diff --git a/src/bin/unicodeblocks.c b/src/bin/unicodeblocks.c
index af5c2966f..2bfd41f34 100644
--- a/src/bin/unicodeblocks.c
+++ b/src/bin/unicodeblocks.c
@@ -7,7 +7,6 @@
 #include "demo.h"
 
 // show unicode blocks. a block is always a multiple of 16 codepoints.
-#define ITERATIONS 24 // show this many pages
 #define BLOCKSIZE 512 // show this many per page
 #define CHUNKSIZE 32  // show this many per line
 
@@ -15,11 +14,43 @@ int unicodeblocks_demo(struct notcurses* nc){
   struct ncplane* n = notcurses_stdplane(nc);
   int maxx, maxy;
   notcurses_term_dimyx(nc, &maxy, &maxx);
-  int i;
-  // show 256 at a time. start with the known-working ascii/latin-1 blocks.
-  uint32_t blockstart = 0;
-  for(i = 0 ; i < ITERATIONS ; ++i){
-    ncplane_erase(n);
+  // some blocks are good for the printing, some less so. some are only
+  // marginally covered by mainstream fonts, some not at all. we explicitly
+  // list the ones we want.
+  const struct {
+    const char* name;
+    uint32_t start;
+  } blocks[] = {
+    { .name = "Basic Latin, Latin 1 Supplement, Latin Extended", .start = 0, },
+    { .name = "IPA Extensions, Spacing Modifiers, Greek and Coptic", .start = 0x200, },
+    { .name = "Cyrillic, Cyrillic Supplement, Armenian, Hebrew", .start = 0x400, },
+    { .name = "Arabic, Syriac, Arabic Supplement", .start = 0x600, },
+    { .name = "Samaritan, Mandaic, Devanagari, Bengali", .start = 0x800, },
+    { .name = "Gurmukhi, Gujarati, Oriya, Tamil", .start = 0xa00, },
+    { .name = "Telugu, Kannada, Malayalam, Sinhala", .start = 0xc00, },
+    { .name = "Thai, Lao, Tibetan", .start = 0xe00, },
+    { .name = "Myanmar, Georgian, Hangul Jamo", .start = 0x1000, },
+    { .name = "Ethiopic, Ethiopic Supplement, Cherokee", .start = 0x1200, },
+    { .name = "Canadian", .start = 0x1400, },
+    { .name = "Runic, Tagalog, Hanunoo, Buhid, Tagbanwa, Khmer", .start = 0x1600, },
+    { .name = "Mongolian, Canadian Extended, Limbu, Tai Le", .start = 0x1800, },
+    { .name = "Buginese, Tai Tham, Balinese, Sundanese, Batak", .start = 0x1a00, },
+    { .name = "Lepcha, Ol Chiki, Vedic Extensions, Phonetic Extensions", .start = 0x1c00, },
+    { .name = "Latin Extended Additional, Greek Extended", .start = 0x1e00, },
+    { .name = "General Punctuation, Letterlike Symbols, Arrows", .start = 0x2000, },
+    { .name = "Mathematical Operators, Miscellaneous Technical", .start = 0x2200, },
+    { .name = "Control Pictures, Box Drawing, Block Elements", .start = 0x2400, },
+    { .name = "Miscellaneous Symbols, Dingbats", .start = 0x2600, },
+    { .name = "Braille Patterns, Supplemental Arrows", .start = 0x2800, },
+    { .name = "Supplemental Mathematical Operators", .start = 0x2a00, },
+    { .name = "Glagolitic, Georgian Supplement, Tifinagh", .start = 0x2c00, },
+    { .name = "Supplemental Punctuation, CJK Radicals", .start = 0x2e00, },
+  };
+  size_t sindex;
+  ncplane_erase(n);
+  for(sindex = 0 ; sindex < sizeof(blocks) / sizeof(*blocks) ; ++sindex){
+    uint32_t blockstart = blocks[sindex].start;
+    const char* description = blocks[sindex].name;
     int chunk;
     if(ncplane_cursor_move_yx(n, 2, 2)){
       return -1;
@@ -29,6 +60,12 @@ int unicodeblocks_demo(struct notcurses* nc){
     if(ncplane_printf(n, "Unicode points %04x–%04x\n", blockstart, blockstart + BLOCKSIZE) <= 0){
       return -1;
     }
+    if(ncplane_cursor_move_yx(n, 3, 3)){
+      return -1;
+    }
+    if(ncplane_printf(n, description) <= 0){
+      return -1;
+    }
     for(chunk = 0 ; chunk < BLOCKSIZE / CHUNKSIZE ; ++chunk){
       if(ncplane_cursor_move_yx(n, 4 + chunk, 2)){
         return -1;