<div class="gmail_quote">On Fri, Sep 7, 2012 at 11:32 PM, Barry Smith <span dir="ltr"><<a href="mailto:bsmith@mcs.anl.gov" target="_blank">bsmith@mcs.anl.gov</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
Oh, another that occurred to me is figures; If slide 27 and 38 included a bunch of figures (stored in some silly subdirectory called figs/ or figures/ or who knows what with absurd names) I have to remember to find out all the file names and copy over all of them and fix the paths (if I use the correct subdirectory name called Figures :-). </blockquote>
</div><br><div>\graphicspath{{./Figures}}</div><div><br></div><div>You do realize you are basically arguing for "templates" instead of "libraries". Instead of using PETSc, people should just copy the GMRES from someone else, then modify so that it calls their MatMult and PCApply, bug fixes and composability be damned.</div>