Dear Bogdan, when you succeed in building Qt-WinGW-ANGLE, please post your steps, or better, an installable package (or ZIP with relevant parts) if that is possible :-) Would be fine for 5.2.
For general development, I am afraid having to build Qt from sources with special adaptations is quite a high burden. I am not familiar with the Olson/ICU timezone database. Are the benefits worth the effort? Or maybe one of us (you?) will have to provide pre-built installable "Qt+ICU for Stellarium" dev packages. Best regards, Georg On Mo, 11.11.2013, 09:09, Bogdan Marinov wrote: > On Wed, Oct 16, 2013 at 2:19 PM, Bogdan Marinov > <daggers...@gmail.com>wrote: > >> Just a FYI: >> >> John Layt's work on a time zone engine for Qt has been merged in its >> codebase and it will be available in Qt 5.2. >> >> >> https://bugreports.qt-project.org/browse/QTBUG-71?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel >> >> https://bugreports.qt-project.org/browse/QTBUG-23509?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel >> >> I'll have a look at it and find out to what degree it will be suitable >> for >> use in Stellarium. Of course, to actually work on time zones in >> Stellarium >> I'll need a computer that supports OpenGL 2.0. Which is somewhat >> possible >> in November. >> > > A small update on this: I've been trying to build Qt with MinGW and ANGLE > support, so that I can run it on Windows with DirectX. > > I took the chance to look at the source code of the Time Zone class. An > unpleasant surprise: while the code itself uses the Olson database clock > names (the familiar Linux tz database), it uses different implementations > on different platforms. The proper Olson/IANA database is used only on > Linux, everything else uses whatever timezones are used on that OS, with > their identifiers mapped to Olson's. > > http://doc-snapshot.qt-project.org/qt5-stable/qtimezone.html > > This means I'll probably have to write my own time zone classes after all. > Fortunately, the existing implementation means that I can "borrow" the > code > of Qt-based tz database reader (with the appropriate attribution, of > course > :)). > > A possible solution is to roll it out in two step: the first step is to > use > Qt time zones, introducing the new GUI to the users; the second step just > will be to re-factor the necessary parts of the back-end (all the boring > data entry would be already done for the first step). This will have the > advantage of quickly shipping a time zone feature without waiting for me > to > finish the more tricky advanced code. > > I am also investigating the option to force Qt to use ICU instead of a > platform-specific implementation. ICU's time zone library is based on the > Olson database and can be updated: > http://userguide.icu-project.org/datetime/timezone#TOC-Updating-the-Time-Zone-Data > Unfortunately, this means using a custom build of Qt on platforms that > don't support the proper database. > > Oh, and the bad news: after more than 9 hours of building, my Qt > installation failed to build successfully. Now I have to fix the problem, > clean, reconfigure and rebuild. > > P.S. The Qt documentation calls it the "Olsen" database, but it's Olson. > I'm going to file a bug report. :D Qt 5.2 is scheduled to ship in early > December. :DD > > Regards, > Bogdan Marinov > ------------------------------------------------------------------------------ November Webinars for C, C++, Fortran Developers Accelerate application performance with scalable programming models. Explore techniques for threading, error checking, porting, and tuning. Get the most from the latest Intel processors and coprocessors. See abstracts and register http://pubads.g.doubleclick.net/gampad/clk?id=60136231&iu=/4140/ostg.clktrk _______________________________________________ Stellarium-pubdevel mailing list Stellarium-pubdevel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/stellarium-pubdevel
