17138 2008-02-01 09:40:58 -0800 REGRESSION: Node highlight not updated properly 2008-02-07 15:37:10 -0800 1 1 1 Unclassified WebKit Web Inspector (Deprecated) 528+ (Nightly build) PC Windows XP RESOLVED FIXED data:text/html,<div style="width: 500px; height: 50px; background: red;">Top</div><div style="position: absolute; top: 1500px; width: 500px; height: 50px; background: green;">Bottom</div> InRadar, PlatformOnly, Regression P1 Normal --- 1 dev+webkit aroben aroben oldest_to_newest 69059 0 dev+webkit 2008-02-01 09:40:58 -0800 After Adam's reworking of the node highlight code yesterday, the highlight no longer updates on Windows when switching dom nodes until the inspected window is focused. Also, while docked, the node highlights can disappear completely at times. For example: 1. Load the testcase in the URL and inspect the Top div - the highlight is correct. 2. Switch to the Bottom div - the hole for the Top div is gone and the entire webview is faded grey. 3. Click in the inspected window - the hole for the Bottom div appears. 4. Dock the inspector. 5. Walk the DOM from <html> down and notice the different highlight behaviors, especially how toggling back and forth between the Top and Bottom divs never causes a hole to appear. 69063 1 dev+webkit 2008-02-01 09:53:52 -0800 (In reply to comment #0) > After Adam's reworking of the node highlight code yesterday [snip] For future reference, that would be the following check ins: <http://trac.webkit.org/projects/webkit/changeset/29886> <http://trac.webkit.org/projects/webkit/changeset/29887> <http://trac.webkit.org/projects/webkit/changeset/29888> 69069 2 mrowe 2008-02-01 10:47:55 -0800 <rdar://problem/5719869> 69692 3 aroben 2008-02-07 11:43:38 -0800 I think the problem is that we translate the context before scrolling the element into view, which could change how much we need to translate by. 69698 4 18988 aroben 2008-02-07 12:10:14 -0800 Created attachment 18988 patch with changelog 69700 5 18988 darin 2008-02-07 12:20:30 -0800 Comment on attachment 18988 patch with changelog r=me + context.translate(-overlayRect.x(), -overlayRect.y()); Should we use save/restore instead? 69701 6 aroben 2008-02-07 12:22:37 -0800 (In reply to comment #5) > (From update of attachment 18988 [edit]) > + context.translate(-overlayRect.x(), -overlayRect.y()); > > Should we use save/restore instead? I think it would be "in addition to" rather than "instead". I think we could save/restore in case the InspectorClient wanted to do more drawing of its own, but since the hope is that all drawing will go through this method I don't think it's really necessary. 69714 7 aroben 2008-02-07 15:37:10 -0800 Fixed in r30081 18988 2008-02-07 12:10:14 -0800 2008-02-07 12:20:30 -0800 patch with changelog 0001-Fix-Bug-17138-REGRESSION-Node-highlight-no.patch text/plain 3270 aroben RnJvbSAyM2VkZGEzMjdjYzFhMzZhN2QwZjAyOTkwMmY3ZWEzZmVkOTk0NmJiIE1vbiBTZXAgMTcg MDA6MDA6MDAgMjAwMQpGcm9tOiBBZGFtIFJvYmVuIDxhcm9iZW5AYXBwbGUuY29tPgpEYXRlOiBU aHUsIDcgRmViIDIwMDggMTU6MDk6MzggLTA1MDAKU3ViamVjdDogW1BGUl0gICAgICAgICBGaXgg QnVnIDE3MTM4OiBSRUdSRVNTSU9OOiBOb2RlIGhpZ2hsaWdodCBub3QgdXBkYXRlZCBwcm9wZXJs eQoKICAgICAgICA8aHR0cDovL2J1Z3Mud2Via2l0Lm9yZy9zaG93X2J1Zy5jZ2k/aWQ9MTcxMzg+ CiAgICAgICAgPHJkYXI6Ly9wcm9ibGVtLzU3MTk4Njk+CgogICAgICAgIFJldmlld2VkIGJ5IE5P Qk9EWSAoT09QUyEpLgoKICAgICAgICBObyB0ZXN0IHBvc3NpYmxlLgoKICAgICAgICAqIHBhZ2Uv SW5zcGVjdG9yQ29udHJvbGxlci5jcHA6CiAgICAgICAgKFdlYkNvcmU6Okluc3BlY3RvckNvbnRy b2xsZXI6OmRyYXdOb2RlSGlnaGxpZ2h0KTogVXBkYXRlIHRoZQogICAgICAgIG92ZXJsYXlSZWN0 IGFmdGVyIHNjcm9sbGluZyB0byBtYWtlIHN1cmUgdGhhdCB3ZSB0cmFuc2xhdGUgdGhlIGNvbnRl eHQKICAgICAgICBieSB0aGUgY29ycmVjdCBhbW91bnQuCi0tLQogV2ViQ29yZS9DaGFuZ2VMb2cg ICAgICAgICAgICAgICAgICAgIHwgICAxNiArKysrKysrKysrKysrKysrCiBXZWJDb3JlL3BhZ2Uv SW5zcGVjdG9yQ29udHJvbGxlci5jcHAgfCAgIDEwICsrKysrKy0tLS0KIDIgZmlsZXMgY2hhbmdl ZCwgMjIgaW5zZXJ0aW9ucygrKSwgNCBkZWxldGlvbnMoLSkKCmRpZmYgLS1naXQgYS9XZWJDb3Jl L0NoYW5nZUxvZyBiL1dlYkNvcmUvQ2hhbmdlTG9nCmluZGV4IDY4ZGZmYTMuLmM5MDU0YzIgMTAw NjQ0Ci0tLSBhL1dlYkNvcmUvQ2hhbmdlTG9nCisrKyBiL1dlYkNvcmUvQ2hhbmdlTG9nCkBAIC0x LDUgKzEsMjEgQEAKIDIwMDgtMDItMDcgIEFkYW0gUm9iZW4gIDxhcm9iZW5AYXBwbGUuY29tPgog CisgICAgICAgIEZpeCBCdWcgMTcxMzg6IFJFR1JFU1NJT046IE5vZGUgaGlnaGxpZ2h0IG5vdCB1 cGRhdGVkIHByb3Blcmx5CisKKyAgICAgICAgPGh0dHA6Ly9idWdzLndlYmtpdC5vcmcvc2hvd19i dWcuY2dpP2lkPTE3MTM4PgorICAgICAgICA8cmRhcjovL3Byb2JsZW0vNTcxOTg2OT4KKworICAg ICAgICBSZXZpZXdlZCBieSBOT0JPRFkgKE9PUFMhKS4KKworICAgICAgICBObyB0ZXN0IHBvc3Np YmxlLgorCisgICAgICAgICogcGFnZS9JbnNwZWN0b3JDb250cm9sbGVyLmNwcDoKKyAgICAgICAg KFdlYkNvcmU6Okluc3BlY3RvckNvbnRyb2xsZXI6OmRyYXdOb2RlSGlnaGxpZ2h0KTogVXBkYXRl IHRoZQorICAgICAgICBvdmVybGF5UmVjdCBhZnRlciBzY3JvbGxpbmcgdG8gbWFrZSBzdXJlIHRo YXQgd2UgdHJhbnNsYXRlIHRoZSBjb250ZXh0CisgICAgICAgIGJ5IHRoZSBjb3JyZWN0IGFtb3Vu dC4KKworMjAwOC0wMi0wNyAgQWRhbSBSb2JlbiAgPGFyb2JlbkBhcHBsZS5jb20+CisKICAgICAg ICAgU29tZSBjbGVhbnVwIG9mIE1hYy1vbmx5IHVzZXIgc3R5bGVzaGVldCBjb2RlCiAKICAgICAg ICAgSSBtb3ZlZCBVc2VyU3R5bGVTaGVldExvYWRlciBvdXQgb2YgRnJhbWUuY3BwIGludG8gaXRz IG93biBmaWxlcywgYW5kCmRpZmYgLS1naXQgYS9XZWJDb3JlL3BhZ2UvSW5zcGVjdG9yQ29udHJv bGxlci5jcHAgYi9XZWJDb3JlL3BhZ2UvSW5zcGVjdG9yQ29udHJvbGxlci5jcHAKaW5kZXggYjMy NGUyYi4uYzYyYjJjMyAxMDA2NDQKLS0tIGEvV2ViQ29yZS9wYWdlL0luc3BlY3RvckNvbnRyb2xs ZXIuY3BwCisrKyBiL1dlYkNvcmUvcGFnZS9JbnNwZWN0b3JDb250cm9sbGVyLmNwcApAQCAtMTU5 MCwxMCArMTU5MCw2IEBAIHZvaWQgSW5zcGVjdG9yQ29udHJvbGxlcjo6ZHJhd05vZGVIaWdobGln aHQoR3JhcGhpY3NDb250ZXh0JiBjb250ZXh0KSBjb25zdAogICAgIGlmICghbV9oaWdobGlnaHRl ZE5vZGUpCiAgICAgICAgIHJldHVybjsKIAotICAgIEZyYW1lVmlldyogdmlldyA9IG1faW5zcGVj dGVkUGFnZS0+bWFpbkZyYW1lKCktPnZpZXcoKTsKLSAgICBGbG9hdFJlY3Qgb3ZlcmxheVJlY3Qg PSBzdGF0aWNfY2FzdDxTY3JvbGxWaWV3Kj4odmlldyktPnZpc2libGVDb250ZW50UmVjdCgpOwot ICAgIGNvbnRleHQudHJhbnNsYXRlKC1vdmVybGF5UmVjdC54KCksIC1vdmVybGF5UmVjdC55KCkp OwotCiAgICAgUmVuZGVyT2JqZWN0KiByZW5kZXJlciA9IG1faGlnaGxpZ2h0ZWROb2RlLT5yZW5k ZXJlcigpOwogICAgIGlmICghcmVuZGVyZXIpCiAgICAgICAgIHJldHVybjsKQEAgLTE2MDUsNiAr MTYwMSw5IEBAIHZvaWQgSW5zcGVjdG9yQ29udHJvbGxlcjo6ZHJhd05vZGVIaWdobGlnaHQoR3Jh cGhpY3NDb250ZXh0JiBjb250ZXh0KSBjb25zdAogICAgIGlmIChyZWN0cy5pc0VtcHR5KCkpCiAg ICAgICAgIHJlY3RzLmFwcGVuZChub2RlUmVjdCk7CiAKKyAgICBGcmFtZVZpZXcqIHZpZXcgPSBt X2luc3BlY3RlZFBhZ2UtPm1haW5GcmFtZSgpLT52aWV3KCk7CisgICAgRmxvYXRSZWN0IG92ZXJs YXlSZWN0ID0gc3RhdGljX2Nhc3Q8U2Nyb2xsVmlldyo+KHZpZXcpLT52aXNpYmxlQ29udGVudFJl Y3QoKTsKKwogICAgIGlmICghb3ZlcmxheVJlY3QuY29udGFpbnMobm9kZVJlY3QpICYmICFub2Rl UmVjdC5jb250YWlucyhlbmNsb3NpbmdJbnRSZWN0KG92ZXJsYXlSZWN0KSkpIHsKICAgICAgICAg RWxlbWVudCogZWxlbWVudDsKICAgICAgICAgaWYgKG1faGlnaGxpZ2h0ZWROb2RlLT5pc0VsZW1l bnROb2RlKCkpCkBAIC0xNjEyLDggKzE2MTEsMTEgQEAgdm9pZCBJbnNwZWN0b3JDb250cm9sbGVy OjpkcmF3Tm9kZUhpZ2hsaWdodChHcmFwaGljc0NvbnRleHQmIGNvbnRleHQpIGNvbnN0CiAgICAg ICAgIGVsc2UKICAgICAgICAgICAgIGVsZW1lbnQgPSBzdGF0aWNfY2FzdDxFbGVtZW50Kj4obV9o aWdobGlnaHRlZE5vZGUtPnBhcmVudCgpKTsKICAgICAgICAgZWxlbWVudC0+c2Nyb2xsSW50b1Zp ZXdJZk5lZWRlZCgpOworICAgICAgICBvdmVybGF5UmVjdCA9IHN0YXRpY19jYXN0PFNjcm9sbFZp ZXcqPih2aWV3KS0+dmlzaWJsZUNvbnRlbnRSZWN0KCk7CiAgICAgfQogCisgICAgY29udGV4dC50 cmFuc2xhdGUoLW92ZXJsYXlSZWN0LngoKSwgLW92ZXJsYXlSZWN0LnkoKSk7CisKICAgICAvLyBE cmF3IHRyYW5zbHVjZW50IGdyYXkgZmlsbCwgb3V0IG9mIHdoaWNoIHdlIHdpbGwgY3V0IGhvbGVz LgogICAgIGNvbnRleHQuZmlsbFJlY3Qob3ZlcmxheVJlY3QsIG92ZXJsYXlGaWxsQ29sb3IpOwog Ci0tIAoxLjUuNC4xOC5nZDBiOAoK