On 2011-11-22 03:30, David E. Wheeler wrote: > https://github.com/theory/text-markup/blob/master/t/base.t#L104 > > So it's expecting "fünky" under `use utf8`, which should be utf8, of course. The value returned is set up at line 84, in this block: > > https://github.com/theory/text-markup/blob/master/t/base.t#L79 > > Which should also be utf8. However, for some reason, one appears *not* to be utf8. FWIW, this test (and others like it in other of my modules) passes in 5.14 and other releases of Perl. Did something change in 5.14.4 to muck with utf8-ness? After a 'use utf8;' the source code is expected to be in utf8. So if you hexview the source, then fünky must take > 5 bytes. -- RuudThread Previous | Thread Next