shell32: FOF_MULTIDESTFILES must be set when copying files into directory

[Reece Dunn]

2008/10/18 Vitaly Perov <vitperov at etersoft.ru>:
> The program is Garant 7 F1 (Russian legal system database). I don't think you
> know this program.
> It is not a real world bug, but there are some significant bugs in
> SHFileOperationW. I don't think a lot of programs use SHFileOperationW to
> move a files, so these fixes maybe don't fix bugs in bugzilla, but is it
> criterion why pathes wouldnt't be applied?
> I solve these bugs for Etersoft. Now in Etersoft wine Garant installs
> perfectly. But I want to fix these bugs in wine.
> I don't think my patches are perfect. I can fix them if you say what's wrong.
> I used to write a tests, but even a tests are ignored. Without any
> explanation. I was resending them, but had no answer again. Sometimes I think
> it's just waste of time.

Having tests (as James has mentioned in this thread) improves the
chances of patches getting in if they relate to bugs in Wine's
implementation of the Windows API. Having applications that are
affected by these bugs can be helpful, but is not a requirement (I
fixed some issues with Wine's implementation of SHCreateStreamOnFile
that didn't have any known applications that affect them, but I did
this by writing test cases for it to highlight where Wine's behaviour
was wrong; I wrote the patches to fix the issues only after I had the
tests in and working).

Have you tried asking on the #winehackers IRC channel? Sometimes, that
is better than the mailing lists to get a response. Especially asking
Alexandre Julliard directly on there.

But also, when you do get a response, listen to it. Tests are
necessary, especially when the code changes some fundamental
behaviour, or code that is (or appears to be) working properly.

Also, when did you submit the other patches? Patchwatcher has been put
in place to run the conformance tests to pick up patches with issues.
If they fail there, they won't be applied.

> I have a lot of patches which I could send, but I think they wouldn't be
> applied just because they don't fix any "real world bug".

All bugs are important, even those that don't have any known
applications that are affected by them. The point is that you need to
prove that Windows is behaving in one way and that Wine is behaving in
another.

It can sometimes be difficult to get patches accepted into Wine, and
sometimes you need to rework a patch or series of patches several
times before they are accepted. The patches you have will be useful to
the community.

If you don't want to continue work on the patches (but I urge you to
be more persistent, and try again), I would suggest you send them
anyway so that other people can work on them so they can be accepted
as a joint effort.

- Reece