Bug 126289 – REGRESSION (r158369): Some pop-up windows open at incorrect size (affects GMail, theblaze.com)

• Status: RESOLVED
• Resolution: FIXED
• Component: WebKit2

[Jim Oase]

For several weeks those spawned windows have been the wrong size, they have been full screen.   Safari works ... webkit, not so much
http://mf.sitescout.com/disp?pid=6AA0C4335D&rnd=[CACHE-BUSTING-ID-HERE]&sid=371

[Jim Oase]

http://nationalreview.com/corner/367435/obamas-half-brother-floored-about-presidents-lying-about-meeting-him-andrew-johnson

Wrong size window

[Alexey Proskuryakov]

I cannot reproduce this with WebKit nightly r161208. When I open <http://mf.sitescout.com/disp?pid=6AA0C4335D&rnd=[CACHE-BUSTING-ID-HERE]&sid=371>, and click on an article link, a page opens that is the same size as the current page.

[Jim Oase]

Created attachment 220309 [details]
The window size with build WebKit-SVN-r158090

Next will be a window size with latest nightly build.   Build 158090 produces the reduced size window.  Build r161259	47.2 MB	2014-01-03 16:52:21 GMT produces a full screen window... see next attachment

[Jim Oase]

Created attachment 220311 [details]
Window size with Build r161259	47.2 MB	2014-01-03 16:52:21 GMT

See previous report for comparison results.

I am using Software  OS X 10.9.1 (13B42)    
Model Name:	MacBook Pro
  Model Identifier:	MacBookPro8,3
  Processor Name:	Intel Core i7
  Processor Speed:	2.2 GHz
  Number of Processors:	1
  Total Number of Cores:	4
  L2 Cache (per Core):	256 KB
  L3 Cache:	6 MB
  Memory:	16 GB
  Boot ROM Version:	MBP81.0047.B27
  SMC Version (system):	1.70f6
  Serial Number (system):	C02FT1KKDF93
  Hardware UUID:	0112776F-3C1A-5407-B085-A7AD4996D69F
  Sudden Motion Sensor:
  State:	Enabled

[Jim Oase]

Created attachment 220313 [details]
Safari view of window with proper size

[Alexey Proskuryakov]

The page on the screenshot looks nothing like <http://mf.sitescout.com/disp?pid=6AA0C4335D&rnd=[CACHE-BUSTING-ID-HERE]&sid=371>.

This probably explains why I can't reproduce.

[Alexey Proskuryakov]

There is a similar issue with GMail, which is almost certainly the same.

<rdar://problem/15744126>

[Gavin Barraclough]

Created attachment 220798 [details]
WIP Gmail fix

[WebKit Commit Bot]

Attachment 220798 [details] did not pass style-queue:

Failed to run "['Tools/Scripts/check-webkit-style', '--diff-files', u'Source/WebKit2/UIProcess/WebPageProxy.cpp', u'Source/WebKit2/UIProcess/WebPageProxy.h', u'Source/WebKit2/WebProcess/WebPage/WebPage.cpp', u'Source/WebKit2/WebProcess/WebPage/WebPage.h', u'Source/WebKit2/WebProcess/WebProcess.cpp', '--commit-queue']" exit_code: 1
ERROR: Source/WebKit2/WebProcess/WebPage/WebPage.h:173:  The parameter name "parameters" adds no information, so it should be removed.  [readability/parameter_name] [5]
Total errors found: 1 in 5 files


If any of these errors are false positives, please file a bug against check-webkit-style.

[Gavin Barraclough]

WIP patch basically reverts https://bugs.webkit.org/show_bug.cgi?id=123557 , that change made the creation parameters retrieved synchronously be window.open be retained, and also passed by CreateWebPage.

This change reverts to the old behavior where the new (inconsistent) state is sent with CreateWebPage - but now, instead of just ignoring it we update the WebPage accordingly.

This is likely to cause some test failures; go go gadget EWS.

[Build Bot]

Comment on attachment 220798 [details]
WIP Gmail fix

Attachment 220798 [details] did not pass mac-wk2-ews (mac-wk2):
Output: http://webkit-queues.appspot.com/results/6680620302336000

New failing tests:
fast/dom/Window/mozilla-focus-blur.html

[Build Bot]

Created attachment 220799 [details]
Archive of layout-test-results from webkit-ews-15 for mac-mountainlion-wk2

The attached test failures were seen while running run-webkit-tests on the mac-wk2-ews.
Bot: webkit-ews-15  Port: mac-mountainlion-wk2  Platform: Mac OS X 10.8.5

[Gavin Barraclough]

Created attachment 220916 [details]
Fix

[Anders Carlsson]

Comment on attachment 220916 [details]
Fix

View in context: https://bugs.webkit.org/attachment.cgi?id=220916&action=review

> LayoutTests/platform/mac-wk2/fast/dom/Window/mozilla-focus-blur-expected.txt:8
> -PASS: The focus should not have been changed!
> -PASS: The focus should not have been changed!
> -PASS: The focus should not have been changed with URL=data:text/html,<script>opener.focus();opener.postMessage("", "*");</script>
> -PASS: The focus should not have been changed with URL=data:text/html,<script>blur();opener.postMessage("", "*");</script>
> +FAIL: The focus should not have been changed!
> +FAIL: The focus should not have been changed!
> +FAIL: The focus should not have been changed with URL=data:text/html,<script>opener.focus();opener.postMessage("", "*");</script>
> +FAIL: The focus should not have been changed with URL=data:text/html,<script>blur();opener.postMessage("", "*");</script>

I think it's better if we skip this test instead of landing new results.

[Gavin Barraclough]

Committed revision 162088.