There is no way to close the Page Info dialog on the Macho-o build of Mozilla. The red close button (tht thing on the top left) is greted out (only minimize and maximize) and there is no ok button or any other buton to click. Source was up to date as of 01/27/2002 01:08. .mozconfig: ac_add_options --disable-tests ac_add_options --disable-debug ac_add_options --enable-optimize ac_add_options --with-glib-prefix=/sw ac_add_options --with-libIDL-prefix=/sw ac_add_options --disable-mailnews ac_add_options --enable-plaintext-editor-only mk_add_options MOZ_CVS_FLAGS="-q -z9 -d :pserver:anonymous@cvs-mirror.mozilla.or g:/cvsroot"

(that checkout time was http://bonsai.mozilla.org time)

->db48x

First, what's mach-o? Second what window manager greys out the close button? That's usually not optional.

Mach-o is one of the ways to build Mozilla on Mac OSX, using the unix build system (gmake, and co) but using some Mac specific stuff. In the end you end up with a native Mac application, so It's Mac OSX's native window manager. Note. this works fine on the CFM build 2002012503 (For more information on mach-o, you can look at http://www.mozilla.org/ports/fizzilla/ but I'm nto sure how helpful that would be. The greying out of the close button is specific to some dialogs ccing Zach.

I thought it was a mac thing. I'll need a mac person to take a look at it.

related: bug 81657

qa for mach-o bugs belongs to me at this time because most qa contacts do not have access to mach-o builds.

qa back to me

So... we should consider adding "close" to the option list we pass to openDialog when opening pageinfo

Confirming. I just saw this in a CVS build built on Mar 11.

any traction here? Can someone with a mach-o tree try what I suggest in comment 9? No build environment required, even; it's a JS-only change....

*** Bug 172463 has been marked as a duplicate of this bug. ***

Comment on attachment 103926 [details] [diff] [review] the patch.... someone needs to test this r=cls pending ongoing discussion of possible deeper widget issue.

cls says that fixes the bug. Taking, with your permission, Daniel. reviews? pinkerton? bryner? As bryner suspected, uninitialized vars were the problem.

Comment on attachment 103926 [details] [diff] [review] the patch.... someone needs to test this this patch is bogus.

Comment on attachment 103944 [details] [diff] [review] More like this, actually sr=bryner. good catch.

Comment on attachment 103944 [details] [diff] [review] More like this, actually r=cls

looks good. r=pink

Comment on attachment 103944 [details] [diff] [review] More like this, actually a=blizzard on behalf of drivers for 1.2final.

fixed on trunk.