Clarify necessity for ideographic spaces.

I left it unclear / open to interpretation on whether to use ideograpic
spaces, but realized that without being specific on its necessity,
developers may implement something that would cause trouble with the
Japanese user. (two words looking like one word, or phrase verification
failing because it can't handle ideographic spaces, etc.)

1 files changed, 6 insertions, 4 deletions

diff --git a/bip-0039/bip-0039-wordlists.md b/bip-0039/bip-0039-wordlists.md
index 4a68e88..1eaf585 100644
--- a/bip-0039/bip-0039-wordlists.md
+++ b/bip-0039/bip-0039-wordlists.md
@@ -10,10 +10,12 @@
 
 ###Japanese
 
-1. Users will most likely separate the words with UTF-8 ideographic space.  
-(UTF-8 bytes: 0xE38080) When generating the seed, normalization as per the spec will
-automatically change these into normal ASCII spaces. Depending on the font, displaying the
-words should use the UTF-8 ideographic space if it looks like the symbols are too close.
+1. **Developers implementing phrase generation or checksum verification must separate words using ideographic spaces / accommodate users inputting ideographic spaces.**  
+(UTF-8 bytes: **0xE38080**; C/C+/Java: **"\u3000"**; Python: **u"\u3000"**)  
+However, code that only accepts Japanese phrases but does not generate or verify them should be fine as is.
+This is because when generating the seed, normalization as per the spec will
+automatically change the ideographic spaces into normal ASCII spaces, so as long as your code never shows the user an ASCII space
+separated phrase or tries to split the phrase input by the user, dealing with ASCII or Ideographic space is the same.
 
 2. Word-wrapping doesn't work well, so making sure that words only word-wrap at one of the  
 ideographic spaces may be a necessary step. As a long word split in two could be mistaken easily