You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
It seems that there aren't too many examples of kakuji readily available on the
107
126
web. In particular, there's no complete font (say, for the jōyō kanji) that I
108
127
know of. There are about 200 or so examples that I've found images of. I'm
@@ -134,7 +153,6 @@ On radical-based decomposition of Chinese characters:
134
153
Radical Packing Language]]
135
154
136
155
* december 8
137
-
** more on kakuji and seal scripts
138
156
Continuing the thread from yesterday, I thought it might be a good exercise to
139
157
start by translating the book's description of kakuji. It's definitely above my
140
158
pay grade, but I'll do my best to translate it in-line (I tend to use the
@@ -201,8 +219,8 @@ This passage (which unfortunately seems to be all the book has on kakuji)
201
219
suggests looking also at the nine-fold seal script. Here are some resources I
202
220
have found:
203
221
- a collection of [[https://commons.wikimedia.org/wiki/Category:Calligraphy_manuals_including_nine-fold_seal_script][images]] of calligraphy manuals including nine-fold seal script
204
-
- a [[https://babelstone.co.uk/Blog/2012/10/khitan-seals.html][site]] with some images of the Khitan seals and a table of potential correspondences with
205
-
standard character forms
222
+
- a [[https://babelstone.co.uk/Blog/2012/10/khitan-seals.html][site]] with some images of the Khitan seals and a table of potential
223
+
correspondences with standard character forms
206
224
- some helpful [[https://digitalorientalist.com/2025/02/25/digital-resources-for-reading-japanese-seals/][notes]] on resources for reading Japanese seals
207
225
- slightly off-topic, but [[https://en.wikipedia.org/wiki/Kufic#Square_Kufic][square Kufic script]] is an Arabic doppelganger!
208
226
I'm sure there are a lot more resources in Chinese, but unfortunately my
@@ -211,3 +229,67 @@ Mandarin is still super basic.
211
229
I was hoping to start looking at some of the available kakuji, breaking down
212
230
radical shapes, and writing some code, but I got a little too interested in seal
213
231
scripts and the translation was tough, so I'll leave that for tomorrow.
232
+
233
+
* december 9
234
+
235
+
Before I fell asleep last night I was doodling kanji, trying to get a concrete
236
+
sense of how kakuji are designed. As the passage above points out, kakuji are
237
+
designed after seal script, so building an intuition of how the contemporary
238
+
glyphs correspond to seal script glyphs is probably a good place to start. As
239
+
I do that, though, I think it'd be fun to draw out some kakuji as well!
240
+
241
+
So... let's make a kakuji editor! Doodling is all well and good, but it'd be
242
+
nice to have a little purpose-built glyph editor. I've recently been trying to
243
+
learn a bit of C using =raylib=, and this is as good of a learning project as any.
244
+
245
+
The basic idea is a simple grid-based editor where you can paint in valid
246
+
kakuji. I haven't yet worked out exactly what I mean by valid (we'll narrow the
247
+
definition down as we go), but for now notice that kakuji must be space-filling
248
+
and be composed entirely of vertical and horizontal strokes.
249
+
#+ATTR_HTML: :alt an example of kakuji using diagonal lines
Edit, added later: here's a smallified version of the image I edited.
62
59
63
-
{{< figure src="/ox-hugo/japan-2025-self-portrait-dithered.jpg" alt="a mirror self-portrait taken by handheld camera in a hotel room" >}}
60
+
{{< figure src="images/december-adventure-2025/japan-2025-self-portrait-dithered.jpg" alt="a mirror self-portrait taken by handheld camera in a hotel room" >}}
The original file was 6280x4160, at around 2.6mb, but I haven't bothered
72
-
checking into the repository. The image above has been scaled down
69
+
checking it into the repository. The image above has been scaled down
73
70
significantly, and with 16 colors comes in at around 82kb.
74
71
72
+
The tricky thing with `ox-hugo` (I write my website in org-mode in emacs) is that
73
+
I sometimes produce images in the org-file directory instead of the hugo assets
74
+
directory (say when writing python code to generate images). These generated
75
+
images need to be copied over to the hugo directory. In cases like static
76
+
pre-selected images, though, I can just throw them into the hugo directory to
77
+
begin with.
78
+
75
79
76
80
## december 5 {#december-5}
77
81
@@ -134,6 +138,10 @@ about their extreme space-filling blockiness when compared to the gentle
134
138
hieroglyphic curves of seal script or the varied visual rhythm of contemporary
135
139
regular script.
136
140
141
+
See, for example, this image I found on [Robundo Publishing's blog](https://robundo.com/robundo/column/archives/268):
142
+
143
+
{{< figure src="/images/december-adventure-2025/robundo_kakuji_22_d.jpg" alt="a manual showing correspondences between kakuji and other scripts" >}}
144
+
137
145
It seems that there aren't too many examples of kakuji readily available on the
138
146
web. In particular, there's no complete font (say, for the jōyō kanji) that I
139
147
know of. There are about 200 or so examples that I've found images of. I'm
@@ -169,9 +177,6 @@ On radical-based decomposition of Chinese characters:
169
177
170
178
## december 8 {#december-8}
171
179
172
-
173
-
### more on kakuji and seal scripts {#more-on-kakuji-and-seal-scripts}
174
-
175
180
Continuing the thread from yesterday, I thought it might be a good exercise to
176
181
start by translating the book's description of kakuji. It's definitely above my
177
182
pay grade, but I'll do my best to translate it in-line (I tend to use the
@@ -241,8 +246,8 @@ suggests looking also at the nine-fold seal script. Here are some resources I
241
246
have found:
242
247
243
248
- a collection of [images](https://commons.wikimedia.org/wiki/Category:Calligraphy_manuals_including_nine-fold_seal_script) of calligraphy manuals including nine-fold seal script
244
-
- a [site](https://babelstone.co.uk/Blog/2012/10/khitan-seals.html) with some images of the Khitan seals and a table of potential correspondences with
245
-
standard character forms
249
+
- a [site](https://babelstone.co.uk/Blog/2012/10/khitan-seals.html) with some images of the Khitan seals and a table of potential
250
+
correspondences with standard character forms
246
251
- some helpful [notes](https://digitalorientalist.com/2025/02/25/digital-resources-for-reading-japanese-seals/) on resources for reading Japanese seals
247
252
- slightly off-topic, but [square Kufic script](https://en.wikipedia.org/wiki/Kufic#Square_Kufic) is an Arabic doppelganger!
248
253
@@ -252,3 +257,69 @@ Mandarin is still super basic.
252
257
I was hoping to start looking at some of the available kakuji, breaking down
253
258
radical shapes, and writing some code, but I got a little too interested in seal
254
259
scripts and the translation was tough, so I'll leave that for tomorrow.
260
+
261
+
262
+
## december 9 {#december-9}
263
+
264
+
Before I fell asleep last night I was doodling kanji, trying to get a concrete
265
+
sense of how kakuji are designed. As the passage above points out, kakuji are
266
+
designed after seal script, so building an intuition of how the contemporary
267
+
glyphs correspond to seal script glyphs is probably a good place to start. As
268
+
I do that, though, I think it'd be fun to draw out some kakuji as well!
269
+
270
+
So... let's make a kakuji editor! Doodling is all well and good, but it'd be
271
+
nice to have a little purpose-built glyph editor. I've recently been trying to
272
+
learn a bit of C using `raylib`, and this is as good of a learning project as any.
273
+
274
+
The basic idea is a simple grid-based editor where you can paint in valid
275
+
kakuji. I haven't yet worked out exactly what I mean by valid (we'll narrow the
276
+
definition down as we go), but for now notice that kakuji must be space-filling
277
+
and be composed entirely of vertical and horizontal strokes.
278
+
279
+
280
+
By space-filling, I
281
+
mean that every grid element must have at least some ink in it. The second rule
282
+
is actually my own -- there are examples on Black Silk's crest list that have
283
+
diagonal strokes (see the image to the left). I'm happy to ignore such cases for
284
+
now. This particular example, incidentally, is no longer used: 總 is listed as
285
+
the pre-simplified form of 総 in Wikipedia's [list of jōyō kanji](https://en.wikipedia.org/wiki/List_of_j%C5%8Dy%C5%8D_kanji), and means whole
286
+
or total.
287
+
288
+
After tinkering for a bit, I got some of the `raylib` boilerplate out of the way
289
+
and ended up with a very preliminary little clickable grid-based editor:
290
+
291
+
{{< figure src="/images/december-adventure-2025/raylib_grid_r.jpg" alt="a dinky editor for filling in black squares in a grid with red ink" >}}
292
+
293
+
To get a bit of a drawing feel, it'd obviously be nice to click and drag,
294
+
brush-like, to draw contiguous blocks of ink. Frustratingly, I seem to be
295
+
running into an issue with how `raylib` interprets my mouse's position when I
296
+
drag. Maybe it's due to my tiling window manager, or maybe it's due to some kind
297
+
of X/Wayland incompatibility, but as soon as I drag my mouse, `raylib` reports
298
+
that my mouse is a solid 50 pixels above where it actually is, until I release
299
+
the left button and move my mouse a bit. It doesn't seem to be an error in my
300
+
code -- I was able to reproduce this in the little program I was writing a month
301
+
ago to learn the basics of `raylib`.
302
+
303
+
Alas, I've already wasted most of my night trying to debug this, so I'll leave
304
+
it here for now. I suppose I can return to this and swap to keyboard controls,
305
+
though it definitely won't feel as smooth. Then finally onto the interesting
306
+
part: understanding the patterns behind the gaps in ink and the radical shapes!
307
+
308
+
309
+
## future adventures? {#future-adventures}
310
+
311
+
Thought I'd collect the little project ideas that tend to pop into my head when
312
+
working on code.
313
+
314
+
- write up some notes on the basics of how C programs are compiled and linked,
315
+
as well as the platform-dependent aspects. _inspired by:_ me realizing I have no
316
+
idea what I'm doing when I'm putting together a Makefile or working on a
317
+
C project on my mac when traveling
318
+
- learn and write a toy program in assembly! and/or forth! and/or uxntal! if I
319
+
can get the kakuji editor working, it might be a good second program to port
320
+
over. _inspired by:_[100r](https://100r.co), learning more about how memory and caches works
321
+
- exploring moving away from org/ox-hugo for static site generation. _inspired
322
+
by:_ as much as I love being able to use org-babel to write and run code from
323
+
directly inside my posts, losing control over the final export is annoying.
324
+
maybe I could just post-process the generated html? or rather, use tools that
0 commit comments