Previous Up No next

Diary, November 2026



Sun Mon Tue Wed Thu Fri Sat
  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


Sunday, November 1, 2026

65 years

I was born on Wednesday, November 1, 1961 at 9:45 in the morning (CET). That means that today, I will be 65 years according to the calendar and 23376 days old. But how old am I really? The tropical year, the time that it takes for the seasons to return, is said to be 365 days, 5 hours, 48 minutes, and 45.19 seconds. Today, around 3:33:57 in the night, it was 65 tropical years since I was born. But the tropical year is shorter than the sidereal year, the time taken by the Earth to orbit the Sun once with respect to the fixed stars, namely 365 days, 6 hours, 9 minutes, and 9.76 seconds. That means that, tomorrow nigth around 1:40:34 it will be 65 sidereal years after I was born. The average anomalistic year, the time taken for the Earth to complete one revolution with respect to its apsides, is 365 days, 6 hours, 13 minutes, and 52.6 seconds. That means that, tomorrow around 6:46:59 it will be 65 average anomalistic years since I was born.

I have reached the age at which people used to retire in the Netherlands, till almost fourteen years ago, when the age was increased. Now I will official retire on February 1, 2029.

I went into the city and shortly visited the exhibition Stap in! Mijn cultuur with works by from local women from various cultural backgrounds.

I met with some of my family in the cafe Het Bolwerk, the oldest cafe in Enschede, known for being the local pub of Willem Wilmink. The band BlankPaint was performing during our stay. As presents, I got some green tea '宜興茗茶' from Yixing, some cinema vouchers, jar of cherry jam, and a 'tie' with the number 65 as can be seen in the picture below. After a short walk through the city center we enjoyed and Indonesian rijsttafel at Nyonya Meneer. Everybody enjoyed the dinner.


(©Pieter)


Tuesday, November 4, 2026

Endive and red chicory hotchpotch

Tonight's dinner: endive and red chicory (radicchio) hotchpotch, served with pork belly slices and seasoned with a touch of curry powder. A successful combination. The endive, the red chicory and the potatoes are from the harvest of the Herenboeren Usseler Es cooperative.

Walking simulator games

I saw the video Custom Engine Forest Showcase and was bit surprised by how realistic it looks. I do not know whether it is completely procedural generated. Some days ago, I came across the video 44 km Through a 10 Trillion km Procedural World from the Infinitra Project which has much variation but is less realistic. A few days before I saw the video Procedural Planet, which has nice animation of ths sky, while the trees are very simplified, but it is planet sized. I have enjoyed adventure games. It has very varied terrain but no buildings and or roads. Some others that I came across are: I like adventure games where you can walk around. I like the exploration type more than the ones where you have to puzzle to make progress (especially if the puzzling requires you to quickly walk from one location to another) or where there is a lot of narrative. I think what I prefer are so-called Walking simulator games. I wonder whether the best parts of the above could be combined and serve as a starting point for a walking simulator game. Lately, I have been thinking about a type that is kind of a labyrinth, a city on multiple levels, where you can see 'roads' but where it is not immediately clear how to get somewhere. I game that would require good navigation and orientating skills. On itch.io there are many free games for Linux tagged Walking simulator. Just one of them is Wave Function Collapse.


Friday, November 6, 2026

Linear interpolation with integers

Linear interpolation is about finding points on a line defined by two points. Usually it is a (non-vertical) line in the XY-plane or a function defined in time going from one value to another between two points in time. Without loss of generality we could asume that the first point is on the origin and the other point at xy-coordinates (A, B) where A is larger than zero. The y-coordinate of the points on the line can be calculated with (x*B)/A. Now when integers are used, as can be seen in the graph below, you get points that are on or below the line. This is due to the fact that integer division, rounds towards zero. A simple method for implementing rounding, is to add half of A before the division, resulting in using the formula (x*b+A/2)/A, as can be seen with the second option in the animation below. Now the points appear above and below the line and the average error is smaller. The end point can be dragged around with the mouse. In mircocontrollers division is usually a slow operations and especially when one has to calculate many interpolated points, it might be a good idea to avoid the division. Because the B/A remains constant, one possiblity is to introduce factor for that and use that as a multiplication factor. Multiplications are usually more efficient (because it can be parallelized while this is not so easy for a division). However, this does not work very well when B is not much larger than A. Divisions with a power of two can be efficiently implemented in a microcontroller to use a right arithmetic shift operation. The idea is to multiply the factor with a power of two and use a right arithmetic shift in the calculation of a point. This method is a type of fixed-point arithmetic. In the fourth option below, this technique is used. As can be seen the points appear on or below the line. When the end point is moved to the bottom right corner, the points often appear far below the line. The correction that was used in the second option, adding half of the the multiplication factor, does not solve the problem. One solution, given in the sixth option, is to use a much higher factor. This seems to work much beter, but still when it is compared in with the second option, it has difference, which can be visualized with clicking the 'Show differences with best' checkbox. An obvious solution seems be to increase the factor even furhter. However, for some cases this might lead to an overflow during the multiplication, which would result in very large errors as is shown in the last option where the integer values are limited to 16-bits and the value of f is truncated whenever the B value becomes to high. (High values of A do not cause truncation problems.) The additional multiplication factor should be selected careful with respect to the maximum y value and the maximum integer value.

y = (B * x) / A
y = (B * x + A/2) / A
f = A / B, y = f * x
f = (8 * A) / B, y = (f * x) / 8
f = (8 * A) / B, y = (f * x + 4) / 8
f = (256 * A) / B, y = (f * x + 128) / 256
f = (2048 * A) / B, y = (f * x + 1024) / 2048 with 16 bits
Show differences with best


Saturday, November 7, 2026

Museumnacht

This evening, I went to the Museumnacht (Museum night) in Enschede. I visited a number of the locations that joined this year.


This months interesting links


Home | October 2026