Previous Up No next

Diary, January 2022



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  31


Saturday, January 1, 2022

13.5° Celsius

The temperature at Twenthe Airport has gone up to 13.5° Celsius, which breaks the previous record of 12.9° on this date in 2012.


Wednesday, January 5, 2022

Memory upgrade

I got a 2Gbyte DDR3 SDRAM memory module from a fellow TkkrLab member and installed it in FJF2 following the YouTube video How to upgrade ram of Acer Aspire One D270. Easy! Do it yourself. I ran some memory test to check the memory module before closing it. Reconnecting the keyboard (twice) was the hardest part.


Friday, January 7, 2022

Paintings and such

Today, we spend some time cleaning the former home of my mother now that we are in the process of selling the house. I took some paintings and objecs with me. Among them:


Tuesday, January 11, 2022

Tropy

I have been looking at the Tropy application to see if I could use it for my research on Peter Struycken. It has some interesting features. It allows you organize your images and allows you to add meta data and notes to these, including selections of them. You can export the data as JSON and the notes are stored with a subset of HTML, which makes it possible to parse them. I want to extract data from the images and have a way to relate, to correlate, and to condens them into well established and supported facts. I also want to go from the facts back to the data and the images they are based on. It seems that Tropy does not support this. This afternoon, I discovered that it is possible to define templates and base these on all kinds ontologies. I had some look at Linked Open Vocabularies and the GND ontology. These are based on Resource Description Framework. (On Hacker News there was some discussion about the semantic web.) I find this model a little limited and rather cumbersome to work with. I thinking about representing the data in the images with (nested) bullet lists in the notes and write some software to extract these.


Friday, January 14, 2022

2-state 4-color BLB

Last Wednesday, I read the blog Another New Record in Self-Cleaning Turing Machines, which is about a new record breaking 2-state 4-color Turing Machine that is a Blanking Beaver, a kind of Busy Beaver. In the past two days, I have been trying to analyze it. This evening, I found that it calculates the following Collatz like sequence:
    n = 1; m = 1;
    while (true)
        if ((n % 2) == 1)
        {
            if (m == 2)
                n = 1;
            else
                n = 2 + 5 * (n + m) / 2;
            m = 1;
        }
        else
        {
            n = 1 + 5 * n / 2;
            m++;
        }
The steps it took me to find this can be found in this program. I started with an interpretting implementation for the Turing machine using a compressed representation of the tape. The idea that in case a number of cells are given the same 'colour' (symbol) to represent it as a single cell with a number representing the count. The implementation is found in the function interpretter. I ran this to get an idea of the patterns that occured. At one point, I replaced the count (in the printout) by a star character and discovered some repeating patterns. I decided to write a version where the turing machine is hard code. See the function hard_coded1. At one point, I made it print out the transitions from state with colour under the head to next state and next colour under the head. This showed that there were only a limited number of transitions. I used this to construct the function hard_coded2 where they are shown in the comments. I added labels for all state-colour combinations and a goto statements for the transistions. In the function hard_coded3 I reordered the statements such that state-colour combinations that often occured to follow each other appeared below each other. I also added while-loops for repeating state-colour combinations, which I rewrote in a more compact form. (The while-loops can be found as comments in the code.) This resulted in less output to be printed, but I took care that the lines in the input still matched those generated with the earlier functions, to make sure I had not introduced errors. I studied the output and noticed some repeating patterns. This lead to the implementation of hard_coded4 function that tries to reimplement the behaviour starting from the 'A'-state, '0'-colour combination. I already had noticed that there was a slight difference depending on the number of ones on the left tape were odd or even. (It took me some effort to get the number of intermediates steps calculated correctly.) However, it did not result in the expected cleaning of the tape. This lead to the development of the function hard_coded5 which does not make a difference between whether the number of ones is odd or even, but stops a little earlier. Now it did produce the expected cleaning of the tape (not exactly). Through analyzing the output produced by this function, I wrote the calc function, which is also shown above. The calculation of the n variable is derived from the code of hard_coded5 and the calculation of the m variable is based on analyzes of the output.


Sunday, January 16, 2022

Border poles

It is since last year September 2 that Conny and I went looking for border poles. This afternoon, we continued our search. We were surprised to find several border poles that we missed last year as they were clearly visible now. For sure the amount of overgrowth is much less than in the four months ago. (I had entered the location of the border poles we wanted to find in a GPX file and unloaded it to my Galaxy Tab 3 tablet.) We encountered the following border poles:


Monday, January 17, 2022

Book

Last Saturday, the lock-down with respect to non-essential shops came to an end. This morning, I could not resist my urge to visit charity shop Het Goed on my way to the office. At 10:10, I bought the following books:


This months interesting links


Home | December 2021