147794f00a80409a41e2fee989218a97bb67b99a
angie
  Fri Apr 1 11:11:29 2011 -0700
Bugs #3475, #3476 (hgTracks map links messed up for SNPs, omimGene):These were side-effects of my commit a8d97808 Mar. 11, in which I
enabled next-prev-exon for all simple beds (which, because they don't
have exons/blocks, just means zap-to-edge), instead of only for dbVar
where it was requested.  It turns out that way too many bed-based
tracks with custom drawItems and/or drawItemAt methods are missing
something necessary for the nextItem stuff to work.  In this fix,
instead of hooking up zap-to-edge for all simple bed, I hook it up
only to GVF (dbVar/ISCA).

diff --git src/hg/hgTracks/gvfTrack.c src/hg/hgTracks/gvfTrack.c
index cdfd100..7c3ee6c 100644
--- src/hg/hgTracks/gvfTrack.c
+++ src/hg/hgTracks/gvfTrack.c
@@ -107,16 +107,17 @@
     hashAdd(nameHash, gvf->name, gvf);
     }
 sqlFreeResult(&sr);
 hFreeConn(&conn);
 slSort(&list, gvfHierCmp);
 tg->items = list;
 }
 
 void gvfMethods(struct track *tg)
 /* Load GVF variant data. */
 {
 bedMethods(tg);
 tg->canPack = TRUE;
 tg->loadItems = gvfLoad;
 tg->itemColor = gvfColor;
+tg->nextPrevExon = simpleBedNextPrevEdge;
 }