[SCM] WebKit Debian packaging branch, debian/unstable, updated. debian/1.1.15-1-40151-g37bb677

[darin]

The following commit has been merged in the debian/unstable branch:
commit 07ba9e2588868e5bbd15d56a034d03e3c34a9172
Author: darin <darin at 268f45cc-cd09-0410-ab3c-d52691b4dbfc>
Date:   Thu Mar 20 00:45:49 2003 +0000

            Reviewed by John.
    
            - fixed 3202780 -- REGRESSION: progress bar, stop button, both get stuck in "loading" state (bartsoft.com)
    
            * WebView.subproj/WebDataSource.m: (-[WebDataSource isLoading]):
            Add back the "is this page complete" check, so that subresource loads don't make us
            think we're loading again, once the page is complete. On the other hand, frames
            must still be checked independent of the "is this page complete" flag to avoid
            reintroducing bug 3200611.
    
    
    git-svn-id: http://svn.webkit.org/repository/webkit/trunk@3867 268f45cc-cd09-0410-ab3c-d52691b4dbfc

diff --git a/WebKit/ChangeLog b/WebKit/ChangeLog
index da4593b..25a624e 100644
--- a/WebKit/ChangeLog
+++ b/WebKit/ChangeLog
@@ -1,10 +1,22 @@
+2003-03-19  Darin Adler  <darin at apple.com>
+
+        Reviewed by John.
+
+        - fixed 3202780 -- REGRESSION: progress bar, stop button, both get stuck in "loading" state (bartsoft.com)
+
+        * WebView.subproj/WebDataSource.m: (-[WebDataSource isLoading]):
+        Add back the "is this page complete" check, so that subresource loads don't make us
+        think we're loading again, once the page is complete. On the other hand, frames
+        must still be checked independent of the "is this page complete" flag to avoid
+        reintroducing bug 3200611.
+
 2003-03-19  Ed Voas  voas at apple.com
 
         Reviewed by Richard.
 
-		Got it working in non-compositing mode as well, so in theory it
-		can work inside a PowerPlant application as well. It's a bit of what
-		I'd consider a hack, but it's pretty straightforward.
+        Got it working in non-compositing mode as well, so in theory it
+        can work inside a PowerPlant application as well. It's a bit of what
+        I'd consider a hack, but it's pretty straightforward.
 
         * Carbon.subproj/CarbonWindowAdapter.m:
         (-[CarbonWindowAdapter setViewsNeedDisplay:]):
@@ -1391,8 +1403,8 @@
 
         Reviewed by Richard.
 
-		Don't use _HIViewSetNeedsDisplayInRect for now. Fixes
-		Jaguar builds.
+        Don't use _HIViewSetNeedsDisplayInRect for now. Fixes
+        Jaguar builds.
 
         * Carbon.subproj/HIViewAdapter.m:
         (-[HIViewAdapter setNeedsDisplay:]):
@@ -1415,9 +1427,9 @@
 
         Reviewed by Richard.
 
-		First cut of Carbon view support. It will need tweaking as we go,
-		but I think I finally have all the assertions taken care of, and
-		I also believe that I have the drawing glitches all sorted out now.
+        First cut of Carbon view support. It will need tweaking as we go,
+        but I think I finally have all the assertions taken care of, and
+        I also believe that I have the drawing glitches all sorted out now.
 
         * Carbon.subproj/CarbonUtils.h: Added.
         * Carbon.subproj/CarbonUtils.m: Added.
diff --git a/WebKit/WebView.subproj/WebDataSource.m b/WebKit/WebView.subproj/WebDataSource.m
index ea0f8b4..6408d8d 100644
--- a/WebKit/WebView.subproj/WebDataSource.m
+++ b/WebKit/WebView.subproj/WebDataSource.m
@@ -82,11 +82,15 @@
 // Returns YES if there are any pending loads.
 - (BOOL)isLoading
 {
-    if (!_private->primaryLoadComplete && _private->loading) {
-        return YES;
-    }
-    if ([_private->subresourceClients count]) {
-	return YES;
+    // Once a frame has loaded, we no longer need to consider subresources,
+    // but we still need to consider subframes.
+    if ([[self webFrame] _state] != WebFrameStateComplete) {
+        if (!_private->primaryLoadComplete && _private->loading) {
+            return YES;
+        }
+        if ([_private->subresourceClients count]) {
+            return YES;
+        }
     }
     
     // Put in the auto-release pool because it's common to call this from a run loop source,

-- 
WebKit Debian packaging