Previous Up No next

Dairy, April 2018



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


Monday, April 2, 2018

Ready Player One

This afternoon, I went to see the movie Ready Player One. The movie was better than I had expected. Yes, it was full of action, but it was not all over the top. The story also felt believable, probably because it is based on a book, although I understand that the plot differs from the book. Also the many references (easter eggs) to movies and games made it interesting.



© Heidi Ulrich

Tuesday, April 3, 2018

TkkrLab: Gigatron

I took the Gigatron micro-computer with me to show it at the office and to have it photographed at TkkrLab. One of the photos is shown here on the right. I spend some time studying the Python script in the Core folder of Gigatron ROM project at GitHub. I could not make much sense of it.


Thursday, April 5, 2018

5 solutions for 9 with 19 integers

In the past month, I worked on a new program for finding solutions for the the Irregular Chocolate Bar problem. In the past two weeks, I did not do much with it, but this afternoon, I decided to give it a try with a simple change. After running for some hours, it found the following five solutions for 9 with 19 integers:

1 51 85 89 95 111 115 120 124 130 150 155 160 165 169 185 191 195 229
6 56 84 85 95 115 116 119 120 130 150 155 160 164 165 185 195 196 224
41 49 76 85 95 99 115 120 121 130 150 155 159 160 165 181 185 195 239
21 56 69 85 95 101 115 119 120 130 150 155 160 161 165 179 185 195 259
26 61 62 78 97 103 106 111 113 134 138 142 146 169 177 183 202 218 254

I took the idea of the new program by looking at a solution for 7 and putting the numbers into a matrix of 6 columns and 7 rows, such that the rows summed up to 60 and the columns up to 70. Each cell is either empty or contains exactly one number. Now it turns out that there are many solutions that have the same pattern with filled and empty cells (when normalized). The columns and rows can be sorted such that the number of filled cell in each row of column increases or stays the same. The program consists of three nested 'loops'. In the outer loop the program tries all possibilties of numbers of filled columns and columns. In the middle loop, given the number of cells that need to be filled in each row and column, if finds every possibility to empty and filled cells. Before continueing, the program checks if the solution is unique by permuting all columns and rows (with same number of filled cells) to see if it already appeared. If this is the case, the third and last loop tries to find numbers matching the pattern of filled cells, such that the columns and rows add up to the correct numbers. Before this, it does figure out a strategy, to fill the cells that need to be filled. The strategy is nothing else than an order in which they are filled. After they are filled, the solution is checked with a new (and much faster) knapsack algorithm.


Friday, April 6, 2018

Book

At 14:42:10, I bought the book Paralipomena Orphica written by Harry Kurt Victor Mulisch in Dutch, with an interaction by Marita Theodora Catharina Mathijsen, edited by Margot Engelen, and published by B for Books in 2016, ISBN:9789085164357, from charity shop Het Goed for € 0.95.


Saturday, April 7, 2018

Going into the city

I went into the city with Andy. After visiting McDonalds, we went to a toy shop, because Andy wanted to look at DVD's. He did not find many at one place and we asked an attendant if there were more. He said that what we had seen was all and that this was due to Netflix. We walked to Concordia to see the exhibition Robson Ateliers. I asked Andy what he saw in the paintings and objects. He gave interesting answers, but wanted to leave quickly. Next we went to bookshop Broekhuis. At 15:28:28, I bought the following two photo-books from sale of art books for € 5.00 each: Next he wanted to go to the Public Library for looking at all the DVD's. At around 15:48 at the Public Library (Centrale Bibliotheek Enschede), I bought the book Vrouwen en Liefde: Een culturele revolutie in ontwikkeling written by Shere Hite, translated from English to Dutch by Wineke Boegborn and Anki Klootwijk from Women in Love: A Cultural Revolution in Progress, published by Sesam in 1990, ISBN:9789024647477, for € 1.00.

Sunday, April 8, 2018

Magnolia

The first flowers of our magnolia have opened today. Yesterday, they almost opened. This weekend it is the first time this year we have temperatures above 20°C. In March we had some very cold periods and I hope they will not happen again.

Gigatron: reading data from ROM

The Gigatron TTL micro-computer has a Harvard architecture, which means that instructions and data are not accessed in the same ways as in the now commonly used Von Neumann architecture. The Gigatron has no instruction for reading data from the ROM, which only contains instructions (and their operands). This evening, I discovered that there is a very clever trick by which it is possible to read data from ROM by clever use of the two step instruction pipe-line and the fact that the pipe-line is not cleared during a jump/branch. This means that the Gigatron will always execute the instruction following the jump expression. The trick is to use a jump instruction to a location in the ROM where there is a load immeditate instruction (loading the data in the accumulator) and let this follow by another jump instruction to (for example) the next memory location. This causes the load immediate instruction to be executed when the program counter has returned to the next memory location. See the following table for a step-by-step flow of events, where the 'normal' flow of execution starts at 0x0804 and the data is stored at 0x0870.

 Program counter  Instruction being fetched  Instruction being execute 
0x0804jump to 0x0870previous instruction
0x0805jump to 0x0806jump to 0x0870
0x0870load immediatejump to 0x0806
0x0806some instructionload immediate

Of course, this idea is only usefull when the address of the location of the load immediate instruction (0x0870) can be calculated. There are indeed jump (branch) instructions where this is possible. Note that the load immediate instruction could also be replaced by some other instruction (as long as it is not a jump/branch instruction).


Tuesday, April 10, 2018

Cosey Fanny Tutti

This morning, I finished reading the auto-biography Cosey Fanny Tutti (by herself), which I started reading on March 19, the day after I bought the book. It was an interesting read. A red thread in the book is her complicated relationship with Genesis P-Orridge, who is called Gen in the book.

No solutions for 10 with 21 integers

Some days ago, I started the new program for the the Irregular Chocolate Bar problem for finding solutions for 10. Today, I saw that it found solutions with 22 integers. Not surprising, because last September, I already found 885 such solutions. It seems that there are no solutions with 21 integers, as the program started searching from 20 integers. Below one of the solutions (with 22 integers) that the program found, is given. This solution is not included in the 885 solutions that were found last year. I am not really surprised about this and expect that the progam will find many more such solutions.
10 set:  10 28 33 38 45 52 53 73 80 81 102 108 115 137 143 150 172 199 200 207 242 252
6 buckets with 420:  45+143+199+33 172+108+102+38 28+242+150 200+73+137+10 252+115+53 80+52+207+81
7 buckets with 360:  172+53+102+33 207+115+38 200+10+150 45+73+242 28+52+81+199 80+137+143 252+108
8 buckets with 315:  45+199+38+33 28+137+150 80+52+81+102 252+53+10 73+242 200+115 172+143 108+207
9 buckets with 280:  52+45+150+33 242+38 115+53+10+102 81+199 137+143 207+73 172+108 80+200 252+28
10 buckets with 252:  108+73+38+33 102+150 242+10 199+53 137+115 28+143+81 207+45 200+52 80+172 252


Friday, April 13, 2018

Easter Egg in the Gigatron

While showing the Gigatron TTL microcomputer to some family members, I hit upon an animation that I had not seen before and that to my knowledge has not been published. I conclude that this must be Easter Egg build into the ROM by Marcel and Walter. I have not yet discovered the exact steps to activate the Easter Egg, but I was able to activate it a second time. I found nothing in the gigatron-rom at GitHub related to the Easter Egg, which seems to imply that the ROM is either different version than the one published there (or that the target files published there are different that the ones produced by the sources). Have fun with you Easter Egg hunt!

Addition Saturday, April 14: I got an email from Marcel informing me that I was the first one to have reported the Easter Egg and that even Walter did not know anything about it.


Saturday, April 14, 2018

China Dreams

I saw the exhibition China Dreams by Willemijn Calis at XPO. The works on display were made during her trip to Dalin, China. I found one of the works quite interesting and contacted her.


Tuesday, April 17, 2018

Books

At 16:32:40, I bought the following two books from charity shop Het Goed:


Thursday, April 19, 2018

AKI takes over

In the afternoon, I went to see the exhibition AKI takes over: "Hotel Bella Arte" at the University of Twente with works by Nils Leibeling and Jelle van Assem.

Almost 30°

At two meter, the maximum temperature at weather station at Twente Airport was 28.7° Celcius and at 10 cm it was 29.9°. It would not surprise me if these are record high temperatures for this day of the year. Most flowers of our magnolia have fallen off. A tulip has opened in the back garden.


Friday, April 20, 2018

Coloured rubber straps

About two weeks, I bought a ball of coloured rubber straps, with the intent to use as hair rubber straps. I got the idea to use two straps with different colours every day from a set of five colours. I also do not want to wear the same colour on two consecutive days. I reasoned that there must be a sequence of ten combinations that I could follow (because there are ten combinations of five colours). When I wrote a program to count all the solutions, I discovered that there was none. Next, I searched for a solution of twenty combinations where each combination is included twice, and I found many. In search of an elegant solution, I wanted to avoid solutions where some colour is not used for some time or used on a long sequence of alternating days. Then I realized that I also did not want to wear a certain combination on days that are close together. However, in each solution of twenty combinations, there was a combination that appeared only three days apart. While thinking about this, I realized that the combinations are like the vertices in the Perterson graph. This also explains why there was no solution for ten combinations, because the Peterson graph has no Hamiltonian cycle. I wondered whether anybody ever studied the existence of paths that visit each vertex exactly n-times, a generalisation of Hamiltonian cycles (for which n equals 1). Maybe one reasons is that many graphs that do not have a Hamiltonian cycle do have a cycle which visits each vertex twice, and/or that there exists a rather simple constraint for which a graph contains a cycle that exactly visits each vertex n times. But adding the requirement that each there must be at least k vertices between two consecutive visits, may make it interesting. One could even state that fraction k+1 divided by the number of vertices defines the Hamiltonianess.

When I tried to find an elegant solution on paper, I realized that there was an error in the part of the algorithm that calculated the distance between reoccurences of the same combination of coloured straps. When I corrected the algorithm, it counted 60 solutions with at least seven other combinations between each combination. The solution I had found on paper, was on of the 60 solutions found by the program. I strongly suspect that all solutions are basically the same, with the colours and the starting position permutated, because there are 10 combinations and 3! ways in arranging the neighbours of a certain combination. I wrote a small fragment of JavaScript to calculate, which combination to wear on the coming ten days (from today). (I am not going to publish the program, because it seems I could have figured out everything on paper.)


Sunday, April 22, 2018

Peterson graph

I thought about making a pendant of the Perterson graph. I had something like the Petersen Graph Pendant in mind, and I thought about making it from a copper wire. Then I wondered if it would be possible to make it out of a single copper wire, where each edge is visited twice in opposite direction. At home I adapted the program I wrote two days ago, because trying this by hand seemed too complicated, and it found 960 'different' solutions of 30 combinations where the minimal distance between two occurences is five. No solutions with a greater minimal distance were found by the program. Five is two less than seven as found with the solution with 20 combinations. I also discovered that 720 of these solutions have five combinations occuring just five positions from each other, and 240 solutions with only three combinations occuring just five positions from each other. There was no further subdivision with respect to the occuring distances between combinations. It would not surprise me if there are basically only a few (two or four) solutions. I have not yet decided whether I am going to switch to the sequence of 30 combinations. Luckily, I can delay the decision till the thirtieth of the month because I found a solution that is similar until that day. The combinations for that sequence are displayed below:


This months interesting links


Home | March 2018