<div class="gmail_quote">On Thu, Apr 9, 2009 at 12:59 PM, Paul Oppenheim (Poppy Linden) <span dir="ltr"><<a href="mailto:poppy@lindenlab.com">poppy@lindenlab.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
<div class="im">Robert Dailey wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
Do you have unit tests that require failure to compile in order to be successful? If you do, placing these particular unit tests in a project with other tests that must compile to succeed will interrupt the process, depending on the order of compilation. <br>
</blockquote>
<br></div>
I don't have that particular case, but I could support it. With my scenario each test is its own binary. This is normally messy, but because tests are built in sub-projects it's not noticeable in the main build. Each group of tests looks like one test in the main build, but are completely separate in the sub-build.</blockquote>
<div><br></div><div>Ah, I see now. It was not immediately obvious to me what your code was doing. I didn't notice it was making each CPP file basically compile itself (Which is still just an assumption).</div><div><br>
</div>Where is the implementation for "set_source_files_properties"?</div>