Abdelrazak Younes wrote: > Abdelrazak Younes wrote: >> I suspected something like that but we had 770 not 237 in addition to >> 302 so I think there was a problem nevertheless.
There is no problem in 1.5. 1.5 has 0x0237 and then 0x0302, and this is correct. If 1.6 is different then there might be a lyx2lyx problem (unless somebody changed the character on purpose, but I doubt that). > Hum, little correction, 770 == 0x302 and 567 == 0x237, so the problem > was not in the 770 character but on the character after which was not > 567 but 264. Well I am a bit confused... 264 is the next character (capital C with circumflex). Maybe you confuse the order of characters in surrogate pairs: It is "base combining", not "combining base". Georg PS: You are the obvious candidate to fix the metrics bug (4252).
