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
|
Blackletter
I started reading the book Designing Type by Karen Cheng. On page 14 it has a section on 'Type Classification' that
talks about the French typographic historian Maximilian Vox. I noticed that his classification, now known as Vox-ATypI classification did not include Blackletter, also known as Gothic script, Gothic minuscule or Gothic type,
which was a script used in Germany and also the Netherlands for official
publications, where it was known as Textualis. Blackletter and also Gaelic type were later added to the classification. I cannot remember
having seen Gaelic type before, which uses the insular G letter form for the letter g. (Another example of how reality is almost always more complex than it
appears on first sight.)
Proclamate Light
I figured out how to use your own fonts on a webpage. I downloaded the
Proclamate Light
font designed by Paul Lloyd. Below an example of the usage with the text of
what I wrote yesterday. Just before the paragraph, I added the following in
the HTML:
<style>
@font-face { font-family: 'proclamate_lightlight';
src: url('proclamate_light-webfont.woff') format('woff');
font-weight: normal;
font-style: normal; }
.blackletter { font-family: 'proclamate_lightlight'; }
</style>
and I used '<p class"blackletter"><font size="+1">' for the paragraph. I am not
sure if this works for all browsers, so if it shows as normal text, it might
by my fault.
I started reading the book Desiging Type by Karen Cheng. On page 14 it has a section on 'Type Classification' that
talks about the French typographic historian Maximilian Vox. I noticed that his classification, now known as Vox-ATypI classification did not include Blackletter, also known as Gothic script, Gothic minuscule or Gothic type,
which was a script used in Germany and also the Netherlands for official
publications, where it was known as Textualis. Blackletter and also Gaelic type were later added to the classification. I cannot remember
having seen Gaelic type before, which uses the insular G letter form for the letter g. (Another example of how reality is almost always more complex than it
appears on first sight.)
I managed to compile a program with the C
compiler that I am developing. The commit 6ac84a5d contains the version of tcc_cc.c with
which the hello.c program can be compiled to the stack_c language, and
compiled to an ELF with the stack_c compiler and the life-bootstrap programs
blood-elf, M1 and hex2. Below the hello.c
program.
#include <stdio.h>
int main(int argc, char *argv[])
{
fputc('H', stdout);
fputc('\n', stdout);
return 0;
}
The include directive is ignored by the compiler and the stdlib.c
file is included. The relevant parts of this file are:
const FILE *stdout = 1;
int sys_int80(int a, int b, int c, int d);
int fputc(int c, FILE *stream)
{
int buffer[1];
buffer[0] = c;
return sys_int80(4, stream, buffer, 1);
}
The function sys_int80 represent the software interrupt that are used
in Linux for system
calls. I am still struggling a bit how array variables are dealt with.
In the above fputc function, in the call to sys_int80 the
arguments stream and buffer are slightly handled differently.
For the variable stream the contents is retrieved, while for the
variable buffer, which is an array, the address is used. The compiler
compiles this to the following stack_c program:
int stdout
void fputc
{
int stream stream =:
int c c =:
int buffer
buffer 0 4 * + c ? = ;
4 stream ? buffer 1 sys_int80 () return
}
void __init_globals__ ;
void main
{
int argv argv =:
int argc argc =:
__init_globals__ ()
72 stdout ? fputc () ;
10 stdout ? fputc () ;
0 return
}
void __init_globals__
{
stdout 1 = ;
return
}
AI agents for software engineering
Today, I came across the article My AI Skeptic Friends Are All Nuts that argues that there are no reasons
not to use AI agents as a software engineer. In the past
half year, I have tried to use agents in my work, but they never have helped me
much. When, this morning, I asked ChatGTP to generate the code for a M4F MCU that
implements the VirtIO code for accessing the GPIO through RPMSG using
FreeRTOS, it produced two
answers that were both incomplete and incorrect. (I have added links to the
acronyms to show that these are not very obsure things, suggesting that what I
ask is not something extremely obscure.) Someone replied, acknowledging that it was outside his domain, that I could break
down my prompt into separate prompts. I followed this approach and again got
some answer that was neither complete nor correct. When I made the suggestion
to ChatGTP to take Section 5.18 of Virtual I/O Device (VIRTIO) Version 1.3 into account, it produced almost
the same code preceded with some babbling from the document, but did not use
anything from the specification, not even the code fragments mentioned in the
section. I understand that a large group of software developers are using
AI agents, probably even more than half, but I get the impression that these
are software developers in domains where there is a lot of repetition. For
areas that are less common, such as embedded software development for specific
devices, AI agents probably do not have enough training materials, also because
a lot of embedded software is proprietary. Examples that are provided by
companies selling embedded devices are often rather simple. For certain
application domains, such as aviation and medical, there are often strict
coding standards that have to be followed.
Book
At 16:47, I bought the book Standbeeld: Kunst en woningbouw edited by
Edwin Boering and
Jaap Huisman, written in Dutch, and
published by Gemeente Amsterdam on Thursday, June 6, 1991
from Het Goed for € 0.99.
KunstenLandschap 2025
I biked around the route of KunstenLandschap 2025. At 15:11, I bought the entrence and got a
wristband, a small book with details about the art work and artists, and a map with the route. There was some rain while biking around. I stayed
inside some shipping container along the route during some of the worst rain. I
found the following artist noteworthy (including the number of the location
where they were shown):
At Rijksmuseum Twenthe, I watched two new
exhibitions. From herman de vries. 70 jaar de
natuur als kunstwerk ('herman de vries: 70 year the nature as artwork'), I found the following
works noteworthy:
- Untitled [first random composition], 1962.
- The ITBON relief (v67-36a), 1967.
- Table XXXIII from , Statistical Tables for Biological, Agricultural, and Medical Research
by Sir Ronald Aylmer
Fisher and Frank
Yates.
- spatial random structure, 1965.
- v70-07 random objectivation, 1970.
- the saviours, 2012.
- holy days [with marion reisner], 2008.
- untitled [collected playa de la caleta, la gomera], 1982.
- mesa (43 objects), 1996-2007.
- Book natural relations: eine skizze, 1968.
- Field work, De Hoge Veluwe, 2025.
- The coven - tree love, melanie bonajo, 2019.
- The coven - trustingingeling, melanie bonajo, 2019.
- The coven - cuddle coven, melanie bonajo, 2019.
- all, 1994.
From the exhibition Ergens tussen hoop en vrees ('Somewhere between hope
and fear'), I found the following works noteworthy:
- The Embium Building with Sky, Rik Smits, 2011.
- Harbour, Laith Al Rahmani, 2024.
Long low slow baden
I went biking again today and followed some part of route of the KunstenLandschap 2025. The highlight of the day was being a subject of the
Long low slow
baden installation by Gemma Luz Bosch. It was rather deeply relaxing and special experience. I
saw the installation last year on June 28 when
it was shown on the Graduation show KABK exhibition.
This afternoon. I first went to Concordia. I first had a look at the shower
tray of the Paviljoen #4 exhibition that
I saw on May 24. The bucket with blue coloured
water had emptied completely and was removed. Next, I saw Doublet #8, a
colaboration between Peggy Franck, who
painted the windows and Christine
Moldrickx, who I guess contributed the two dia projectors. I found the
exhibition somewhat minimalistic. On the first floor, I saw the exhibition
Kunstblikken: De Spiegel Binnenin with works by high school students
that they made in collaboration with some artists. I found one lino print
rather interesting. Next, I (re)visited the following locations:
- 3: Kunst in Stilte en Licht, which very different than last
Saturday, because there were lots of people all walking along the table.
The bird cage with bird that was hanging above the table, was also gone.
- 4.
- 9: Long low slow baden. I waited about half an hour before it was
my turn. It was worth the waiting.
- 12: Bloemen Geuren niet voor u, where I smelled the following
'chemicals': Citronellol, linalool, geraniol, α-Isomethyl ionone, farnesol, eugenol,
and benzyl
salicilate.
At Rijksmuseum Twenthe, I walked around and
found the following works noteworthy:
- v67-36b random objectivation, herman de vries, 1967.
- White, herman de vries, 1962.
- belladonna (digitized 16mm film), herman de vries, 1983.
- Inflammatory Essays, Jenny Holzer, 1979-1982.
- The Undercurrent, Rory Pilgrim, 2019.
First tropic day
Today, the temperature at Twenthe Airport has gone
up to 30.4° Celsius, which does not beat the record of 32.8° of 1964,
but it does make the first tropic day according to the definition of the
Royal Netherlands Meteorological Institute. The predition is that tomorrow it
will even become hotter and probably break the record for that day.
Home
| May 2025
