[Mono-devel-list] Re: Analysis of W3C regressions

[Atsushi Eno]

Hi Andrew,

Ok, as we talked on IRC that number of regression stuff was my
foolish mistake :( Now things should be fixed (actually those
tests just exposed other bugs that used to be there). As long
as I tried existing Test/System.Xml/standalone_tests there is
no regression anymore.

There is still a case that your W3C test says regression,
but am not sure what this means. Here I paste the entire content:

*** Test failed:	rmt-e2e-15h	type:invalid 
nonValidatingPassed:True,validatingPassed:True	
Element content can't contain entity reference if replacement text is
character reference to whitespace

Maybe description on what is logged onto "failed.lst" (in README)
would be helpful? :-)

Thanks
Atsushi Eno


Andrew Skiba wrote:
> Hello.
> 
> Atsushi Eno wrote:
>> I need concrete number unlike just saying "there are regressions".
> 
> I did not just say "there are regressions". I spent few hours on 
> debugging, digging svn and creating patch that can be applied to svn 
> latest, to find what exactly led to these regressions.
> 
> And concrete number is 62 regressions out of them I know the origin of 
> 21, this is revision 41840.
> 
>> For now what I got from r41840 is
> 
> The regressions were introduced by r41840, so you should get r41839 to 
> see the difference ;-)
> 
> I attached patch which undoes the r41840, did you try to apply the patch 
> to svn latest and run w3c tests? You can see that it's really a part of 
> 41840 by issuing
> 
> svn diff -r41839:41840>41840.diff
> 
> then compare 41840.diff to the patch I have sent.
> 
> Andrew.
> 
>