[CWB] unicode problems with Greek and OCS

[Stefan Evert]

> On 22 Mar 2015, at 15:11, Hardie, Andrew <a.hardie at lancaster.ac.uk> wrote:
> 
> Total mystery to me. Possibly the result of cosmic rays.
> 
> One possibility that occurs to me now is that, if your corpus contained any characters that have been relatively recently added to Unicode (which is possible for historical docs in Cyrillic?) that the character tables in GLib had not been updated in the old build, but they have in the new (????)

Yes, that's what I just thought:  Recompiling CWB may have used a more recent version of gcc or a newer GLib build.  That doesn't seem to be uncommon, though it usually goes the other way (i.e. bugs suddenly pop up after upgrading dependencies).

Good to hear that this mysterious problem seems to have been resolved!

Cheers,
Stefan