> One issue I did note is that the alignment between the cross from > plpoin and plstring is slightly different. (I noticed this by accidently > leaving the plpoin call in as well as the plstring one). > > I haven't investigated further yet to know which one is "right" but > it would be worth a look. >
Hi Andrew: I have changed the subject to alignment for obvious reasons. I have been keeping track of such alignment issues with examples/python/test_circle.py. Would you add at least a plpoin and plstring page to that example with the cross? Here is the current status with -dev xwin. Of course, the two pages, "Large Circle + Light Diagonal Cross with plptex" and "Plplot Encoded Unicode Indexed Asterisk with plptex" do not work for this non-unicode device driver. "Ascii asterisk with plptex" and "Plplot Encoded Hershey Indexed Asterisk with plptex" produce a yellow vertical which is slightly misaligned with the red vertical behind it. The next two pages, "Asterisk with plsym" and "Asterisk with plpoin" produce a slightly different alignment, then "Plplot Encoded Hershey Indexed Asterisk with plstring" goes back to the original alignment. These are all such subtle effects that they might be dismissed as the result of roundoff error in the 16-bit internal representation of PLplot positions, but I am not entirely sure of that interpretation, and you may find a different story if you repeat the asterisk experiments with a cross instead. Here is the current status for the unicode-aware -dev qtwidget and -dev xcairo. "Large Circle + Light Diagonal Cross with plptex" produces a "crumpled" circle for the largest three magnifications but the others are fine for -dev qtwidget. I reported this as a Qt bug years ago to trolltech but never got a response, and I would be interested if anyone else here gets a good result (round circles for the highest three magnifications) for any version of Qt. -dev xcairo produces round circles at all magnifications. According to the gucharmap display for generic sans (which is transformed to DejaVu Sans on my system), the "Light Diagonal Cross" symbol has excellent vertical and horizontal centering within its box so that has been the symbol I used to tune up the vertical alignment of both the qt and cairo devices. You can see the results of that tuning on this page with the intersection of the various magnified crosses in complete agreement in position. In contrast, you can also see how the center of the circle drifts closer to the bottom (i.e., shows spurious alignment issues) as it is magnified, but that is because it is not cleanly vertically centered in its box according to the gucharmap display for DejaVu Sans. (By the way, if you try this example with the wxGC version of -dev wxwidgets you can see from how the center of the crosses drift with magnification demonstrating that both the vertical and horizontal positions need tuning for at least this version of -dev wxwidgets.) The asterisk pages of this example demonstrate consistent alignment between -dev xcairo and -dev qtwidgets. "Ascii asterisk with plptex" and "Plplot Encoded Unicode Indexed Asterisk with plptex" show the asterisk moving upward with magnification, but the gucharmap display explains that because the asterisk is centered high in its box (at least for the DejaVu Sans font). So I think those results are fine, and the much better centering you get with -dev xwin for "Ascii asterisk with plptex"is simply because the Hershey fonts have a vertically centered asterisk within its box. What actually bothers me for -dev xcairo and -dev qtwidgets is the remaining asterisk pages display consistently centered asterisks in complete disagreement with the plptex results and my explanation above about how the asterisk glyph is vertically misaligned within its box according to gucharmap. What do those text code paths know that plptex does not know? Has there been some extra spurious vertical adjustment in those cases? Or is there something missing from the plptex code path that causes misalignment unless you do some extra tuning for each device? 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 __________________________ ------------------------------------------------------------------------------ Try before you buy = See our experts in action! The most comprehensive online learning library for Microsoft developers is just $99.99! Visual Studio, SharePoint, SQL - plus HTML5, CSS3, MVC3, Metro Style Apps, more. Free future releases when you subscribe now! http://p.sf.net/sfu/learndevnow-dev2 _______________________________________________ Plplot-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/plplot-devel
