Closed Bug 43908 Opened 24 years ago Closed 24 years ago

Need to check where drag initiates in bookmarks tree

Categories

(SeaMonkey :: Bookmarks & History, defect, P2)

x86
Windows 98

Tracking

(Not tracked)

VERIFIED FIXED

People

(Reporter: bugzilla, Assigned: slamm)

Details

(Whiteboard: [rjc][nsbeta3+])

Attachments

(1 file)

Build ID: 2000062608

In the bookmarks tab of the sidebar, you can click plain white space (without 
selecting an item) and drag and a drag session will be initiated.  Similarly, 
you can select an item and then just starting from pretty much anywhere else in 
the tree and the drag still starts.  The same holds true for the tree in the 
Manage Bookmarks window.  We need to ensure that the drag is actually 
initiating from the item itself that's being dragged. I don't know if this is 
specific to bookmarks, or rather the general tree dragging code that seems to 
be reused in various places.
FYI, here's hyatt's comments (see bug 43367 and bug 43134):
  "These are bugs in the event handlers of individual trees. ... It was
  incorrect of me to hide clicks on the scrollbar from the DOM.  I fixed
  this when I moved to the new tree.  I'm not doing hiding scrollbar
  events any more, and now people have to patch their tree event handlers
  to check to see if a click actually occurred on a cell. Clicking on the
  blank space in a tree would have exhibited the same problem."
Yep, I didn't see that. Thanks John.

RJC: since slamm's on sabbatical, would this be easy for you to fix?  If not, 
feel free to remove yourself from whiteboard/cc -- it's not critical at all.
Severity: normal → minor
Whiteboard: [rjc]
over to rjc
Assignee: slamm → rjc
Keep this on slamm's list... with [rjc] in the status area of the bug.
Assignee: rjc → slamm
this was fixed.
Status: NEW → RESOLVED
Closed: 24 years ago
Resolution: --- → FIXED
Actually, this throws an error, which does abort the drag, but not in the best 
way.

  JavaScript error: chrome://communicator/content/bookmarks/bookmarksDD.js line 
80: 
    event.target.parentNode.parentNode.parentNode.parentNode has no properties

I think that this is the right fix

Index: mozilla/xpfe/components/bookmarks/resources/bookmarksDD.js
===================================================================
RCS file: /cvsroot/mozilla/xpfe/components/bookmarks/resources/bookmarksDD.js,v
retrieving revision 1.12
diff -u -r1.12 bookmarksDD.js
--- mozilla/xpfe/components/bookmarks/resources/bookmarksDD.js  2000/08/24 
00:17:28     1.12
+++ mozilla/xpfe/components/bookmarks/resources/bookmarksDD.js  2000/09/11 
03:09:05
@@ -42,7 +42,7 @@
   // if the click is on the tree proper, ignore it. We only care about clicks 
on items.
 
   var tree = document.getElementById("bookmarksTree");
-  if ( event.target == tree )
+  if ( event.target == tree || event.target.localName == "treechildren" )
     return(true);         // continue propagating the event
     
   var childWithDatabase = tree;


Does that look correct, Steve "5 days and counting" Lamm.
Status: RESOLVED → REOPENED
Keywords: patch
Resolution: FIXED → ---
thanks for digging deeper.  it looks good to me (r=blake).  I don't know how to 
check this in, though...I didn't write the patch, it's from a netscape 
employee, but the bug isn't nsbeta3+.  Brendan, can I check this in with your 
approval?
Keywords: approval
Pending r=slamm, nominating nsbeta3 to cover the Netscape side of procedures.
There is a patch, which is very straightforward and (if judged correct) has 
minimal risk, snuffing out an obvious error.

(Heh. Perhaps I should just drop hints in the future ...).
Keywords: nsbeta3
Do the right thing: check the patch in.

/be
nav triage team:
OK, wise people have a patch and say the patch is good.  Check it in this week 
or forever hold your peace.  +, P2
Priority: P3 → P2
Whiteboard: [rjc] → [rjc][nsbeta3+]
fix checked in, but I think ray's progid changes broke bookmarks dragging (so 
may be hard to verify)
Status: REOPENED → RESOLVED
Closed: 24 years ago24 years ago
Resolution: --- → FIXED
okay, ray's second batch of progid->contractID checkins fixed dragging, and 
this is now working properly (dragging only initiates in the proper place, and 
no more console errors).

vrfy fixed in new win98 build
Status: RESOLVED → VERIFIED
Product: Browser → Seamonkey
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: