Am Samstag 15 Januar 2011, 15:08:28 schrieb Thiago Macieira: > On Saturday, 15 de January de 2011 12:28:38 Ingo Klöcker wrote: > > On Friday 14 January 2011, Thiago Macieira wrote: > > > On Friday, 14 de January de 2011 21:49:01 Ingo Klöcker wrote: > > > > On Friday 14 January 2011, Thiago Macieira wrote: > > > > > On Friday, 14 de January de 2011 13:46:19 Sebastian Trueg wrote: > > > > > > However, the query parser still uses static QRegExp objects > > > > > > which seems a bad idea, isn't that right? > > > > > > > > > > static QRegExp are a bad idea regardless of whether threading is > > > > > involved or not. > > > > > > > > Why? > > > > > > Because it's static. Static non-PODs aren't a good idea. > > > > Where static means file static, right? > > File-static, class-static or function-static, though function-static is the > least evil of those.
Again another crash -- with Qt 4.6.2 -- that should not happen if I understood everything correctly: term.contains(QRegExp("^[a-zA-Z0-9\\-\\.]+\\.[a-zA-Z]{2,6}")) http://websvn.kde.org/tags/KDE/4.4.2/kdebase/workspace/plasma/generic/runners/locations/locationrunner.cpp?view=markup line 99 https://bugs.kde.org/show_bug.cgi?id=247117 The code has changed since then (for KDE 4.6), so asking users to retry with a newer KDE version would not get valuable information. Still I tried to reproduce the crash with the old code and with Qt 4.7.1 but was not able, so is that fixed there?