rebar issues

[Matt Chapman]

Hi,

I'm having some rebar issues with a particular application (Canon
File Viewer Utility).

The rebar is meant to look something like:

|0---------------            |1------|2---|
|3------|4------|5------|6------          |
(and there are some extra hidden bands)

Instead in wine the result is:

|0---------------                       |||
|3------                               ||||

with the remaining bands having 0-size child windows (and there's
important functionality in those bands).  The minimum sizes
provided by the application seem to be right.  The application then
requests maximisation of 0,3,4,5,6 in that order, which isn't
implemented in WINE so that should not do anything, but I'd still
expect the minimum sizes to be honoured.

The output of wine --debugmsg +rebar is at:

http://www.cse.unsw.edu.au/~matthewc/rebar.gz (62K)

I'm currently trying to understand the source, but if anyone is
familiar with the rebar and has any suggestions that would be
much appreciated.

Thanks,
Matt

[Matt Chapman]

OK, I managed to fix the problem by doing this:

Index: dlls/comctl32/rebar.c
===================================================================
RCS file: /home/wine/wine/dlls/comctl32/rebar.c,v
retrieving revision 1.83
diff -u -r1.83 rebar.c
--- dlls/comctl32/rebar.c	9 Feb 2004 22:07:43 -0000	1.83
+++ dlls/comctl32/rebar.c	21 Feb 2004 07:03:15 -0000
@@ -1497,10 +1497,7 @@

/* separator from previous band */
    cxsep = (cntonrow == 0) ? 0 : SEP_WIDTH;
-
-	/* Header: includes gripper, text, image */
-	cx = lpBand->cxHeader;
-	if (lpBand->fStyle & RBBS_FIXEDSIZE) cx = lpBand->lcx;
+	cx = lpBand->lcx;

if (infoPtr->dwStyle & CCS_VERT)
        dobreak = (y + cx + cxsep > adjcy);

However, I'm not sure I understand the original code (i.e. initial cx in
REBAR_Layout = header only for the !RBBS_FIXED_SIZE case), and hence
don't know whether this will break something else.  Can anyone comment?

Matt

On Sat, Feb 21, 2004 at 11:40:00AM +1100, Matt Chapman wrote:
> Hi,
>
> I'm having some rebar issues with a particular application (Canon
> File Viewer Utility).
>
> The rebar is meant to look something like:
>
> |0---------------            |1------|2---|
> |3------|4------|5------|6------          |
> (and there are some extra hidden bands)
>
> Instead in wine the result is:
>
> |0---------------                       |||
> |3------                               ||||
>
> with the remaining bands having 0-size child windows (and there's
> important functionality in those bands).  The minimum sizes
> provided by the application seem to be right.  The application then
> requests maximisation of 0,3,4,5,6 in that order, which isn't
> implemented in WINE so that should not do anything, but I'd still
> expect the minimum sizes to be honoured.
>
> The output of wine --debugmsg +rebar is at:
>
> http://www.cse.unsw.edu.au/~matthewc/rebar.gz (62K)
>
> I'm currently trying to understand the source, but if anyone is
> familiar with the rebar and has any suggestions that would be
> much appreciated.
>
> Thanks,
> Matt

[Mike Hearn]

On Sat, 21 Feb 2004 18:10:09 +1100, Matt Chapman wrote:
> However, I'm not sure I understand the original code (i.e. initial cx in
> REBAR_Layout = header only for the !RBBS_FIXED_SIZE case), and hence
> don't know whether this will break something else.  Can anyone comment?

Hi Matt,

Are you sure the bug isn't that RBBS_FIXED_SIZE should be set somewhere
and isn't (or that this field is used in more than just that case)?

There is a great little set of programs called Control Spy floating around
the net which acts as a workout for the common controls library, you might
want to get hold of that (I have a copy if you can't find it via google)
and see if the rebar always behaves correctly even with your fix.

Reading the code and MSDN some more to better understand the code would be
a good plan as well.

thanks -mike

[Matt Chapman]

On Sat, Feb 21, 2004 at 01:19:47PM +0000, Mike Hearn wrote:
> On Sat, 21 Feb 2004 18:10:09 +1100, Matt Chapman wrote:
> > However, I'm not sure I understand the original code (i.e. initial cx in
> > REBAR_Layout = header only for the !RBBS_FIXED_SIZE case), and hence
> > don't know whether this will break something else.  Can anyone comment?
>
> Hi Matt,
>
> Are you sure the bug isn't that RBBS_FIXED_SIZE should be set somewhere
> and isn't (or that this field is used in more than just that case)?

No, I'm sure that the minimum size should be respected regardless, I've
verified this with Control Spy on Windows (thanks for the reference).
> There is a great little set of programs called Control Spy floating around
> the net which acts as a workout for the common controls library, you might
> want to get hold of that (I have a copy if you can't find it via google)
> and see if the rebar always behaves correctly even with your fix.

I've done some testing with Control Spy and my patch seems to fix the above
case without breaking anything else that I can see.

(CCing to wine-patches)

Matt