From: Andrew R. <and...@us...> - 2004-02-17 14:30:42
|
On Mon, Feb 16, 2004 at 01:18:05PM -0800, Alan Irwin wrote: > On 2004-02-16 18:44-0000 Andrew Ross wrote: > > > I've just committed a change to rename plarrows2 / plsarrow to plvect > > and plsvect to avoid any possible confusion with the old version. I've > > backed out the changes to plarrows so it will still operate just as it > > always did. This way we can depreciate and remove plarrows (which was > > never documented or part of the common API) without affecting the new > > interface. > > > > I've also added a fortran binding very much in the same style as the > > fortran plcon[t012] interface. All 3 examples produce identical results. > > Thanks, Andrew. > > I just now cvs updated, and make install with all your recent changes went > fine. > > Since I am not an API expert, I will confine my comments just to example 22, > and specifically c/x22c.c since the C++ and f77 versions mimic that. > > * Thanks for the colour change. I think it looks better. > > * That default arrow on the first page is probably something you inherited, > but I still think it looks fairly ugly. I suggest using the arrow defined > for the current second page the default. Then you could eliminate a page and > have the default style produce what is now the second page, and the > user-defined arrow be filled as in the current 3rd page. I've changed the default arrow style. I don't think anyone could disagree. The old plarrow still uses the old style if anyone cares. > * I would suggest using a more interesting vector field to serve for the > (new) first and second pages. Example 8 plots an interesting "rosen" > function. One possibility is to use the gradient of that function as the > vector field. > > * For similar "interest" reasons, I would suggest using the gradient of > the shielded potential of two charges in a conducting sphere (example 9, 5th > page) for your final page. > > * For that final page, I would also surround it by a circle (just like > example 9, 5th page) rather than a box. > > Andrew, I have suggested a large number of C example changes that entail a > fair amount of work, but as soon as I get some other PLplot projects > finished up this week (java, python, and gd linking issues, and rpms), and > assuming you agree with the example changes, I would be willing to help with > that effort as well as propagation of the 22nd C example to the other > interfaces. I've committed an alternative x22c.c. See what you think. I agree mine examples were a little dull. Not convinced these are better though. A good polar example seems particularly hard. Suggestions welcome. Andrew |