<html><body>
<p>Crass,</p>
<p>I'm currently working on setting up a reviewboard for citadel</p>
<p>( http://www.reviewboard.org/ )</p>
<p>maybe once its up and running we can use it for this too.</p>
<blockquote>
<div class="message_header"><span>Mo Mär 29 2010 11:50:15 EDT</span> <span>von "crass" <crass@berlios.de> </span> <span class="message_subject">Betreff: Re: [Freeassociation-devel] SVN fails regression tests?</span></div>
<div class="message_content"><tt></tt><br /> <tt>Art,</tt><br /> <tt></tt><br /> <tt>I have the impetus to fix some, if not all, of the failing tests.</tt><br /> <tt>Considering that other developers on this list seem not very responsive</tt><br /> <tt>until something stops working for them, how do you recommend me</tt><br /> <tt>submitting my proposed changes?</tt><br /> <tt></tt><br /> <tt>I see two main methods of going about this:</tt><br /> <tt>1. I directly commit my changes as I determine them suitable to do so,</tt><br /> <tt>then I resolve issues people encounter later and revert the changes if</tt><br /> <tt>need be.</tt><br /> <tt></tt><br /> <tt>2. I post patches to the list for review by interested parties, then</tt><br /> <tt>commit. Ideally I'd get feedback from at least one person, but I'm</tt><br /> <tt>skeptical of that. I would like to get feedback because I'm not an</tt><br /> <tt>expert in RFC-2445, though I do read it.</tt><br /> <tt></tt><br /> <tt>I've been making changes to the tests themselves that are clearly</tt><br /> <tt>wrong, on the test side (mainly failures due to incorrect line</tt><br /> <tt>endings). A lot of the rest of the failing tests seem to be related to</tt><br /> <tt>timezone conversions. From what I've seen of the tests they do appear</tt><br /> <tt>to be geared towards standards compliance, but I don't think they are</tt><br /> <tt>failing because we are trying to maintain compatibility with other</tt><br /> <tt>software. And from what I've seen from the code, I don't see that many</tt><br /> <tt>compatibility hacks (at least labeled). So, I'm a little concerned</tt><br /> <tt>that I might not realize that a test is failing because of a</tt><br /> <tt>compatibility hack. In this case, I'll probably just have to break</tt><br /> <tt>compatibility and restore it when notified along with a comment making</tt><br /> <tt>note of the hack. What do you think about this?</tt><br /> <tt></tt><br /> <tt>Glenn</tt><br /> <tt></tt><br /> <tt>On Sun, 28 Mar 2010 07:42:16 -0400</tt><br /> <tt>"IGnatius T Foobar" <room_freeassociation-devel@uncensored.citadel.org></tt><br /> <tt>wrote:</tt><br /> <tt></tt><br />
<blockquote>
<blockquote><tt>Anyone noticed that the regression tests fail in SVN? I think we </tt><br /> <tt>should have a policy that all code commits cause no tests to fail </tt><br /> <tt>(after we get the tests to a point where they are passing).</tt><br /> <tt>Anyone else have opinions on this? Is there any reason not to fix</tt><br /> <tt>the regression tests? </tt><br /> <tt></tt><br /></blockquote>
<tt>They were broken when we (the current maintainers) originally</tt><br /> <tt>inherited the code. While I agree that it needs to be fixed, I'm not</tt><br /> <tt>quite sure where to begin. </tt><br /> <tt></tt><br /> <tt>I suppose the first step is to determine whether it's the code or</tt><br /> <tt>the tests that are broken. </tt><br /> <tt></tt><br /> <tt>It's also possible that the tests are built to determine standards</tt><br /> <tt>compliance, and they've just always been that way because there are</tt><br /> <tt>things we just don't do according to the standards. If that's the</tt><br /> <tt>case then they'll always fail because we prioritize interoperability</tt><br /> <tt>with other software that's actually out there in the real world.</tt><br /> <tt></tt><br /> <tt></tt><br /> <tt>Dunno. Suggestions are welcome. Code contributions are even more</tt><br /> <tt>welcome. :) </tt><br /> <tt></tt><br /> <tt>-- Art </tt><br /> <tt></tt><br /> <tt></tt><br /> <tt>------------------------------------------------------------------------------</tt><br /> <tt>Download Intel® Parallel Studio Eval</tt><br /> <tt>Try the new software tools for yourself. Speed compiling, find bugs</tt><br /> <tt>proactively, and fine-tune applications for parallel performance.</tt><br /> <tt>See why Intel Parallel Studio got high marks during beta.</tt><br /> <tt><a href="http://p.sf.net/sfu/intel-sw-dev" target="webcit01">http://p.sf.net/sfu/intel-sw-dev</a></tt><br /> <tt>_______________________________________________</tt><br /> <tt>Freeassociation-devel mailing list</tt><br /> <tt>Freeassociation-devel@lists.sourceforge.net</tt><br /> <tt>https://lists.sourceforge.net/lists/listinfo/freeassociation-devel</tt><br /> <tt></tt><br /></blockquote>
<tt>------------------------------------------------------------------------------</tt><br /> <tt>Download Intel® Parallel Studio Eval</tt><br /> <tt>Try the new software tools for yourself. Speed compiling, find bugs</tt><br /> <tt>proactively, and fine-tune applications for parallel performance.</tt><br /> <tt>See why Intel Parallel Studio got high marks during beta.</tt><br /> <tt><a href="http://p.sf.net/sfu/intel-sw-dev" target="webcit01">http://p.sf.net/sfu/intel-sw-dev</a></tt><br /> <tt>_______________________________________________</tt><br /> <tt>Freeassociation-devel mailing list</tt><br /> <tt>Freeassociation-devel@lists.sourceforge.net</tt><br /> <tt>https://lists.sourceforge.net/lists/listinfo/freeassociation-devel</tt><br /> <br /></div>
</blockquote>
<p> </p>
</body></html>