Commit 9bfe5a applies a patch sent in by Torsten Martinsen, a clued-up git user. So perhaps this is already an early sign that our switch to git is going to pay off.
That patch improved the detection of Qt5 (required when the experimental option -DPLPLOT_USE_QT5=ON is set). For the previous version, I had to set all of PATH, CMAKE_LIBRARY_PATH, and CMAKE_INCLUDE_PATH appropriately to find Qt5 properly, but now for the changed build-system code, I only need to set PATH (and presumably not even that, if Qt5 is installed in a system location). To help test this commit, I epa_built Qt5.3.1. Compared to the previous Qt5 epa_build configuration (for Qt5.2.1) this configuration eliminated a lot of unneeded (by PLplot) Qt5 components. That work reduced the Qt5 build time down from roughly an hour to less than 20 minutes. Once Qt5.3.1 was available to me I successfully tested commit 9bfe5a on Linux using the CMake option -DPLPLOT_USE_QT5=ON, and the target test_all_qt (which does what its name implies). To discuss the overall status of PLplot qt device driver versus the Qt library, the preferred version of Qt is still Qt4 which has no character alignment or any other serious issues from the PLplot perspective. In contrast, for the Qt5 case I apply an empirically determined uniform vertical offset to the position of characters to work around the Qt5 alignment bug(s). The result is pretty good both for Qt5.2.1 and 5.3.1, but someone else has reported bad character alignment issues (where the offsets vary strongly from character to character so one uniform offset does not compensate very well) for 5.3.0. That said, a lot of users do have a need to try Qt5 so if you are in that group, I suggest you carefully pick the version (i.e., you should probably avoid 5.x.0 versions from our bad report for 5.3.0). At some point the more-or-less uniform vertical alignment issue for Qt5 will become fixed, and I will need to remove my compensation for that Qt5 bug in the PLplot code. But we may have to wait to, say Qt5.5.x, for the required Qt5 fix(es) to occur. (We had similar problems for the Qt4.x series until Qt4.5 or so.) In sum, use Qt5 with caution if you must. If development follows what happened with Qt4, we should expect a reliable version near Qt5.5 or so. For now if Qt5 is not a necessity, Qt4 is a more reliable choice. Alan __________________________ Alan W. Irwin Astronomical research affiliation with Department of Physics and Astronomy, University of Victoria (astrowww.phys.uvic.ca). Programming affiliations with the FreeEOS equation-of-state implementation for stellar interiors (freeeos.sf.net); the Time Ephemerides project (timeephem.sf.net); PLplot scientific plotting software package (plplot.sf.net); the libLASi project (unifont.org/lasi); the Loads of Linux Links project (loll.sf.net); and the Linux Brochure Project (lbproject.sf.net). __________________________ Linux-powered Science __________________________ ------------------------------------------------------------------------------ Want excitement? Manually upgrade your production database. When you want reliability, choose Perforce Perforce version control. Predictably reliable. http://pubads.g.doubleclick.net/gampad/clk?id=157508191&iu=/4140/ostg.clktrk _______________________________________________ Plplot-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/plplot-devel
