On segunda-feira, 4 de fevereiro de 2013 15.30.24, Knoll Lars wrote: > > Importing ICU into qtbase is fine with me. Anyhow, I don't particular like > > hard dependency to ICU from Qt5Core, since it forces everyone deploying a > > Windows helloworld to also ship icudt49.dll with 17,5 MB (!). > No, please not. I don't want to add such a huge module to 3rdparty. At most, > I could live with a git submodule.
Also note that the data library is only an inconvenience now because we have duplicated data: the Unicode tables and CLDR locale data are in ICU and in QtCore. Once we remove our own copy of that data, the total size would be reduced. I also think that we should configure ICU on Windows to reduce its data size: we don't need all of it. > > I know that you can still configure with -no-icu. So how about making the > > usage of ICU a runtime decision? Has anybody looked into this already? > > And are we up to maintain the old, internal backend also in future > > versions?> It used to be dynamically loaded in Qt 4. We changed in Qt 5 because ICU doesn't know how to make releases that are binary compatible. It's extremely hard to find the library and the functions inside. Besides, there is no failure mode. If the library isn't found, QtCore must crash / abort / qFatal. Since that's used for everything (QCoreApplication triggers it), it's pointless to dynamically load it. It just makes the loading process slower. -- Thiago Macieira - thiago.macieira (AT) intel.com Software Architect - Intel Open Source Technology Center
signature.asc
Description: This is a digitally signed message part.
_______________________________________________ Development mailing list Development@qt-project.org http://lists.qt-project.org/mailman/listinfo/development
