Re: Fwd: Fwd: Assume
"Daniel M. Pomerantz" <[email protected]>
| Newsgroups | gmane.comp.windows.dotnet.nunit.devel |
|---|---|
| Message-ID | <[email protected]> |
Thanks Charlie. I'm not really in need of this feature, and whether the test never shows up or shows up as not run is not too important to me. I would think showing up as not run would be more ideal, but again, not my request. I'm just trying to suggest ways to implement it. As for Assumption not meaning what it should mean, I'm fine with changing the name. I think Kelly only picked the name because of the theory document. I prefer something built on the idea of "Conditional," but I am really ambivalent here. Just trying to forward ideas. dmp Charlie Poole wrote: > Hi Daniel, > > >> Just brainstorming here.... >> >> Right now the runners have three results: Pass, Fail, and >> Did Not Run. >> Sorta like a nullable Boolean. I don't see that the runner >> would have to alter that at all. If the Assumptions fail, >> then this test falls into the "Did Not Run" category. If >> some one chose to use ASSumptions rather than ASSerts, then >> this is the desired results, and it's up to the developer to >> deal with it. If they wanted a failure, they would not have >> used ASSume. >> > > But in this case, you don't want a "Did not Run" You don't > want the test to show up at all. Which is my point. Since the > test is created from your iterator at test-load-time but run > at test-run-time, the test will be listed and shown as not > run. You're building tests here and what I thik you really > want is for the test not to be built - to never exist! > > > >> I can see potential bennefit of eiter wrapping a Test inside >> of another so that the outter test only passes if the second >> one runs, but then I wonder, why bother going through all >> those hoops if the end result was /a/ failing test. Then >> there's the question of should a continuous build system >> report these skipped tests any differently than other skipped >> tests. I just don't see if. Maybe I'm not seeing the big >> picture, but if a developer wanted a test with invalid data >> to fail, he would use Assert. If he wants it to skip, this >> gives him an option that wasn't previously there. Please >> feel free to tell me what I've overlooked something. >> > > My only point is that the word Assume does't mean what you > are doing here in normal English. > > Charlie > > >> I thought of something else which may or may not be helpful here. >> Extending the theme created by the ExpectedException and >> Expicit attributes, I thought a [Conditional] attribute might >> be helpful. It might serve no purpose other than >> information, but if a test is decorated as Conditional, >> Assumptions cause a "Not-Run" situation. If the method is >> not decorated as such, ASSumptions cause a failure. Any >> bennefit seen here? >> >> dmp >> >> Vaughn, Clifton wrote: >> >>> I've read one paper, but I don't recall it explaining great detail >>> about the Assume.That. I was still left with the question >>> >> of how the >> >>> runner would indicate that the data didn't meet an >>> >> assumption when you >> >>> expected it to. Maybe I'm still missing something though. >>> >>> Thanks >>> >>> cliff >>> >>> -----Original Message----- >>> From: [email protected] >>> [mailto:[email protected]] On Behalf Of >>> Kelly Anderson >>> Sent: Friday, February 15, 2008 3:14 PM >>> To: [email protected] >>> Subject: Re: [nunit-developer] Fwd: Fwd: Assume >>> >>> On Fri, Feb 15, 2008 at 11:29 AM, Vaughn, Clifton >>> <[email protected]> wrote: >>> >>> >>>> I don't think it needs to cause an error, but it should signal that >>>> >>>> >>> the >>> >>> >>>> data didn't meet the assumptions unless along with the data you >>>> >>>> >>> included >>> >>> >>>> a flag to signal when an error should be thrown. >>>> >>>> Either a signature like this: >>>> >>>> Check.That(meetsAssumptions, 2+2, Is.EqualTo(data) ) >>>> >>>> Or meetsAssumptions is added as a condition. >>>> >>>> Thanks >>>> >>>> >>> Cliff, out of curiosity, have you read any of the Theory >>> >> papers yet? I >> >>> am curious only as to the context of your comment. >>> >>> -Kelly >>> >>> >>> >> ---------------------------------------------------------------------- >> >>> -- >>> - >>> This SF.net email is sponsored by: Microsoft Defy all challenges. >>> Microsoft(R) Visual Studio 2008. >>> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ >>> _______________________________________________ >>> nunit-developer mailing list >>> [email protected] >>> https://lists.sourceforge.net/lists/listinfo/nunit-developer >>> >>> >>> >> ---------------------------------------------------------------------- >> >>> --- This SF.net email is sponsored by: Microsoft Defy all >>> >> challenges. >> >>> Microsoft(R) Visual Studio 2008. >>> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ >>> _______________________________________________ >>> nunit-developer mailing list >>> [email protected] >>> https://lists.sourceforge.net/lists/listinfo/nunit-developer >>> >>> >>> >>> >> -------------------------------------------------------------- >> ----------- >> This SF.net email is sponsored by: Microsoft Defy all >> challenges. Microsoft(R) Visual Studio 2008. >> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ >> _______________________________________________ >> nunit-developer mailing list >> [email protected] >> https://lists.sourceforge.net/lists/listinfo/nunit-developer >> >> > > > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Microsoft > Defy all challenges. Microsoft(R) Visual Studio 2008. > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > _______________________________________________ > nunit-developer mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/nunit-developer > > > ------------------------------------------------------------------------- This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/