|
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Monday, 1 April 2024 | Dereel | Images for 1 April 2024 |
| Top of page | ||
| next day | ||
| last day |
|
distress: disaster?
|
Topic: technology, photography, opinion | Link here |
Into the office first thing this morning and wanted to continue with yesterday's photo processing. That involves waking distress.lemis.com, the ThinkCentre running Microsoft, which is usually hibernated. That goes via rdesktop, so I first hear the sound of the DVD drive seeking.
But then nothing. More problems with this bloody Microsoft? No, the machine wasn't running at all. Even the power light was off.
More searching, and ultimately came to the conclusion that the box had some hardware failure. Not to worry, I have a number of these boxen. Took out the disk and put it in another one, same model. And it came up happily with no repairs needed: after all, it had been hibernated. Success!
Well, not so fast. When I started up DxO PhotoLab, it wanted me to enter the license key. Why? It's identically the same system that I have been using all along:
|
|
|||||||||
Huh? I have licensed the thing already. OK, dig out the license key and painstakingly type it in, something that I have hated for decades. And it worked, so clearly I didn't mistype.
But things still weren't right. My first task was to straighten this image:
Yes, it's on its side, but it takes a single keystroke (c-l) to fix that. The real issue, a little slower, is to get it exactly rectangular. That's straightforward. Select one of the almost descriptive symbols at the top of the window. But they were gone!
|
|
|
|||||||||
They should look like this:
|
|
|
|||||||||
Why? Something in the back of my head said “ViewPoint”. Right, that could be it. Back to my list of keys, searched the DxO menus, and luckily found a Help item “DxO ViewPoint → Activate DxO ViewPoint”. And yes, that did it. But what a pain!
Still, “Perfectly Clear“ didn't want to be outdone. It, too, decided that I wasn't registered.
|
|
|||||||||
But unlike DxO, it didn't give me the option to register! How do I register? It didn't offer the possibility. After some searching on the web, found a way: select the barely visible “About” towards the right at the bottom of the main window, which brings up this window:
|
|
|||||||||
OK, click on Activate. Yes, it comes up with another window that already knows the license key and email address:
|
|
|||||||||
But the “ACTIVATE” button was greyed out, and nothing that I could do persuaded it to let me submit the data! Off searching for answers, and finally found a way to send a message to their sales team (support has ended). “Please prove that you are human and fill out this CAPTCHA”! And yes, EyeQ, that's the correct information. See if I care. Clearly nobody can steal it.
Damn you, EyeQ, and all those who insult people like that. I won't fill out a CAPTCHA, ever again. Sent off a rather stiff email to support@eyeq.photos and sales@eyeq.photos, asking what to do and telling them what I thought of their attempts to insult their customers. And support@eyeq.photos bounced. Yes, my product has reached end of life (but should work for ever), but surely they have other customers.
And that was only half of the pain. What about Yvonne? She has always had issues; clearly Microsoft likes her no more than she likes Microsoft. But in fact things went relatively smoothly, once we persuaded the bloody thing to react at all. Trying from my system and a shell, I got
=== grog@hydra (/dev/pts/26) ~/Mail 79 -> rdesktop -u yvonne distress
ATTENTION! Found a certificate stored for host 'distress', but it does not match the certificate
received from server.
Review the following certificate info before you trust it to be added as an exception.
If you do not trust the certificate the connection atempt will be aborted:
Subject: CN=distress
Issuer: CN=distress
Valid From: Sat Mar 30 13:06:41 2024
To: Sun Sep 29 12:06:41 2024
Certificate fingerprints:
sha1: e5821dd31e6a4e640b95ee93a1ec815430fa7304
sha256: edb83be75b8c33cb86e31fe2d40b1ce8fa95842e1f0e90216b024f5073407ebc
Do you trust this certificate (yes/no)? yes
Connection established using SSL.
disconnect: Logout initiated by user.=== grog@hydra (/dev/pts/26) ~/Mail 80 -> rdesktop -u yvonne distress
Core(warning): Certificate received from server is NOT trusted by this system, an exception has been added by the user to trust this specific certificate.
Connection established using SSL.
Clipboard(error): xclip_handle_SelectionNotify(), unable to find a textual target to satisfy RDP clipboard text request
Protocol(warning): process_pdu_logon(), Unhandled login infotype 1
That must have something to do with the change in something, and by comparison it's acceptable. After a bit of playing around, things worked, but only when I wasn't already logged in. Previously it would produce a popup asking “shall I disconnect grog?”, but now it just silently fails. And it seems that it no longer goes to sleep, so I have to explicitly put it to sleep (which, I discover, works with just a handful of mouse gestures).
So: things are almost working again, when there should have been no problems whatsoever. But this is Microsoft. And why did the programs want revalidation now? It wasn't the disk; it came up out of hibernation as if nothing had happened. And it's barely the hardware. The new machine has a PCI graphics card in addition to the on-board graphics (it was once teevee.lemis.com), and it has only 16 GB of memory, and of course the Ethernet interface has a different MAC address, requiring updates to /etc/ethers:
--- ethers 2023/02/18 00:33:55 1.10
+++ ethers 2024/04/02 01:36:41
@@ -5,4 +5,4 @@
6c:f0:49:09:7a:4d teevee
00:21:86:21:ab:7e despair
00:21:cc:d0:9e:0e euroa
-44:39:c4:90:52:20 distress
+6c:0b:84:04:0a:5c distress
But clearly there's something about the hardware environment that triggers the license check. What is it? And why? It would make more sense to store the activation information on the disk. As it is, there's clearly a loophole: change the hardware once a month and you can use DxO for free.
One thing's clear: I really need to migrate distress to a virtual machine, as I had planned when I built hydra. That way I'd save power, probably have more processing power, and I could work around this kind of pain.
|
Bibi castrated!
|
Topic: politics, general, fiction | Link here |
The news of the day: Bibi has been castrated! It seems that it was just completing the job started 11 years ago. As the punchline of an unrelated joke goes, would you eat that pig all at once?
Oh, and admire the delicacy of the reports. They've changed “castration” to “hernia operation”.
|
Whan that Aprille with his shoures sote
|
Topic: general, history, gardening, opinion | Link here |
Sixty years ago I was preparing for the GCE 'O' Level exams. One of the set works for English Literature was the Prologue to the Canterbury Tales, and in a fit of misunderstanding I thought I could pass the exam by memorizing the entire prologue (I failed).
But the weather today reminded me of the first two lines:
Whan that Aprille with his shoures soteThe droghte of Marche hath perced to the rote,
In modern English: when April with its sweet showers has pierced the drought of March to the root.
And somehow that seemed to fit. Here a photo taken this morning:
|
|||||||||||||||||||||||
According to the Bureau of Meteorology, 0.8 mm of rain over the whole month at Ballarat Aerodrome, the closest measuring station to us. By contrast, the average rainfall in March is 42 mm, and in April 51.2 mm.
But now it's April, and as if to honour Geoffrey Chaucer, it rained heavily. The Bureau measured 46.8 mm, 58 times as much in the whole of March, or 1,755 times the March daily average.
That's their view, of course, and they told me so:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
But despite their claims, that's the rainfall for Ballarat, not Dereel. We measured only 28 mm, still a welcome change and a refill for our water tanks. On the other hand, I measured 5 mm of rain last month.
| Tuesday, 2 April 2024 | Dereel | Images for 2 April 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Another disaster
|
Topic: technology, photography, opinion | Link here |
Response from not one, but two people at EyeQ today, saying that the license details were correct, but because the product is no longer supported, they can't do anything about it. They did offer me a discounted upgrade to Radiant, the new replacement, but that would still cost money.
OK, one thing's for sure: changing hardware messes up licenses. So let's do what I had been planning all along and set up a virtual machine on hydra. What shall I call it? Yvonne was for discard or dismiss. She would have liked disgust too, but we've already had that one. But by the time I asked her, I had already started with the name disaster.lemis.com, so one of the next ones will be Yvonne's choice.
First step, of course: make a copy of distress' disk. Put it in dereel, which I haven't used for months, and do a dd, which went at the reassuring speed of about 103 MB/s, effectively wire speed.
Then create a new VirtualBox VM. Last time I tried this, I ran into confusion between hydra and eureka. But once I started the right version of VirtualBox (on hydra), and also discovered how to include a native disk image (it must have a name ending in .hdd), things went relatively smoothly, and soon I had it up and running. Microsoft didn't like the change of IP address, but it didn't know why, just “Windows can't find the problem”. People, “Windows” is the problem. And everything I tried Just Worked—quite a success story. Doubtless the devil is in the detail, and I'll find other issues.
|
50 years of microprocessors
|
Topic: history, technology, opinion | Link here |
It has been well over 50 years since Intel introduced its first microprocessor, but for me the first one was the 8008. Spent a while comparing it with my most recent processor:
| Feature | I8008 | Ryzen 7950X | factor | |||
| Introduction | April 1982 | September 2022 | ||||
| Transistors | 3,500 | 13,140,000 | 3750 | |||
| CPUs | 1 | 32 | 32 | |||
| Clock speed | 0.0005 GHz | 4.5 GHz | 9000 | |||
| MIPS | 0.03 | 220,000 | 7333333 | |||
| Process size | 10 μm | 5 nm | 2000 | |||
| Die size | 16 mm² | 264 mm² | 16.5 |
At first I was confused by the process size. 10 and 5? Oh, different units. The entire I8008 processor could be
And in the course of my investigations, came across this detailed description of the chip layout, which looks very interesting. I should read it some time.
| Wednesday, 3 April 2024 | Dereel | Images for 3 April 2024 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
Where's my New York Times?
|
Topic: general, technology, opinion | Link here |
No surprisingly, I got no response to my open letter to the New York Times. During the week they relent and keep their 1-800 phone number manned beyond dawn until 10:00 (in the morning), so called up there, where a grating artificial voice asked me for the 10 digit code. Huh? What's that?
Somehow worked past that and was connected to Robert, who told me that all was OK with my subscription. I explained (hopefully to his comprehension) that I wasn't getting any email. He said that he would contact technical support (medium wait), after which he said that I should now be receiving email. Despite my requests, he didn't say whether they had found any issue. So now I just need to wait
|
Netanyahu: Sorry, killed wrong aid workers
|
Topic: politics, opinion | Link here |
The world (or a any rate, USA, Canada, UK, Poland and Australia) are up in arms (no, that's an inappropriate metaphor) about the killing of 6 aid workers. Even Anthony Albanese got up on his hind legs and condemned the killing of Lalzawmi "Zomi" Frankcom. The Poles lamented the death of Damian_Sobol. The USA and Canada mournedJacob Flickinger. Nobody mentioned Saifeddin Issam Ayad Abutaha.
And for the first time ever, Bibi admitted a mistake. They had killed non-Palestinian aid workers! That can happen in war, especially if three separate cars in a convoy at some distance from another are targeted by precision missiles, but it should only happen to the Palestinians.
Bibi, does anybody outside Israel believe your drivel? You're actively engaging in genocide. This strike succeeded in removing 200 tons of greatly needed food from Gaza. To quote from this article
“This is not an isolated incident,” said U.N. humanitarian coordinator James McGoldrick, citing the killing of at least 196 humanitarian workers in the occupied West Bank and Gaza since October. “This is nearly three times the death toll recorded in any single conflict in a year.” The 196 includes more than 175 U.N. staffers, U.N. Secretary General António Guterres said Tuesday.
When will people come to the recognition that Israel's barbarism is unacceptable and not even useful to them. No wonder Hamas attacked them. Not a justification, of course, but you can't make people like you by destroying them.
|
CJ off the air
|
Topic: technology, general, opinion | Link here |
Mail from CJ Ellis today:
Hi Greg It seems that I have had no phone signals since we last spoke & you
had been in touch with Broadband for me to be able to have message bank ..
I thought it might have been the storm that has stop my phone , from having
signals , but to even now there is nothing there .. Sorry to trouble you
,but it has been a saga with broadband since we have changed to them with
the phone service ..
What's wrong now? Clearly the loss of VoIP has nothing to do with the bad weather. Checked the status: both lines registered. Tried to call him. Nothing. After 2 minutes, the attempt timed out, but it should have given some other indication before that.
Dammit, call Aussie Broadband support again, and was connected to somebody who on asking admitted to the name Semanjit Kaur. Once again this incredible difficulty that all support personnel seem to have with handling calls on behalf of another customer. At least this time I'm on their list as an approved contact, but she had trouble with the phone number. I had a number ending in 3, from their web site and also my own records. She found the same number with a 2 in the last position.
More puzzling, then she decided that she had been looking in the wrong place and that the number I had was correct. For some reason she didn't try to call it. Instead she wanted to check the router configuration. Can't access it. Oh yes, I can. I'll update the dial plan.
I explained to her that the dial plan had nothing to do with incoming calls, but she thought it would help anyway. OK, but it meant getting CJ to power cycle the router, so that was all we could do. I sent a mail message to CJ explaining how to power cycle it, and that was all that we could do for today.
Does this really have to be this difficult?
|
Bigger disaster
|
Topic: technology, opinion | Link here |
Carried on playing around with disaster.lemis.com today. It was my intention to use it seriously, and 1 CPU wouldn't cut it. But hydra has 32 logical CPUs, so we could easily allocate 8 to disaster. Did that, and fired up DxO PhotoLab. Please activate. OK, we're getting used to that. But when I tried to activate “ViewPoint”, it told me “too many activations”! Three activations on the same disk! Another support request, and all I could do for today.
But I did try processing Saturday's photos. 140 images processed in 5 minutes! 28 images per minute, where I previously only had about 5. That's really worthwhile. Only: for reasons I haven't understood, it uses stupid amounts of CPU time when idle. Here disaster and eureso (FreeBSD, reinstated to see if I could get it to work):
PID USERNAME THR PRI NICE SIZE RES STATE C TIME WCPU COMMAND
37150 grog 37 20 0 17G 17G select 28 493:14 241.39% VirtualBoxVM
37024 grog 32 20 0 2501M 2244M select 8 2:53 0.87% VirtualBoxVM
It wasn't like that all the time; it fluctuated wildly and dropped as low as 20%, still unreasonably high. But that's not that important: I can put it to sleep most of the time. Could it be something to do with GPU emulation?
During this time, the task manager showed something interesting:
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
The top output coincided with the 0% usage, but for non-obvious reasons it was using up to 25% CPU, corresponding to 2 full CPUs. Why? Somehow this must have something to do with virtual machines.
| Thursday, 4 April 2024 | Dereel → Geelong → Dereel | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
|
More Aussie VoIP pain
|
Topic: technology, opinion | Link here |
So why didn't CJ Ellis contact me about the success of power cycling his modem? Checked the status page. Both phones not registered. Did something go wrong with the router? Is he maybe completely off the net?
What a time! I was in a hurry because of an appointment in Geelong, but called up Aussie support and was told that there was a two minute wait—every two minutes for 8 minutes. Finally I was connected with Pranil, who asked all sorts of questions that I didn't know (CJ's mobile phone number, for example, which for Aussie's purposes doesn't exist). Finally I got through to him that I just wanted to check whether CJ was on the net or not, but he kept returning to VoIP configuration. Finally he checked: yes, he was unable to establish contact with CJ's router. That probably means that it was off the air. What are you going to do about it? It looks like a misconfiguration on Aussie's part. Can somebody go on site? Sure, when do you want to go? Oh, they expect me to drive the 55 km and do the work! To fix something that they had done!
Yet Another formal complaint, correcting his text that referred to VoIP rather than the network link (“Internet”), with a request that this time they would send me the text of the complaint. And that would take up to ten days to resolve! He said that he would put it on priority, but that didn't help. 45 minutes on the phone for something that didn't even concern me.
|
Cataract surgery follow-up
|
Topic: health, opinion | Link here |
Off to Geelong today for another eye examination. It was much slower than last time, and it took over half an hour before Bridget, the orthopist, called me in. Before she got a chance to do anything, I mentioned the apparently incorrect focus of my right (“new”) eye. She checked me and found that yes, I had an offset of -0.5 dioptres spherically—almost exactly what I had guessed (well, -0.5 to -0.6). But she also found astigmatism of +1 dioptre at 100°. That doesn't match any miscalculation in the infinity point, and Bridget thought that it could be due to continuing inflammation of the eye. It seems that it takes 4 weeks for complete healing, thus presumably the timing of the drops.
It wasn't until we were done that I discovered that that was about all. Presumably she was there to measure my eye. But I had thought that today's appointment was a precursor to next week's operation on my left eye, when in fact they weren't interested in that at all.
Then, with the inevitable dilating drops, into see Shivesh Varma, David Fabinyi's locum tenens, the man with the handshake of steel. He put me through the usual tests, noted eye pressures of 12 mm Hg (left) and 13 mm (right), apparently excellent values. I mentioned the difference of focus, and he checked through the lists and came up with a value of -0.45 dioptres. I asked him to ensure that this didn't get applied to my left eye, and he said that he would ask David to call me before the procedure.
After we had left, got a phone call from him: in fact, my spherical correction is effectively 0 dioptres, since the cylindrical correction is +1 dioptre, giving a focus range from -0.5 to +0.5 dioptres. That's an interesting consideration; maybe things will clear up when the eye has finished healing. In the meantime, of course, there's the question of where the -0.45 dioptres comes from. Still worth a talk with David, though I really don't want to wait until the right eye has finished healing.
|
More Aussie pain
|
Topic: technology, opinion | Link here |
Back home at 15:00 and tried to catch up with the day's work. First this diary. I was in the middle of describing yesterday's pain with CJ's phone when the doorbell rang. CJ, with a car full of hardware. It seems that he didn't just power cycle the router, he disconnected a number of cables. Did he put them back correctly?
Anyway, took the router in and connected it to dereel, which is seeing more use lately. All works, the phones register, as the Aussie status page shows. And I still can't call in! So effectively we're where we were before I called Aussie the first time.
More investigations: I can make some outgoing calls. I could call my mobile phone, but not my “home phone” (also VoIP with Aussie Broadband). So something is still wrong. What do I do now? Postpone. Sorry, CJ, I know that you went with Aussie on my recommendation (something that I will never do to anybody again), but I don't have time. Leave the hardware here and I'll look at it tomorrow.
But there are so many things that don't make sense here. I need to add to this list as I discover them:
Clearly Aussie's support staff are incapable. Why reconfigure a dial plan if the issue is with calling in? Did Semanjit understand what she was doing?
What did CJ do? He seems to be having comprehension issues. He claimed that Aussie's voice menu didn't give him the choice of selecting “support”. Tried it from his phone. Works as expected. And what did he disconnect yesterday, and how did he reconnect it?
Looking at his call logs was strange. There must be something wrong with this: Why does the usage log show calls from CJ to himself?
| 18-03-2024 | 0395662250 | 2m 14s | $0.15 | |||
| 01-04-2024 | 0353182413 | 4m 38s | $0.00 |
That's a call from 0353182413 to 0353182413, the same number. 4½ minutes long! How can that work? On the same day, we have the same thing on the other line, 0353244269:
| 01-04-2024 | 0353244269 | 1m 44s | $0.00 |
They're too polite to give times (something that MyNetFone did much better), but what does it mean? And of course there are no recent calls; for some reason it takes up to 48 hours to update the call register, something that should happen in real time.
|
named death
|
Topic: technology, opinion | Link here |
As if I didn't have enough to do, after CJ left I discovered that I hadn't received any external mail for nearly 6 hours. Local mail was still coming through, so checked things on lax.lemis.com (also mail.lemis.com). At first sight things looked OK. Nothing queued bar the usual messages from MAILER_DAEMON to non-responsive spam sites, maillog showed the usual rejections because of name lookup issues. Sent a message from freebsd.org. It didn't arrive!
A bit more searching and I discovered that all DNS lookups were failing: named had stopped. Why? No idea, except that I added disaster.lemis.com to the configuration yesterday. Nomen est omen? Restarted and things worked normally. Another thing to check on when I have time.
| Friday, 5 April 2024 | Dereel | Images for 5 April 2024 |
| Top of page | ||
| previous day |
|
Aussie: no support any more!
|
Topic: technology, opinion | Link here |
So what's wrong with CJ's VoIP service? The obvious thing to do would be to try it on my own ATA and see what happened. And indeed, it still didn't work. The same symptoms: call in rang, but the caller got a busy signal, and it stopped trying to call after a few seconds. Clearly a service configuration issue.
So I sent Aussie a summary of what has happened so far, and then called them at 14:20, and was connected to a very quiet Mohammed after only 4 minutes. The same old pain, at least 10 minutes of identification before we even got started. But Mohammed couldn't find the message I sent. What address did I use? support@aussiebb.com.au. Ah, that's an old, worn-out magic word. Now I should send support requests to accounts@aussiebroadband.com.au. OK, did that, but where did the first message go?
While waiting for the new message to arrive, asked why I hadn't had a confirmation of my complaint. It seems that they sent it to CJ, who is off the air!
Finally I got a response to my first support request:
Date: Fri, 5 Apr 2024 03:27:48 +0000 (UTC)
<support@aussiebb.net>: host smtp1.wide.net.au[121.200.0.25] said: 550 5.1.1
<support@aussiebb.net>: Recipient address rejected: User unknown in local
recipient table (in reply to RCPT TO command)
Return-Path: <grog@lemis.com>
Received: from eureka.lemis.com (121-200-11-253.79c80b.mel.nbn.aussiebb.net [121.200.11.253])
by lax.lemis.com (Postfix) with ESMTP id 3B3A328088
for <support@aussiebb.net>; Fri, 5 Apr 2024 03:18:11 +0000 (UTC)
On the one hand, that's symptomatic of what has become of Aussie: no support any more. But look at the times! It took over 9 minutes for their mail system to reject the message!
So he read the email and asked for CJ's email address, which is a little difficult for non-native English speakers (though Mohammed spoke excellent English). He decided to test the connection on their “test bench”, whatever that may mean. And it passed. So it must be the router.
What router? This line was configured on my ATA. I tried to make it clear to him that if the problem moves with the service and not with the ATA, it must be a configuration problem with the service, not the ATA. But I failed. He kept asking his “level 2” people, who presumably have little more understanding than he. And this despite the request in my email:
Please forward this to your VoIP specialists.
He went through all sorts of rigmaroles, including another video view of the ATA configuration screen. I've seen this before, so I put the phone on a tripod, at which point it obligingly rebooted.
More talking, and at some point he said that the service couldn't work at my address because VoIP is tied to the Internet service. That's nonsense of course, and I tried to explain to him how VoIP works. I don't think I succeeded.
Then he went back to the ATA configuration screen. It has frozen, he said. Clearly he had never seen things on a tripod before.
After understanding that, he created a new service with the same number, so all I had to do was type in the new password. It ended in l, but he had read it as 1. And it registered anyway, or at least the ATA claimed that it did. Can I factory reset the ATA? Emphatic NO! Somehow I can't get it into their heads that the problem can't be with the ATA, and think of the damage it would do.
In the end, we agreed that CJ would pick up the equipment and take it home with him, and that they would mess with the configuration there—he still thinks that it's an ATA configuration issue!. But at least it gets it out of my hair.
Still, I could give CJ one of my spare VoIP lines to tide him over. Took a look at his router again. Both lines registered! Tried them out. Yes, they work! And so does the one I have on my ATA.
How can that happen? My best guess is that somebody at Aussie fixed things without telling me. Still, clearly good news. CJ came and picked up his hardware and went home. Called me at 18:45. Both lines down! In addition, I checked my lines. Now the one I had turned off for the ATA test is showing the same symptoms: call in rings, but the caller gets busy signal. And call out works.
What does all this mean? It could be that Mohammed was correct after all: the VoIP configuration depends on the IP address, maybe a misguided attempt to lessen abuse. That would also explain why I was never able to configure Aussie VoIP on my mobile phone. All of this points to a configuration issue on the SIP server or proxy, which happen to be the same machine. Still, sufficient to the day is the evil thereof. So far this week I must have spent at least 8 hours fighting this problem, and all sorts of other things have had to wait as a result.
|
disaster processing photos
|
Topic: photography, technology | Link here |
disaster.lemis.com seems to work well now, though I still need to confirm my software licenses. But one was a surprise:
|
|
|||||||||
“Perfectly Clear“ has decided that it's licensed after all! I need to tread carefully now to ensure that it doesn't change its mind.
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 | Greg's photos | Copyright information |