Re: Regression tests versus the buildfarm environment

<br /><br /> On 08/11/2010 02:12 PM, Peter Eisentraut wrote: <blockquote
cite="mid:1281550347.26522.13.camel@vanquo.pezone.net"type="cite"><br /><blockquote type="cite"><pre wrap="">Even if
youdon't, changing this would only mean that you
 
couldn't safely run "make check" concurrently in multiple branches.
</pre></blockquote><pre wrap="">
That's exactly the point.  The original discussion is here:
<a class="moz-txt-link-freetext"
href="http://archives.postgresql.org/message-id/491D9935.9010200@gmx.net">http://archives.postgresql.org/message-id/491D9935.9010200@gmx.net</a>

</pre></blockquote><br /><br /> You original email said:<br /><br /><blockquote><tt>For some historic reasons, I have
mylocal scripts set up so that they </tt><tt>build development instances using the hardcoded port 65432.<br
/></tt></blockquote><br/> I think my response would be "Don't do that".<br /><br /> Having said that, maybe we could
reasonablyuse something like DEF_PGPORT + 10 * major_version + minor_version in the calculation and advise buildfarm
memberswith multiple animals to keep their port ranges say, 200 or more apart.<br /><br /> But maybe we should just
stickwith my earlier advice :-)<br /><br /> cheers<br /><br /> andrew<br /><br /><br />