<div dir="ltr"><div class="gmail_extra"><div class="gmail_quote">On Mon, Jan 25, 2016 at 2:07 AM, Tom Lane <span
dir="ltr"><<ahref="mailto:tgl@sss.pgh.pa.us" target="_blank">tgl@sss.pgh.pa.us</a>></span> wrote:<br
/><blockquoteclass="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><span
class="">CoreyHuinker <<a href="mailto:corey.huinker@gmail.com">corey.huinker@gmail.com</a>> writes:<br /> >
Incidentally,is there a reason behind the tendency of internal functions<br /> > to avoid parameter defaults in
favorof multiple pg_proc entries?<br /><br /></span>Yes: you can't specify defaults in pg_proc.h.<br /><br /> We work
aroundthat where absolutely necessary, see the last part of<br /> system_views.sql. But it's messy enough, and
bug-proneenough, to be<br /> better avoided --- for example, it's very easy for the redeclaration<br /> in
system_view.sqlto forget a STRICT or other similar marking.<br /><br /> Personally I'd say that every one of the
existingcases that simply has<br /> a default for the last argument was a bad idea, and would better have<br /> been
donein the traditional way with two pg_proc.h entries.<br /><br /> regards, tom lane<br
/></blockquote></div><br/></div><div class="gmail_extra">...which answers my follow up question where make_interval was
gettingthe defaults I saw in the table but not the .h file. Thanks!</div><div class="gmail_extra"><br /></div></div>
