Re: listview: Fix LISTVIEW_CancelEditLabel to not save the changed text

[Nikolay Sivov]

Jeff Latimer wrote:
> ---
>  dlls/comctl32/listview.c       |    2 +-
>  dlls/comctl32/tests/listview.c |    2 +-
>  2 files changed, 2 insertions(+), 2 deletions(-)
> Hi, Jeff.

This is obviously wrong. Test will fail on native systems after your patch.
What is a problem with current code, any bugs reported?

listview: Fix LISTVIEW_CancelEditLabel to not save the changed text

[Jeff Latimer]

On 06/12/09 18:56, Nikolay Sivov wrote:
> Jeff Latimer wrote:
> > ---
> >  dlls/comctl32/listview.c       |    2 +-
> >  dlls/comctl32/tests/listview.c |    2 +-
> >  2 files changed, 2 insertions(+), 2 deletions(-)
> > Hi, Jeff.
>
> This is obviously wrong. Test will fail on native systems after your 
> patch.
> Why do you say that?  It tests ok on XP.  It was not being run on XP due to
> if (!load_v6_module(&ctx_cookie, &hCtx))
> so it looked if it was executing ok.  When I forced it to run, the 
> original failed.  In any case the test comment says it is to check that 
> the string "test" does not become "test1", however the test is
> ok(strcmp(buff, test1) == 0, "Expected label text not to change\n");
> which is the opposite.
> What is a problem with current code, any bugs reported?
> No bug.

listview: Fix LISTVIEW_CancelEditLabel to not save the changed text

[Nikolay Sivov]

Jeff Latimer wrote:
> On 06/12/09 18:56, Nikolay Sivov wrote:
> > Jeff Latimer wrote:
> > > ---
> > >  dlls/comctl32/listview.c       |    2 +-
> > >  dlls/comctl32/tests/listview.c |    2 +-
> > >  2 files changed, 2 insertions(+), 2 deletions(-)
> > > Hi, Jeff.
> >
> > This is obviously wrong. Test will fail on native systems after your 
> > patch.
> > Why do you say that?  It tests ok on XP.
> > No, it doesn't.
> >   It was not being run on XP due to
> > if (!load_v6_module(&ctx_cookie, &hCtx))
> > Then it's not supposed to run on your system.
> > so it looked if it was executing ok.  When I forced it to run, the 
> > original failed.
> > You broke a logic This test needs v6 module to run.
> >   In any case the test comment says it is to check that the string 
> > "test" does not become "test1", however the test is
> > ok(strcmp(buff, test1) == 0, "Expected label text not to change\n");
> > which is the opposite.
> > So the comment is wrong, not the test condition. Please send a patch to 
> > change comment.