Greg
fiction food and drink
health gardening
general history
opinion photography
politics technology
Greg's diary
April 2024
Translate this page
Select day in April 2024:
Su Mo Tu We Th Fr Sa
1 2 3 4 5 6
7 8 9 10 11 12 13
14 15 16 17 18 19 20
21 22 23 24 25 26 27
28 29 30
Select month:
2023 May Jun Jul Aug
2023 Sep Oct Nov Dec
2024 Jan Feb Mar Apr
Today's diary entry
Diary index
About this diary
Previous month
Greg's home page
Greg's photos
Network link stats
Greg's other links
Copyright information
    
Groogle

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:

 
https://lemis.nyc3.digitaloceanspaces.com/grog/Photos/20240401/big/DxO-1.png
Image title: DxO 1
Complete exposure details
Dimensions: 484 x 385, 20 kB
Dimensions of original: 484 x 385, 20 kB
Display this image:
thumbnail    hidden   alone on page
Display all images on this page as:
thumbnails    this size
Show for Monday, 1 April 2024:
thumbnails    small images    diary entry

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:


https://lemis.nyc3.digitaloceanspaces.com/grog/Photos/20240331/big/Wrong-Easter-Lamb-3-orig.jpeg
Image title: Wrong Easter Lamb 3 orig          Dimensions:          5184 x 3888, 3186 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 Sunday, 31 March 2024, thumbnails          All images taken on Sunday, 31 March 2024, small
Diary entry for Sunday, 31 March 2024 Complete exposure details

 

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!

 
https://lemis.nyc3.digitaloceanspaces.com/grog/Photos/20240401/big/DxO-2.png
Image title: DxO 2
Complete exposure details
Dimensions: 661 x 32, 2 kB
Dimensions of original: 661 x 32, 2 kB
Display this image:
thumbnail    hidden   alone on page
Display all images on this page as:
thumbnails    this size
Show for Monday, 1 April 2024:
thumbnails    small images    diary entry

They should look like this:

 
https://lemis.nyc3.digitaloceanspaces.com/grog/Photos/20240401/big/DxO-4.png
Image title: DxO 4
Complete exposure details
Dimensions: 764 x 40, 5 kB
Dimensions of original: 764 x 40, 5 kB
Display this image:
thumbnail    hidden   alone on page
Display all images on this page as:
thumbnails    this size
Show for Monday, 1 April 2024:
thumbnails    small images    diary entry

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.

 
https://lemis.nyc3.digitaloceanspaces.com/grog/Photos/20240401/big/Perfectly-Clear-1.png
Image title: Perfectly Clear 1
Complete exposure details
Dimensions: 757 x 377, 121 kB
Dimensions of original: 757 x 377, 121 kB
Display this image:
thumbnail    hidden   alone on page
Display all images on this page as:
thumbnails    this size
Show for Monday, 1 April 2024:
thumbnails    small images    diary entry

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:

 
https://lemis.nyc3.digitaloceanspaces.com/grog/Photos/20240402/big/Perfectly-Clear-1.png
Image title: Perfectly Clear 1
Complete exposure details
Dimensions: 698 x 525, 21 kB
Dimensions of original: 698 x 525, 21 kB
Display this image:
thumbnail    hidden   alone on page
Display all images on this page as:
thumbnails    this size
Show for Tuesday, 2 April 2024:
thumbnails    small images    diary entry

OK, click on Activate. Yes, it comes up with another window that already knows the license key and email address:

 
https://lemis.nyc3.digitaloceanspaces.com/grog/Photos/20240402/big/Perfectly-Clear-2.png
Image title: Perfectly Clear 2
Complete exposure details
Dimensions: 402 x 225, 3 kB
Dimensions of original: 402 x 225, 3 kB
Display this image:
thumbnail    hidden   alone on page
Display all images on this page as:
thumbnails    this size
Show for Tuesday, 2 April 2024:
thumbnails    small images    diary entry

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 sote
The 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:


https://lemis.nyc3.digitaloceanspaces.com/grog/Photos/20240401/big/Drought-of-March.jpeg
Image title: Drought of March          Dimensions:          4608 x 2782, 4877 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 Monday, 1 April 2024, thumbnails          All images taken on Monday, 1 April 2024, small
Diary entry for Monday, 1 April 2024 Complete exposure details

 

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:


https://lemis.nyc3.digitaloceanspaces.com/grog/Photos/20240402/big/Double-rain.jpeg
Image title: Double rain          Dimensions:          1080 x 2340, 112 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 Tuesday, 2 April 2024, thumbnails          All images taken on Tuesday, 2 April 2024, small
Diary entry for Tuesday, 2 April 2024 Complete exposure details

 
https://lemis.nyc3.digitaloceanspaces.com/grog/Photos/20240402/big/Double-rain-detail.jpeg
Image title: Double rain detail          Dimensions:          1025 x 500, 42 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 Tuesday, 2 April 2024, thumbnails          All images taken on Tuesday, 2 April 2024, small
Diary entry for Tuesday, 2 April 2024 Complete exposure details

 

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:


https://lemis.nyc3.digitaloceanspaces.com/grog/Photos/20240403/big/disaster-performance.png
Image title: disaster performance          Dimensions:          1791 x 1063, 952 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 Wednesday, 3 April 2024, thumbnails          All images taken on Wednesday, 3 April 2024, small
Diary entry for Wednesday, 3 April 2024 Complete exposure details

 


https://lemis.nyc3.digitaloceanspaces.com/grog/Photos/20240403/big/disaster-performance-detail.png
Image title: disaster performance detail          Dimensions:          1026 x 652, 32 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 Wednesday, 3 April 2024, thumbnails          All images taken on Wednesday, 3 April 2024, small
Diary entry for Wednesday, 3 April 2024 Complete exposure details

 

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:


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
next day
last 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:

 
https://lemis.nyc3.digitaloceanspaces.com/grog/Photos/20240405/big/PerfectlyClear-1.png
Image title: PerfectlyClear 1
Complete exposure details
Dimensions: 662 x 352, 16 kB
Dimensions of original: 662 x 352, 16 kB
Display this image:
thumbnail    hidden   alone on page
Display all images on this page as:
thumbnails    this size
Show for Friday, 5 April 2024:
thumbnails    small images    diary entry

“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.


Saturday, 6 April 2024 Dereel Images for 6 April 2024
Top of page
previous day
next day
last day

Aussie: You are not worthy
Topic: technology, opinion Link here

Another day half wasted trying to get Aussie Broadband to fix their configuration problem. Once again a message explaining the situation and asking them to do something about it.

Then called up Aussie, and for a change I was connected to an Australian, Harrison. He wasn't the brightest: I gave him the details at the outset, but he wanted to hear my date of birth no less than three times and CJ's address twice. After a quarter of an hour he still had not managed to identify me, and I asked to be connected to his supervisor.

After a few minutes I was connected to Paul, who also couldn't find the message I sent. I suggested that they had serious problems with their email system and he agreed, yes, the system isn't logical. But even after he found it, and asking my date of birth another 3 times and CJ's address at least once, he refused me access. He wouldn't even admit that they had a customer called CJ Ellis, and he couldn't give me any details—even not why he didn't like my credentials. The privacy laws prohibit that. He would also not tell me where I could turn to solve the problem at hand. In the end, after 42 minutes, I hung up. Aussie (represented by Paul, surname protected for privacy reasons) has refused support for reasons that they don't want to divulge.

So what went wrong there? Did something in their system raise a red flag? Or, more likely, is he completely incompetent and unable to find his way round Aussie's support infrastructure? I think I'd go for the second choice. I thought of calling up again and seeing if the next “consultant” could find his way round the system better. But I couldn't stand the pain. Instead I sent CJ an email:

I've spent another 1½ hours today trying to resolve your phone problems. Unfortunately, today they have decided to refuse to let me work on your behalf. I spent 45 minutes on the phone today and was brushed off by a supervisor who didn't like something about my credentials, but who was too polite to say what.

So: I don't know where to go from here. Your lines are still down, and Aussie is too proud to do anything about it. The best you can do is send them an email and tell them what you think.

Sorry that I recommended Aussie to you. I won't do that to anybody again. My only excuse is that they once used to be a good company, but they're doing everything they can to change that. I'll look for an alternative VoIP service for you.

What went wrong here? As I said to CJ, once they were a good ISP. Now they've become inefficient bureaucrats who would make Telstra envious. And of course the faults haven't been fixed. But why did this happen? It happened to Internode, also once a bright star in the ISP sky. As Simon Hackett put it, How to ignore a customer without even trying. Ironically, that experience drove him to Aussie. I wonder if he's still as happy with them now. I note that the problems I have now are similar to what I complained about last September, just worse. And the solutions that I mention would still apply. The only thing that I would add is that support people seem to be set to believe that all problems are misconfiguration of customer hardware, even when there's hard evidence to the contrary. Yes, it's difficult to find good support personnel, but that's what escalation procedures are for.

So: what do I do now? I should drop Aussie like a hot potato, but who knows if there's a better ISP? And how do I get my /24 routed? But it does look like a good idea to look for a better VoIP supplier. Mañana.


disaster? despise!
Topic: technology, photography, opinion Link here

Yvonne had some photos to process yesterday, and she did it ondisaster, which worked well. I checked the processing times, and they seemed significantly faster than her normal processing on distress, but I forgot to write them down—something like one photo per minute, because all her photos are done with DxO PhotoLab's DeepPRIME feature. But that increase in processing speed didn't seem as dramatic as I had noticed

But why limit myself to 8 CPUs? hydra has 32 logical CPUs. I could easily increase the number on distress to 16.

And, probably, invalidate my license keys. OK, how about cloning it? And yes, that's easy. Shut down the machine, right click on it, select Clone and follow the instructions. And without much ado I had a new machine, which Yvonne decided to call despise.lemis.com. But DxO PhotoLab decided that my license key was invalid, and so did Microsoft. I have DxO's license key written down, and their support team was relatively responsive with the key for “ViewPoint” (which I haven't applied yet), so once I'm happy we could stick with despise.

And how did Yvonne go? Worked out of the box modulo setting DxO in 30 day trial mode. And it was much faster. The only issue was memory: it only had 16 GB of memory, and running 14 concurrent conversions (the maximum) it was hitting memory limits. So I should probably add another 8 GB of memory.

My own access to the machine was less satisfactory:

=== grog@hydra (/dev/pts/29) ~/Photos/20240405 942 -> /home/local/bin/dordesktop despise 3790x2110+0 &
Connecting to despise

ATTENTION! The server uses and invalid security certificate which can not be trusted for
the following identified reasons(s);
...

Do you trust this certificate (yes/no)? Core(error): Failed to read response from stdin
Segmentation fault (core dumped)
Disconnected from despise, status 139

And that was repeatable. rdesktop really dumped core for no obvious reason. But that, too, will find a solution.


Sunday, 7 April 2024 Dereel Images for 7 April 2024
Top of page
previous day

DST ends
Topic: general, photography, technology, opinion Link here

Daylight Saving Time ended today. Lots of clocks to change, including my cameras.

In days gone by I have used this horrible OI.Share, mobile phone app to set the time on the OM-D cameras. But it's such a pain. It has the advantage of setting it almost exactly correct (why do mobile phones, which use NTP, still have times that vary by a couple of seconds?). But is it that important? Yvonne's Olympus OM-D E-M5 Mark III proved to be 3 minutes fast. That's 1 second per day since DST started, but should I (or Yvonne) constantly reset it? I'd be for resetting only when it really makes sense, and then I can do it with programs after the event, like this case 14 years ago


https://lemis.nyc3.digitaloceanspaces.com/grog/Photos/20100417/big/Evening-19.jpeg
Image title: Evening 19          Dimensions:          3008 x 2000, 528 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 Saturday, 17 April 2010, thumbnails          All images taken on Saturday, 17 April 2010, small
Diary entry for Saturday, 17 April 2010 Complete exposure details

 
https://lemis.nyc3.digitaloceanspaces.com/grog/Photos/20100417/big/Evening-7.jpeg
Image title: Evening 7          Dimensions:          3024 x 4032, 2176 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 Saturday, 17 April 2010, thumbnails          All images taken on Saturday, 17 April 2010, small
Diary entry for Saturday, 17 April 2010 Complete exposure details

 

Setting the time manually was so much faster, though several cameras managed to change 2 hours instead of one, presumably a key bounce issue. What I did discover, however, was that the date on the E-30 was wrong. Some time last month it seems that it decided that the month was December, as the mdir output shows:

53102888 ORF  10993760 2024-03-10  13:25
53102889 ORF  13378519 2024-03-10  13:25
53152890 ORF  10993854 2024-03-15  13:25
53152891 ORF  13225868 2024-03-15  13:25
53162892 ORF  10993377 2024-03-16  13:25

Was it only the month? The time shows 13:25 every time, mean solar noon, so it probably was. I wonder how that happened. But the camera is now over 15 years old, so I'll just keep my eye on the date when I turn it on.


More fun with Aussie
Topic: technology, opinion Link here

To my surprise, two messages from Aussie Broadband support accounts today. One was an acknowledgement that I had complained about not being allowed access to CJ's account yesterday with no suggestion about what went wrong. That's enough. A stiff reply with a threat to take the matter to the TIO if they don't respond by Tuesday.

The other message was more to the point:

Unfortunately we are unable to access the modem as our logs are show a
connection to a dell devices at the moment.

Are you able to confirm the Ethernet coming from the NBN box is connected to
the WAN port on the modem.

This silly “modem” word again. Clearly he's talking about the Netcomm router. But that does look strange. CJ's machine is made by Dell (how did they identify it? MAC address?). Has he wired it up wrong. Sent him off a message:

You'll have seen this message below, which I copied you on.  I think
they're still barking up the wrong tree, but one thing makes me
wonder: they talk about seeing a Dell device.  Your computer is Dell,
right?  Are you sure you connected the cables up correctly?  As I said:

> 1.  NTD connected to a Netcomm router (via Ethernet, of course).
> 2.  Router connected to a Microsoft computer and two phones.

In English: the box on the wall is connected to the red connector on
the router.  The computer is connected to one of the yellow connectors
on the router.  Is this what you have done?  If not:

1.  Write down how you connected the cables, and tell me by email.
2.  Fix it.

This should be done as quickly as possible.

And the response? A total of three, all saying nothing and not answering what I thought were clear instructions, except to say that he had connected the computer directly to the NTD to get any connection at all, and that the phones (on the router) were still not working. By the end of the day we still had nothing. I'm beginning to think that the current issues are of CJ's making.


Baked chicken thighs again
Topic: food and drink, opinion Link here

Baked chicken thighs again today. We did them two months ago in the “hair dryer” air fryer, and they came out more or less as planned: meat temperature 78° after 21 minutes at 210°:


https://lemis.nyc3.digitaloceanspaces.com/grog/Photos/20240201/big/Chicken-thighs.jpeg
Image title: Chicken thighs          Dimensions:          5263 x 3888, 4314 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, 1 February 2024, thumbnails          All images taken on Thursday, 1 February 2024, small
Diary entry for Thursday, 1 February 2024 Complete exposure details

 

But they seemed to have required a higher meat temperature. OK, as planned, this time 30 minutes to achieve 82°.

That took 13 minutes, and the thighs didn't look as brown as they could. Clearly I had placed the temperature probe in the wrong place. Turned off the thermometer and went by eye, stopping when they seemed OK:


https://lemis.nyc3.digitaloceanspaces.com/grog/Photos/20240407/big/Roasted-thighs-2.jpeg
Image title: Roasted thighs 2          Dimensions:          5184 x 3888, 5806 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 Sunday, 7 April 2024, thumbnails          All images taken on Sunday, 7 April 2024, small
Diary entry for Sunday, 7 April 2024 Complete exposure details

 

That was after about 18 minutes! And certainly they were cooked enough. The thermometer showed a temperature round 95°. They tasted fine, but I'm still dissatisfied. Why the difference? These thighs were smaller, but that wouldn't significantly affect the time it took to brown them. Maybe the start temperature (21°)? I didn't note it last time.


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

Valid XHTML 1.0!

$Id: diary-apr2024.php,v 1.7 2024/04/08 04:06:44 grog Exp $