Closed Bug 257036 Opened 20 years ago Closed 20 years ago

window.open sets wrong vertical size, depending on event-source

Categories

(Core Graveyard :: GFX: Win32, defect)

x86
Windows 2000
defect
Not set
major

Tracking

(Not tracked)

RESOLVED WORKSFORME

People

(Reporter: ken2006, Unassigned)

References

()

Details

Attachments

(4 files)

User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.8a3) Gecko/20040824 Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.8a3) Gecko/20040824 The example URL contains 3 pop-up event methods: 1) in-page (inline) executed 2) A-tag inline onclick invocation 3) A-tag onclick function call All of them call excatl;y the same window.open command, however, only #2 generates the correct verical dimension of the new window (for me anyway). The others generate a size that is too large (roughly 2x). Reproducible: Always Steps to Reproduce: 1.Load given URL 2.Click on sample links. this only started sometime into the 1.8 builds. 1.7 is okay. I'm rating this major for two reasons: app-programmers (me) depend on correctly sized windows, and, I feel it may point to a lower level problem (that may have other un-anticipated / yet-unknown event/DOM consequences). ps - this is truly a bizarre problem :-)
Changed sample to add two variants of function call: onClick="testFunc" onClick="testFunc()" Example two *does* work, but one does not (behaves same as page-inline sample).
Flags: blocking1.8a4+
worksforme with linux trunk 2004091107 only drivers can set blocking flags(+). you can request(?) them.
Flags: blocking1.8a4+
Not sure if this is the same or not, but with Mozilla/5.0 (Windows; U; Windows NT 5.0; rv:1.7.3) Gecko/20040913 Firefox/0.10 on http://demo:guide@nddemo.peregrine.com/nm/ the toolbar is taller by about the height of the status bar as compared to previous versions (eg Firefox 0.8.x, 0.9.x, etc...) You will notice a white band at the bottom of the window that should not be there.
This may be related to changes made for bug#22183...
Worksforme on Linux as well. Over to Win32 GFX.
Assignee: general → win32
Component: DOM: Level 0 → GFX: Win32
Hmm, the behavior on this changed slightly after I deleted panels.rdf and restarted... now, only link number two on the demo URL opens with the wrong size.. the body-inline script (runs on page load) works fine, so does link number 1 and three... This bug may be related to bug #260098, who behavior also changed after I deleted panels.rdf.
I still see the issue at http://demo:guide@nddemo.peregrine.com with Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.7.3) Gecko/20041026 Firefox/1.0RC1 The code to open the window is: var toolbar_height = 100; var toolbar_width = 400; var tool_x = (screen.width - toolbar_width) / 2; var tool_y = 15; var attr = "width=" + toolbar_width + ",height=" + toolbar_height + ",resizeable=no" + ",left=" + tool_x + ",top=" + tool_y; self.ktoolbar = window.open("", ktoolbarName, attr); If I remove 20 from the toolbar_height, it comes out at the right height. Or is the status bar supposed to fill in that space? This has worked properly for years. Not sure why it is breaking now.
Flags: blocking-aviary1.0?
I'm unable reproduce this problem on windows XP on today's build. Is this win2K only?
I will try today's build in Win2k.
Today it is worse than my last update. Using: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.8a5) Gecko/20041105 Firefox/0.9.1+ I am attaching a screenshot.
unable to reproduce with the latest windows builds. If you can reproduce, please provide a testcase that fails with the latest branch build. thanks.
Status: UNCONFIRMED → RESOLVED
Closed: 20 years ago
Flags: blocking-aviary1.0? → blocking-aviary1.0-
Resolution: --- → WORKSFORME
Matthew, try deleting your panels.rdf file with 'er shut down then restart Moz - does that work?.... You could also try deleteing compreg.dat and XUL.MFL - these files seem exteremly vulnerable to corruption during a crash and cause all kinds of ... problems.
I deleted everything in C:\Documents and Settings\mdarwin\Application Data\Mozilla before I ran the last test case. Re: Comment #11 - Please see http://demo:guide@nddemo.peregrine.com/nm/ - Unblock popup windows - Then press reload - Screen shot from Comment #10 is produced
This is ok with IE6 and NS7.2, just with Firefox 1.0, there is an extra white bar at the bottom.
I have simplified this test case as much as possible and put it at http://www.mdarwin.ca/test/toolbar.html <HTML> <SCRIPT LANGUAGE="Javascript"> window.open("", "foo", "width=400,height=100"); </SCRIPT> <TITLE> Window opening test </TITLE> <BODY> Enable pop-ups to see this test case. </BODY> </HTML> In Mozilla 1.7.3 it opens the correct height. In Firefox 1.0 it opens ~25 pixels too big.
The initial comment in this bug sounds like bug 262411, which is still open and for good reason. Subsequent comments seem unrelated -- an observation that Mozilla has a minimum window size? Please don't take this comment as an invitation to similarly confuse the issue in 262411. I'm merely trying to reassure Ken about his closed bug. Matthew: are you referring to the fact that Mozilla has a minimum window size? It always has. If it's a bit larger in Firefox than in Moz 1.7.3, it doesn't seem like cause for alarm. Try the attached testcase. With Firefox 1.0 I get results like these: spec: 85x400, actual inner: 150x400, outer: 154x406 spec: 95x400, actual inner: 150x400, outer: 154x406 spec: 105x400, actual inner: 150x400, outer: 159x406 spec: 115x400, actual inner: 150x400, outer: 169x406 spec: 125x400, actual inner: 150x400, outer: 179x406 spec: 135x400, actual inner: 150x400, outer: 189x406 spec: 145x400, actual inner: 150x400, outer: 199x406 spec: 155x400, actual inner: 155x400, outer: 209x406 spec: 165x400, actual inner: 165x400, outer: 219x406 spec: 175x400, actual inner: 175x400, outer: 229x406 As you can see the overall window height is pinned to at least 100 pixels, and the content window to 150. A little funny stuff happens at sizes in between. As it's always been. Differences between Mozilla and Firefox could be related to chrome. You'll certainly get different numbers if you have adjusted, for example, your dom.disable_window_open_feature.* prefs.
Ever since Netscape 3.x, the minimum visible window hight has always been ~100 pixels. The code I wrote 5 years ago worked just fine until Firefox .10. I want to create a window with the absolute smallest vertical height. It makes no sense that the minimum window height is 20 pixels larger now, according to your test case: Firefox 1.0: spec: 85x400, actual inner: 150x400, outer: 122x400 spec: 95x400, actual inner: 150x400, outer: 122x400 spec: 105x400, actual inner: 150x400, outer: 127x400 spec: 115x400, actual inner: 150x400, outer: 137x400 spec: 125x400, actual inner: 150x400, outer: 147x400 spec: 135x400, actual inner: 150x400, outer: 157x400 spec: 145x400, actual inner: 150x400, outer: 167x400 spec: 155x400, actual inner: 155x400, outer: 177x400 spec: 165x400, actual inner: 165x400, outer: 187x400 spec: 175x400, actual inner: 175x400, outer: 197x400 Mozilla 1.7.3: spec: 85x400, actual inner: 150x400, outer: 102x400 spec: 95x400, actual inner: 150x400, outer: 102x400 spec: 105x400, actual inner: 150x400, outer: 107x400 spec: 115x400, actual inner: 150x400, outer: 117x400 spec: 125x400, actual inner: 150x400, outer: 127x400 spec: 135x400, actual inner: 150x400, outer: 137x400 spec: 145x400, actual inner: 150x400, outer: 147x400 spec: 155x400, actual inner: 155x400, outer: 157x400 spec: 165x400, actual inner: 165x400, outer: 167x400 spec: 175x400, actual inner: 175x400, outer: 177x400
Just for curiosity, I tried running your test case on IE and Netscape Communicator 3.04 Neither loads test test case javascript properly, so I've attached as screenshot instead. This is under Windows 98.
here are the results under Mozilla 1.8a5 Windows 2000: spec: 85x400, actual inner: 100x400, outer: 128x408 spec: 95x400, actual inner: 100x400, outer: 128x408 spec: 105x400, actual inner: 105x400, outer: 133x408 spec: 115x400, actual inner: 115x400, outer: 143x408 spec: 125x400, actual inner: 125x400, outer: 153x408 spec: 135x400, actual inner: 135x400, outer: 163x408 spec: 145x400, actual inner: 145x400, outer: 173x408 spec: 155x400, actual inner: 155x400, outer: 183x408 spec: 165x400, actual inner: 165x400, outer: 193x408 spec: 175x400, actual inner: 175x400, outer: 203x408
Finally I've reached the bottom of this. Yes there is a problem with short windows in Firefox 1.0. Because of bug 259192, a window opened given an inner height (as in 'height=100') will be padded by the height of a statusbar (22 pixels in my case). A very short window will not display the statusbar (bug 255388) though space is still allocated for it. Rumour has it this has all been fixed on the trunk in bug 217477, but not in Firefox 1.0. Alrighty then. *This* bug, 257036, still lives on as bug 262411: something very odd is happening with window sizes under certain conditions. Comments in this bug from comment 7 onward are actually about that other problem, and to those three bugs I scatter thee.
Product: Core → Core Graveyard
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: