Re: AppIcon patch revised...

[Sungjin Chun]

Hi,

Thanks for your interest.

In XGContextEvent.m, the ReparentNotify event
is processed as you said.

----- Original Message ----- 
From: "Dan Pascu" <dan@services.iiruc.ro>
To: "Sungjin Chun" <ninja@linuxone.co.kr>
Cc: <discuss-gnustep@gnu.org>
Sent: Wednesday, February 07, 2001 8:21 AM
Subject: Re: AppIcon patch revised...


> 
> Be sure not to forget that piece of code (it didn't seem to be in this
> patch, but it may be in another place already):
> 
>         case ReparentNotify:
>             window->parent = event.xreparent.parent;
>             break;
> 
> so that you will be able to relay the event to window maker if
> necessary.
> 
> The only code I saw in the patch is this (I don't know if it is related
> to what I mean or not):
> 
>   window->parent = root->ident;
> 
> but if its related, the parent window should not statically assigned,
> but taken from the ReparentNotify event sent to you by window maker when
> the appicon is created inside wmaker. Thus you need a code similar to
> what I depicted at the top of this email. Else window maker will not
> receive the event you relay to it, but instead one of your own windows
> (root->ident) will get it, and window maker will never know that some
> mouse action took place on that icon and will not drag it.
> 
> Again, I may be wrong, and the root->ident may be a result of a
> ReparentNotify assignment somewhere else in the already existing code.
> 
> 
> On  6 Feb, Sungjin Chun wrote:
> > Hi, 
> > Again I modified my patch. Previous one has the 
> > problem; Appicon sometimes does not correctly
> > displayed. This is fixed in this newer one.
> > And this time I've separated gui part and xgps
> > part and real modification from previous one is
> > in the xgps part. Please look into the patch and
> > any comments are welcome as always.
> > 
> > Thanks in advance.
> 
> -- 
> Dan
> 
>