From: Alan W. I. <ir...@be...> - 2005-04-21 21:05:57
|
Per, I am changing the subject line to something more appropriate. Comments in context below.... Alan On 2005-04-21 18:37+0200 Per Persson wrote: > I noticed the following paragraph in the release notes: > The driver is unicode-enabled, however it currently only supports the > default OS X fonts which are not TrueType. > > That statement might be a misinterpretation of a comment I made some time > ago, let me try to clarify what I meant. > > I recommended that the driver only use the default fonts, by which I the > fonts that come preinstalled with OS X and thus are guaranteed to exist. > > AquaTerm, and hence the driver, support any font installed on OS X which may > be in a number of different formats including, but not limited to, TrueType. > BTW, TrueType was developed by Apple (and to some extent MS) in the 80s. > Sadly, while OS X is fully unicode aware, not all font families support the > full range of glyphs, particular the common ones such as Helvetica, Times and > Courier. It is good that aqt.c has the possibility of TrueType fonts since they tend to have a much richer set of characters than postscript fonts. As a telling example of this, compare results for ps.c (and probably aqt.c although I don't have access to it) on example 23 with what is possible with gd.c devices such as png that have access to TrueType fonts via plfreetype.c (See http://plplot.sourceforge.net/examples/demo23.php for png results.) The first 4 pages of that example are specifically optimized for type 1 postscript fonts so ps.c does well there, but as you get beyond those, there are a lot of symbols missing compared to what is available for the TrueType fonts. And that example just scratches the surface of what is possible with TrueType. For example, with TrueType fonts you are also able to do a variety of languages such as in http://plplot.sourceforge.net/examples-data/demo24/x24.01.png. As I see it, the current problem with aqt.c is it follows what is done for ps.c and hard-codes the font names. Furthermore, because they are hard-coded you are forced to go with the lowest-common-denominator approach (postscript fonts). But ps.c is a special case where there is not a lot of choice other than postscript fonts (at least until Rafael is able to implement his TrueType on-the-fly conversion idea for ps.c). To increase flexibility, I think it would be better for aqt.c to follow the font configuration model used by plfreetype.c. That would allow your users to ./configure the fonts (with default lowest-common-denominator postscript font names but with TrueType fonts allowed for those users that have them installed). The build-time font configuration would need to modify cf/aqt.ac to do this following what is done for fonts in cf/plfreetype.ac. Also, to increase font flexibility at run time (as opposed to software build configuration time) you might want to use environment variables (as we do with plfreetype.c) to allow users to select fonts that are different from the ones that are build-time configured. Anyhow, that is my $0.02 worth on how Hazen or you could proceed to make the TrueType fonts both build-time and run-time configurable for aqt.c when the user happens to have them installed. I think the results would be well worth the effort. Alan __________________________ Alan W. Irwin email: ir...@be... phone: 250-727-2902 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); PLplot scientific plotting software package (plplot.org); the Yorick front-end to PLplot (yplot.sf.net); the Loads of Linux Links project (loll.sf.net); and the Linux Brochure Project (lbproject.sf.net). __________________________ Linux-powered Science __________________________ |