wonderful. You people need to do something about the response times.”
VB Girl: “Well, lots of people are using the machines”.
Me: “Sure, so you need to dimension your system accordingly.”
VB Girl: “Wonderful!” (walks off).
| Saturday, 28 May 2005 | Echunga | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Hadn't planned to do much today, but got a call asking for help on an embedded video system by somebody I know in Adelaide, so down there to take a look. Things were relatively easy to fix, in particular the first one (forgot to pass O_APPEND to open), one of the few one-bit fixes I've done. Still, it took all afternoon, and when I got home I was rather tired. I wonder why.
| Sunday, 29 May 2005 | Echunga | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Brew day again today, and it's showing that I haven't brewed in while. For the first brew I used 3 litres too much mashing water, and I forgot the sparge manifold; strangely, it didn't seem to make any difference to the wort.
Brewing wouldn't be nearly as much work if I only did one batch. Time to think bigger batches through again.
| Monday, 30 May 2005 | Echunga | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Into town for the ADUUG lunch today, at the Cathedral Hotel opposite St. Paul's cathedral in North Adelaide. What monstrous schnitzels! I didn't know they made cows big enough.
ADSL has come to Echunga! The Telstra web site confirms it, and, more importantly, so does Internode, who pointed it out to me. I put in an order with Telstra two months ago, and subsequently tried to contact them to find out whether the lines were suitable or not. Their basic answer: “Don't call us, we'll call you”. They haven't done so yet; I'll be interested to see when (if) they do.
Line quality is still an issue, of course. Got some information about that, apparently from a Telstra internal web site, and not from Jodie of Internode. The line I wanted to provision has the following information:
[Found a 1 meter bridge tap, ignoring] [Found a 2445 meter bridge tap, adding 3.5dB loss fudge factor] [Found a 387 meter bridge tap, adding 3.5dB loss fudge factor] [Found a 1316 meter bridge tap, adding 3.5dB loss fudge factor] 1387 meters of 64PIQL cable at 9.43 dB/Km loss (13.07941 dB) 555 meters of 64PEIUT cable at 7.34 dB/Km loss (4.0737 dB) 100 meters of 64CPIUT cable at 8.54 dB/Km loss (0.854 dB) 1067 meters of 51PIQL cable at 11.9 dB/Km loss (12.6973 dB) 110 meters of 40CPFUT cable at 10.79 dB/Km loss (1.1869 dB) 1443 meters of 64CPFUT cable at 7.19 dB/Km loss (10.37517 dB) 2 meters of 40PEIUT cable at 11.52 dB/Km loss (0.02304 dB) 2 meters of 40PIUT cable at 13.81 dB/Km loss (0.02762 dB) [Adding 10.5dB of fudge factor for 3 bridge tap(s) greater than 25 meters] Total cable length: 4666 meters Total cable losses: 52.81714 dB at 300kHzAnother one has:
[Found a 387 meter bridge tap, adding 3.5dB loss fudge factor] [Found a 3073 meter bridge tap, adding 3.5dB loss fudge factor] [Found a 177 meter bridge tap, adding 3.5dB loss fudge factor] [Found a 688 meter bridge tap, adding 3.5dB loss fudge factor] 1387 meters of 64PIQL cable at 9.43 dB/Km loss (13.07941 dB) 555 meters of 64PEIUT cable at 7.34 dB/Km loss (4.0737 dB) 100 meters of 64CPIUT cable at 8.54 dB/Km loss (0.854 dB) 1067 meters of 51PIQL cable at 11.9 dB/Km loss (12.6973 dB) 110 meters of 40CPFUT cable at 10.79 dB/Km loss (1.1869 dB) 1443 meters of 64CPFUT cable at 7.19 dB/Km loss (10.37517 dB) 2 meters of 40PEIUT cable at 11.52 dB/Km loss (0.02304 dB) 2 meters of 40PIUT cable at 13.81 dB/Km loss (0.02762 dB) [Adding 14dB of fudge factor for 4 bridge tap(s) greater than 25 meters] Total cable length: 4666 meters Total cable losses: 56.31714 dB at 300kHz
It's worth looking at these numbers carefully. Telstra specifies a maximum loss of 56 dB, so the first one is OK, the second is not. Never mind that British Telecom, for example, specifies a 60 dB limit, and elsewhere they've shown that you can run ADSL over lines with as much as 80 dB losses, or that it works better if you have no voice channel. What I find much more interesting is that they add a 3.5 dB “fudge factor” for each bridge tap, but estimate the losses down to 5 places of decimals. Obviously these are calculations, not measurements.
I discovered that a bridge tap is a parallel wire between two posts, which apparently improves voice quality but causes echoing in DSL lines. They're supposed to remove them during provisioning, which would bring both lines down to a loss of 42.31714 dB, well in the range of ADSL.
So what now? It seems that Internode don't route directly: they do it with a GRE/IP tunnel, which requires the use of nos-tun, a program I had never heard of, but which seems to have been there forever. Now to find out how to interface it with a firewall.
| Tuesday, 31 May 2005 | Echunga | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
The Bureau of Meteorology has been crowning itself in glory again. Here are extracts from a few recent weather forecasts:
FORECAST FOR THE ADELAIDE METROPOLITAN AREA Issued at 5:15 am on Monday, 30 May 2005 For Monday. Next three days for Adelaide City : Tuesday : Min 7 Max 17 Cloudy. Mainly fine. FORECAST FOR THE ADELAIDE METROPOLITAN AREA Issued at 11:30 am on Monday, 30 May 2005 Tuesday : Cloudy with a light shower or two developing during the morning. Cool with light northeast to northwest winds, shifting moderate southwest to south around midday. FORECAST FOR THE ADELAIDE METROPOLITAN AREA Issued at 5:15 am on Tuesday, 31 May 2005 Tuesday : Cloudy with light showers. Cool with light winds, shifting moderate southwest to south around midday. FORECAST FOR THE ADELAIDE METROPOLITAN AREA Issued at 11:35 am on Tuesday, 31 May 2005 Tuesday : Cloudy with a light shower or two. Cool with light to moderate southwest to southerly winds.
In fact it drizzled all day, which at least gave us some rain. We could still do with much more.
More thinking about ADSL today. One of the other things I also need to do is to set up the pf firewall to try out Bob Beck's spamd, and so it seemed reasonable to try to do both at the same time. Also heard from Daniel O'Connor, who suggested using the gif0 interface instead of nos-tun. Installing pf is easy enough: it's now in the base system, so all I need to do is to convert my ipfw rules. That will take longer.
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.
| Top of page | Previous month | Greg's home page | Today's diary entry | Next month | Greg's photos | Copyright information |
