Xiaofan Chen wrote: > > Review is static analysis - studying a change, understanding what it > > does and why, and how that affects any surrounding code. > > > > Testing is dynamic analysis - observing the results of a change, > > usually either trying to confirm progressions or detect regressions. > > > > You and everyone else can do review in gerrit, even if nobody does > > testing. > > That is what I am doing now. But my level of proficiency in > coding limits me only to relatively simple patches.
It is a really great contribution anyway! Even if it's only a little here and there and only for some commits, it's still *something* and that is always much much better than nothing! > For things I can test and I care, I will actually try to test as > well, eg, the mpsse codes and the J-link patches during the 0.6.0 > release cycle. Yep, I remember. Testing is also a really great contribution! Especially when everyone has different equipment. > You are certainly a better reviewer than I. Thanks! And I think you are a much better tester than I. > > The reality of any hardware-related project is that testing is much > > more tricky than with software-only projects, so it's easy to be lazy > > about creating tests and that's fine. > > I am a hardware engineer myself, so I put testing to be higher status. I don't think it's fair to say that either one type of analysis is universally better than the other - I think the two very much complement each other. Review can miss problems that testing can catch, and testing can miss problems that review can catch. The ideal is to have as much as possible of both! :) Testing sometimes has the very nice advantage that it can be automated. That's the very best - when nobody has to do manual work, and problems can be detected "for free". > But I agree with you it is tricky to test OpenOCD > against different adapter and different targets. Yes. I guess everyone is aware, and somehow it is what it is. All testing done by anyone is in any case very valuable. > > Conversely, if someone does testing but have not done any review, > > that is *also* valuable information to store in gerrit. That is what > > the review score 0 is for. "I have not done review of this, but I add > > this comment." > > I think that actually can be a "1", "tested by" is good to have. It could be nice to indicate testing in a structured way, but until there is some convenient way I completely agree that giving a +1 review score works fine as an indicator for successful testing. //Peter ------------------------------------------------------------------------------ Symantec Endpoint Protection 12 positioned as A LEADER in The Forrester Wave(TM): Endpoint Security, Q1 2013 and "remains a good choice" in the endpoint security space. For insight on selecting the right partner to tackle endpoint security challenges, access the full report. http://p.sf.net/sfu/symantec-dev2dev _______________________________________________ OpenOCD-devel mailing list OpenOCD-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/openocd-devel
