<html>
<head>
<style>
.hmmessage P
{
margin:0px;
padding:0px
}
body.hmmessage
{
FONT-SIZE: 10pt;
FONT-FAMILY:Tahoma
}
</style>
</head>
<body class='hmmessage'>Suggestion: Can we get our CI build server to 'treat warnings as errors'? Then anyone who introduces warnings will get a flip on the nose immediately.<BR><BR>
Referring back to the Code Standards we once had : do NOT introduce 'stub' or 'future use' fields or functions. Code for what you need _now_. Tomorrow may never come. And, judging from our code, in too many cases it never did.<BR>
<BR>
I have a refactoring tool that greys out unused code in Visual Studio, so I see it very clearly.<BR>
<BR>
At some point, I used to just remove it without further ado, but nowadays.. what do you all say?<BR><BR>Best regards,<BR>Stefan Andersson<BR>Tribal Media AB<BR> <BR>Join the 3d web revolution : <A href="http://tribalnet.se/" target=_blank>http://tribalnet.se/</A><BR> <BR><BR><BR><BR>
<HR id=stopSpelling>
<BR>
> Date: Fri, 27 Jun 2008 12:32:46 +1200<BR>> From: mariusz@nowostawski.org<BR>> To: opensim-dev@lists.berlios.de<BR>> Subject: Re: [Opensim-dev] OpenSim codebase statistics.<BR>> <BR>> Mike Mazur wrote:<BR>> <BR>> [...]<BR>> <BR>> > There are actually more unit tests (though not many more). Try pointing<BR>> > nunit at bin/OpenSim.Region.ScriptEngine.Common.Tests.dll as well.<BR>> <BR>> Great - thanks. 20 extra existing tests added to the stats. I hope more <BR>> will come.<BR>> <BR>> >> (note: building is done on the 64-bit Opteron Linux server, and all<BR>> >> the build time is in seconds and the warnings are on Linux with mono<BR>> >> 1.9.1, so they may vary slightly between particular compiler used).<BR>> > <BR>> > From what I understand there are some 32-bit binaries that are provided<BR>> > in the SVN checkout. I'm not sure to what extent they're referenced at<BR>> > build time, but since you're building on a 64-bit platform, perhaps they<BR>> > contribute some of the warnings?<BR>> > <BR>> > Somebody more familiar with this can probably clarify further.<BR>> <BR>> As far as I can see none of the warnings is related to the 32-bit vs. <BR>> 64-bit dependencies - all the warnings are compiler warnings that are <BR>> cause by the source code not following the ECMA rules, such as:<BR>> The variable is declared/assigned but its value is never used<BR>> The private field is declared/assigned but its value is never used<BR>> The private method is declared but never used<BR>> The method hides inherited member<BR>> something obsolete is used,<BR>> and so on.<BR>> <BR>> What I meant is that various compilers may or may not be religous about <BR>> all of these warnings, and also it may depend on the particular compiler <BR>> flags being used. We use the standard build process as it is setup on <BR>> the SVN trunk itself.<BR>> <BR>> <BR>> -- <BR>> cheers<BR>> Mariusz<BR><BR></body>
</html>