On Mon, Apr 13, 2009 at 04:58:12PM +0200, Sebastian Harl wrote: > Hi Kevin, > > On Sun, Apr 12, 2009 at 03:57:44AM -0500, kevin brintnall wrote: > > diff --git a/src/librrd.sym.in.in b/src/librrd.sym.in.in > > index b4e8484..0a7a699 100644 > > --- a/src/librrd.sym.in.in > > +++ b/src/librrd.sym.in.in > > @@ -32,6 +32,7 @@ rrd_new_context > > rrd_open > > rrd_parsetime > > rrd_proc_start_end > > +rrd_random > > rrd_read > > rrd_resize > > rrd_restore > > > diff --git a/src/rrd_tool.h b/src/rrd_tool.h > > index f58cd78..ddcd2a9 100644 > > --- a/src/rrd_tool.h > > +++ b/src/rrd_tool.h > > @@ -119,6 +119,8 @@ int rrd_fetch_fn_libdbi(char *filename, enum cf_en > > cf_idx, > > char *a, > > char *b); > > > > + long rrd_random(void); > > + > > #endif /* _RRD_TOOL_H */ > > > > #ifdef __cplusplus > > Why did you include rrd_random in librrd.sym (which tells the compiler > to export that symbol) but then added the definition to rrd_tool.h which > is supposed to be a "private" header?
I don't intend that any third party callers would invoke it.. It's exported because rrdcached also needs that symbol. You're probably right - we should put it into a public header instead. -- kevin brintnall =~ /kbr...@rufus.net/ _______________________________________________ rrd-developers mailing list rrd-developers@lists.oetiker.ch https://lists.oetiker.ch/cgi-bin/listinfo/rrd-developers
