From: Alan W. I. <ir...@be...> - 2008-01-29 23:55:15
|
Today (revision 8195) I finished the infrastructure for supporting the installed Ada bindings and examples, and as a result the installed tree test of the Ada examples work for the first time on my (Debian testing) platform. Note I have followed the recommendations of http://www.ada-france.org/debian/debian-ada-policy.html for the install locations of the Ada bindings source files and the Ada library information files which are generated during the plplotadad library build. I encourage everybody (especially Jerry) to try install tree tests of the Ada examples following http://www.miscdebris.net/plplot_wiki/index.php?title=The_PLplot_install-tree_test to make sure there are no non-Linux platform issues. In the installed Ada examples Makefile I have separated the -aI (the install directory location of the Ada source for the libplplotadad build) and -aL (the install directory location of the Ada library information files generated by the libplplotadad build) options from the normal compiling and linking information delivered by pkg-config. The reason is gnatmake wants the -aI and -aL flags in an extremely special order, and I am not sure how to get pkg-config to deliver this order. Advice is welcome on this issue since ideally these details would be wrapped in pkg-config results which normally are hidden from the user. Despite the somewhat cumbersome nature of the current Makefile build directions, the result on my platform is good. The "make" command in examples/ada builds all the Ada examples rapidly with no repeat compilations (a good sign that gnatmake is finding all the required files it needs). And running plplot-test.sh from the parent directory gives good-looking results for all the Ada examples (except for some colour issues, see below). This new build-system infrastructure for the install tree is a big step forward for our Ada bindings and examples. All that is required to change those bindings and examples from experimental (with ENABLE_ada OFF by default) to mature (with ENABLE_ada ON by default) is to fix the current colour issues (which show up for the second page of example 2 and for example 5) and implement the remaining standard examples to test substantial additional parts of the Ada interface API. Currently there are 9 standard Ada examples implemented. Only 21 to go! 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); PLplot scientific plotting software package (plplot.org); 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 __________________________ |
From: Jerry <lan...@qw...> - 2008-01-30 03:49:51
|
On Jan 29, 2008, at 4:55 PM, Alan W. Irwin wrote: > Today (revision 8195) I finished the infrastructure for supporting the > installed Ada bindings and examples, and as a result the installed > tree test > of the Ada examples work for the first time on my (Debian testing) > platform. > Note I have followed the recommendations of > http://www.ada-france.org/debian/debian-ada-policy.html for the > install > locations of the Ada bindings source files and the Ada library > information > files which are generated during the plplotadad library build. Great news. Thanks, Alan > > I encourage everybody (especially Jerry) to try install tree tests > of the > Ada examples following > http://www.miscdebris.net/plplot_wiki/index.php? > title=The_PLplot_install-tree_test > to make sure there are no non-Linux platform issues. I'll do this soon. > > This new build-system infrastructure for the install tree is a big > step > forward for our Ada bindings and examples. All that is required to > change > those bindings and examples from experimental (with ENABLE_ada OFF by > default) to mature (with ENABLE_ada ON by default) is to fix the > current > colour issues (which show up for the second page of example 2 and for > example 5) and implement the remaining standard examples to test > substantial > additional parts of the Ada interface API. Currently there are 9 > standard > Ada examples implemented. Only 21 to go! I spent a little time on these weird color issues a while back but haven't come up with anything. Since it doesn't happen in all examples I thought I'd see if it turns up in any more as I get them ported to Ada to see if some kind of pattern emerges. I'm not messing with the color table as far as I know so it seems like some weird voodoo at work--but doesn't it always. I have completed the Ada example 8 (not yet submitted) and am working on example 9 which is harder than the others. With that done, we will have 1 through 10 and 12 working. Jerry |