Greg
animals brewing
food and drink gardening
general health
history language
music multimedia
opinion photography
politics Stones Road house
technology
Greg's diary
recent entries
Translate this page
Select day in June 2018:
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:
2017 Sep Oct Nov Dec
2018 Jan Feb Mar Apr
2018 May Jun Jul Aug
Today's diary entry
Diary index
About this diary
Greg's home page
Greg's photos
Network link stats
Greg's other links
Copyright information
    
Groogle

Tuesday, 19 June 2018 Dereel Images for 19 June 2018
Top of page
next day
last day

Max Lehey: Requiescat in pace
Topic: general Link here

I've been meaning to visit my uncle Max and his sister Frieda for at least two months now, and I keep postponing it, at least because I wanted to visit them both on the same trip, and they live over 200 km away from each other.

Today Frieda called up: the problem has simplified itself. Max died in his sleep last night. He had just turned 90 years old.

Damn! Apart from the sorrow about his death, that's another case where procrastination has irreversible consequences. I wasn't overly close to Max, but I had always admired him for his cheerfulness and—as I see it—happy way of life. In that respect he was an example to us all.


Happy anniversary, FreeBSD
Topic: technology Link here

In the morning's mail today:

06/19     Charlie Root <root@FreeBSD.org> born in Portland, Oregon, United States, 1993
06/19     FreeBSD project born, 1993

The first one an insider joke. 25 years already! How time flies!


Lunch with Lorraine
Topic: general, food and drink Link here

Off to Lorraine Carranza's for lunch today. I had been putting off the invitation until after I visited Max and Frieda, but clearly that wasn't a good idea.

This time I took a camera (the Olympus E-PM1) and took some photos.


This should be Yvonne-Lorraine-3.jpeg.  Is it missing?
Image title: Yvonne Lorraine 3          Dimensions:          4108 x 3024, 2716 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, 19 June 2018, thumbnails          All images taken on Tuesday, 19 June 2018, small
Diary entry for Tuesday, 19 June 2018 Complete exposure details

 

She has a really nice view to the north, the real “Grassy Gully” after which the road (some distance away) is named:


This should be Grassy-Gully.jpeg.  Is it missing?
Image title: Grassy Gully          Dimensions:          999 x 270, 155 kB
Make a single page with this image Hide this image
Make this image a thumbnail Make thumbnails of all images on this page
Display small version of all images on this page
All images taken on Tuesday, 19 June 2018, thumbnails          All images taken on Tuesday, 19 June 2018, small
Diary entry for Tuesday, 19 June 2018 Complete exposure details

 

She has also picked up some antiques since we were last there. There's an organ, about 100 years old:

 
This should be Organ-2.jpeg.  Is it missing?
Image title: Organ 2
Complete exposure details
Dimensions: 446 x 606, 147 kB
Dimensions of original: 3024 x 4108, 4236 kB
Display this image:
thumbnail    hidden   alone on page
Display all images on this page as:
thumbnails    this size
Show for Tuesday, 19 June 2018:
thumbnails    small images    diary entry

We still don't know what the purpose of this one was:

 
This should be Cabinet.jpeg.  Is it missing?
Image title: Cabinet
Complete exposure details
Dimensions: 403 x 669, 157 kB
Dimensions of original: 2118 x 3517, 2014 kB
Display this image:
thumbnail    hidden   alone on page
Display all images on this page as:
thumbnails    this size
Show for Tuesday, 19 June 2018:
thumbnails    small images    diary entry


Snipe blunder
Topic: photography, technology, opinion Link here

I've been looking at the M.ZUIKO DIGITAL ED 75 mm f/1.8 lens for a while now. Not the most versatile lens, but I can see a use for it. But the cost is high (new prices round $900), so I have been waiting for a cheap one to come by for under $500.

Today was the day: one appeared on eBay starting at $399. OK, bid up to $499? No, I could get an automatic 5% discount, so I could offer up to $525. Did that via JustSnipe, and then went off to lunch with Lorraine.

Came back, and the lens had sold for $515. But not to me! After much searching, JustSnipe admitted to an “Error #20000”. Thanks, JustSnipe. That's the last time.


More Hugin investigations
Topic: technology, photography Link here

More playing around with Hugin today. Yesterday I used the GUI's “Align” function, but there's also the option to run it directly. My guess was that the difference was due to using “Align”, so today I used cpfind. Surprise, surprise! Only adjacent control points.

So how was this cpfind run?

=== grog@eureka (/dev/pts/28) /photowork/Hugin-build-eureka 92 -> ps aux | grep cpfind
grog       69358  49.9  2.0   801448  671604  -  RNs  12:22pm      0:06.34 cpfind --multirow -o /tmp/ap_res6w6ICm /tmp/ap_inprojv8AcxR

OK, let's run cpfind against that project file (/tmp/ap_res6w6ICm /tmp/ap_inprojv8AcxR, and yes, it was still there). I got all the control points!

But wait. What about that --multirow option? Try it with that. Yup, it suppresses the control points beyond the neighbours.

Why? According to the man page,

       --multirow
           Enable heuristic multi row matching (default: off)

       --linearmatch
           Enable linear images matching (default : all pairs)

What I'm getting here is the result of --linearmatch, not --multirow. And that's not the documented behaviour. But I tried it with the Microsoft version, and sure enough, the behaviour was the same. So now I have:

Still, it's clear now that I'll have to do some code reading. It's interesting to note that --multirow comes directly before --linearmatch in the man page. What if it's like that in the code too, and somehow a break; got lost?


ALDI return day
Topic: general Link here

Time for lots of things to go back to ALDI:

 
This should be ALDI-returns-4.jpeg.  Is it missing?
Image title: ALDI returns 4
Complete exposure details
Dimensions: 447 x 605, 165 kB
Dimensions of original: 3888 x 5258, 4662 kB
Display this image:
thumbnail    hidden   alone on page
Display all images on this page as:
thumbnails    this size
Show for Tuesday, 19 June 2018:
thumbnails    small images    diary entry
 
This should be ALDI-returns-1-detail.jpeg.  Is it missing?
Image title: ALDI returns 1 detail
Complete exposure details
Dimensions: 616 x 438, 133 kB
Dimensions of original: 4334 x 3081, 2521 kB
Display this image:
thumbnail    hidden   alone on page
Display all images on this page as:
thumbnails    this size
Show for Tuesday, 19 June 2018:
thumbnails    small images    diary entry

That's the induction cooker (at the last minute I decided to return it after all), a frying pan that, despite claims, doesn't work with induction, two mixing bowls that Yvonne thought could serve as Chinese soup bowls (they're too big), a “stock pot” (i.e. saucepan) that is in fact perfectly OK except that it's not the size I wanted, the dicer that just doesn't work, and the range hood that I bought at the same time as the induction cooker, and which I decided was more pain to try than the advantage that it promised.


Wednesday, 20 June 2018 Dereel
Top of page
previous day

A new disk for teevee
Topic: multimedia, technology Link here

I'm collecting a lot of video files on teevee that I don't want to erase. And the file system is getting full:

=== grog@tevee (/dev/pts/3) ~ 23 -> df /spool
Filesystem     1048576-blocks      Used     Avail Capacity  Mounted on
/dev/ada0p4         2,784,209 2,721,807    34,559    99%    /teevee/spool

So I bought a 6 TB drive to replace it. I'm gradually getting the routine of this:

  1. Put the disk into a different machine as a second disk, in this case the old tiwi.

  2. Partition as the disk in teevee, with the exception of the larger last partition:

    === root@tiwi (/dev/pts/0) ~ 385 -> DRIVE=ada1
    === root@tiwi (/dev/pts/0) ~ 386 -> gpart create -s gpt $DRIVE
    === root@tiwi (/dev/pts/0) ~ 387 -> gpart add -s 64k -t freebsd-boot $DRIVE
    === root@tiwi (/dev/pts/0) ~ 388 -> # Install boot loader.  Note that -i is the partition number
    === root@tiwi (/dev/pts/0) ~ 389 -> gpart bootcode -b /boot/pmbr -p /boot/gptboot -i 1 $DRIVE
    === root@tiwi (/dev/pts/0) ~ 390 -> gpart add -s 83886080  -t freebsd-ufs $DRIVE
    === root@tiwi (/dev/pts/0) ~ 391 -> gpart add -s 30g -t freebsd-swap $DRIVE
    === root@tiwi (/dev/pts/0) ~ 392 -> gpart add -t freebsd-ufs $DRIVE
    === root@tiwi (/dev/pts/0) ~ 393 -> gpart show $DRIVE
  3. Run newfs. In the case of the /spool file system, which needs significantly different parameters, first get the newfs line from dumpfs on the old machine:

    === grog@teevee (/dev/pts/7) ~ 6 -> dumpfs -m /spool
    # newfs command for /spool (/dev/ada0p4)
    newfs -L Videos -O 2 -U -a 64 -b 32768 -d 32768 -e 2048 -f 4096 -g 2097152 -h 64 -i 1847296 -k 1152 -m 1 -o space -s 5702060032 /dev/ada0p4

    The size (-s) is incorrect for the new file system, but I can just omit it to use the rest of the disk. And of course the device name needs to be checked too.

  4. Mount the teevee file systems via NFS on tiwi and copy the files:

    cd /teevee/spool
    rsync -vaH * /spool | tee -a /var/tmp/rsynced-backup
    cd /mnt            # teevee:/ was mounted here
    rsync -vaH --delete-after * /teevee | tee -a /var/tmp/rsynced-roo

    I did these in parallel. Why --delete-after with the root file system? I had started the first time with tar instead of rsync.

All went well, and when I came back after dinner the root file system had been copied. But what did I see with /spool? Can't copy file: file has vanished!

Looking at what had happened was sobering. Here's what I had some time before dinner:

=== grog@tiwi (/dev/pts/3) ~ 23 -> df
Filesystem     1048576-blocks      Used     Avail Capacity  Mounted on
/dev/ada0p2            19,831    11,756     6,488    64%    /
devfs                       0         0         0   100%    /dev
/dev/ada0p4            19,831    11,020     7,224    60%    /destdir
/dev/ada0p5           412,373   332,095    47,288    88%    /home
linprocfs                   0         0         0   100%    /compat/linux/proc
eureka:/               39,662    27,077     9,411    74%    /eureka
eureka:/home        2,820,026 2,513,661    80,763    97%    /eureka/home
eureka:/Photos      7,629,565 4,604,297 2,948,972    61%    /Photos
teevee:/               64,456    51,685     7,614    87%    /mnt
/dev/ada1p2            64,456    48,650    10,649    82%    /teevee
/dev/ada1p5         5,568,407   421,444 5,091,278     8%    /spool
teevee:/spool       2,784,209 2,721,807    34,559    99%    /teevee/spool

But now I had:

=== grog@tiwi (/dev/pts/3) ~ 23 -> df
Filesystem     1048576-blocks      Used     Avail Capacity  Mounted on
teevee:/spool       2,784,209         0 2,756,366     0%    /teevee/spool

2.7 terabytes of data lost! How could that happen? The log from copying the root file system gave the clue. After completing the sync, I had:

...
deleting spool/ports/
deleting spool/camera/dontdotar
deleting spool/camera/K32.S.01
deleting spool/camera/K32.S
deleting spool/camera/
deleting spool/Videos/Winnetou/Winnetous-Sohn-20180223-193000.txt
...

I had made not one, but three blunders:

  1. I had mounted teevee:/spool under the local tiwi:/teevee, the destination of the root file system copy.
  2. I had specified --delete-after in the sync.
  3. (As I later discovered) I hadn't backed up teevee:/spool for three weeks.

The first blunder was the worst. If I had mounted the file systems sanely, nothing would have gone wrong, and in principle --delete-after makes perfect sense. But as it was, I had deleted more data than I had ever seen (in total) before about ten years ago.

Set to restoring the backup. I'll have to download the rest over the course of the next few days.


eBay blunder
Topic: general, technology, opinion Link here

eBay has had some good discounts recently, coming up to the end of the financial year. In particular, they had 20% a whole selection of household appliances from The Good Guys. I had wanted to buy a new freezer, and the Good Guys were one option. 20% off made them by far the best deal. OK, go for it: select “Buy It Now”, enter the discount code, and press “Pay Now”.

WRONG, FOOL! First you need to press “Apply discount”. And as it was, without so much as a by your leave, I paid the full price, $309.80 too much. Grrrr!

Yes, this is PEBKAC, but the web page could have told me that I had outstanding inputs, like so many others. But there's nothing for it. Contact seller, please cancel, and I'll buy again with the discount.

But of course the only way to contact the seller was by “email”, in other words a web form. Filled that out and then spent a lot of time trying to find a phone number, but no, they're not even in the phone book! I can contact individual branches, but that's all.

OK, call up eBay and see if they can apply the discount. No, sorry, no way, says June. But since it was at least partially their fault (though I don't think she completely understood the details of what I told her), they could give me a discount voucher for $310 (why not $309.80? I got the feeling that the decimals were too complicated). Applicable to a single payment (not necessarily purchase) in the next month. Under the circumstances, that sounded like the best option, so I took it.


Communicating with old fogeys
Topic: technology, general, opinion Link here

Mark Newton came by on the IRC discussion about the fridge purchase, and brought a valid point:

why on earth would they want to pay people to sit in call centres?

I suppose that makes sense, but there should be a way to contact people in emergencies. In this particular case it wouldn't have been so bad if it hadn't been for the fact that the discount expires tonight. He went on to say:

people older than about 50 spent their formative years where the only way they could interact with a company is via phone, and that's the vector they expect. People younger grew up dealing with companies using the web. Over-50s are a small enough minority that it's not worth investing in people and systems to cater for them; if they change their behavior, they stop being a minority at zero cost to the supplier.

OK, I'm over 50—way over 50. How old is Mark? He could be in that age group, or close to it. But on reflection, no, I don't usually like to communicate by phone. I'd far prefer email, and I've ranted about that in the past. It's just on athis occasion, because I needed an immediate response.

Thinking further, people in my age group spent their formative years without the option of getting very much done by phone. 50 years ago I spent many fruitless journeys to people's houses to see if they were there. Either they didn't have a phone, or it didn't occur to me to call them. And if you wanted to do any kind of business, you had to be physically present. It wasn't until I returned to Australia 21 years ago that I discovered I could get some business done by phone. But it's interesting to see that a means of communication has come and gone in that time.


Investigating cpfind
Topic: photography, technology, opinion Link here

OK, why does cpfind --multirow only find control points in adjacent images? Took a look at the code. My guesses weren't very accurate.

Firstly, this is C++, not C, so processing options is an order of magnitude more complicated than a simple switch statement. But round main.cpp(399), I found:

    if(doLinearMatch + doMultirow + doPrealign>1)
    {
        std::cerr << hugin_utils::stripPath(argv[0]) << ": The arguments --linearmatch, --multirow and --prealigned are" << std::endl
             << "  mutually exclusive. Use only one of them." << std::endl;
        return false;
    };

Admire that };, of which there are many. But clearly I didn't have both --linearmatch and --multirow. Looked a bit further and found the processing for multirow, but like everything in C++, it is so convoluted that you need to know all the internals before you start. Gave up and sent a message to the developers' list. We'll see what happens.


Happy anniversary, FreeBSD
Topic: technology Link here

Yesterday, I was told, was the 25th anniversary of the naming of the FreeBSD project:

06/19     Charlie Root <root@FreeBSD.org> born in Portland, Oregon, United States, 1993
06/19     FreeBSD project born, 1993

But no, it seems that the time was based on a message that David Greenman sent:

To: interim@bsd.coe.montana.edu (Interim 0.1.5)
Subject: Re: "386BSD" trademark (fwd)
From: David Greenman <davidg@implode.rain.com>
Date: Sat, 19 Jun 93 17:26:02 -0700

So, like the first datagram over the ARPANET, it really happened on the following day based on UTC. So another day to celebrate!


This page contains (roughly) yesterday's and today's entries. I have a horror of reverse chronological documents, so all my diary entries are chronological. This page normally contains the last two days, but if I fall behind it may contain more. You can find older entries in the archive. Note that I often update a diary entry a day or two after I write it.     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. 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.


Greg's home page This month Greg's photos Greg's links

RSS 2.0 Valid XHTML 1.0!