Re: msvcrt: Few failing tests for O_TEXT files

[Richard Cohen]

Vitaliy Margolen wrote:
> - _lseek(tempfd, -1, FILE_END);
> - ok(_read(tempfd,btext,LLEN) == 1, "_read expected 0 got '\n'\n");

That doesn't look right.

Richard.

msvcrt: Few failing tests for O_TEXT files

[Vitaliy Margolen]

Saturday, October 8, 2005, 5:53:13 AM, Richard Cohen wrote:
> Vitaliy Margolen wrote:
> > - _lseek(tempfd, -1, FILE_END);
> > - ok(_read(tempfd,btext,LLEN) == 1, "_read expected 0 got '\n'\n");
>
> That doesn't look right.

What's not right about it? It moves to the last byte and tries to read it. Which
is '\n' and should not be read. Tested on windows and that's what it does.

Vitaliy

msvcrt: Few failing tests for O_TEXT files

[Richard Cohen]

Vitaliy Margolen wrote:
> Saturday, October 8, 2005, 5:53:13 AM, Richard Cohen wrote:
> > > - ok(_read(tempfd,btext,LLEN) == 1, "_read expected 0 got '\n'\n");
>
> What's not right about it?

ret = _read(tempfd,btext,LLEN);
ok(ret == 0, "_read expected 0, got %d\n", ret);

^
           |

:)

Richard.

msvcrt: Few failing tests for O_TEXT files

[Vitaliy Margolen]

Saturday, October 8, 2005, 10:05:48 AM, Richard Cohen wrote:
> Vitaliy Margolen wrote:
> > Saturday, October 8, 2005, 5:53:13 AM, Richard Cohen wrote:
> > > > - ok(_read(tempfd,btext,LLEN) == 1, "_read expected 0 got '\n'\n");
> >
> > What's not right about it?
>
> ret = _read(tempfd,btext,LLEN);
> ok(ret == 0, "_read expected 0, got %d\n", ret);
>        ^
>        |
>
> :)

Ah, doh. Just forgot to change an error message ;)

Vitaliy

msvcrt: Few failing tests for O_TEXT files

[Uwe Bonnes]

> > > > > "Richard" == Richard Cohen  writes:

Richard> Vitaliy Margolen wrote:
    >> + _lseek(tempfd, -1, FILE_END); + ok(_read(tempfd,btext,LLEN) == 1,
    >> "_read expected 0 got '\n'\n");

Richard> That doesn't look right.

Why do we like to play "question and answer" without giving the answer?

What oesn't look right, how would you code it?