New Wahoo 12.00 build with rendering and Speed Dial improvements
By Ruarí Ødegaardruario. Friday, September 2, 2011 1:07:16 PM
Some core highlights include implementation of fully CSS 2.1 compliant display:list-item
Regarding that last point, if the fix works as expected, Opera should not go past the cache size. And, after a crash, it should restore to the correct size. If you can check the cache size after a crash, and also after correctly closing Opera following a crash, that would be much appreciated.
On the Desktop side we have made the animations much nicer when you move Speed Dials around.
Of course there are plenty of other great fixes, so scroll down and have a look!

WARNING: This is a development snapshot: It contains the latest changes, but may also have severe known issues, including crashes, and data loss situations. In fact, it may not work at all.
Download
Changelog
Core
- CORE-39767 (CSS 2.1 compliant display:list-item implementation)
CORE-39302 (The cache can grow past its limit)- CORE-32820 (SVG optimizations)
- CORE-12447 (overflow: hidden ignored for table-layout: fixed)
- CORE-17961 (Missing URL in redirection message)
- CORE-23382 (NFD é rendered incorrectly, and worse in 2.4 than 2.2)
- CORE-25431 (offsetLeft & offsetTop values are 0,0 for blank anchor tags)
- CORE-25611 (computedStyle returns an empty string for text-shadow property)
- CORE-25902 (border-radius doesn't clip replaced content)
- CORE-26985 (Cursor doesn't change to pointer when hovering links until the page has completely loaded)
- CORE-27522 (border-radius with different border widths)
- CORE-28875 (border on inline-table disappears if it has overflow: hidden)
- CORE-28949 (:target selector applies even after hash removed)
- CORE-30401 (Initialization of SVG DOM objects different from what the SVG 1.1 second edition mandates)
- CORE-31539 (overflow: hidden ignored on table if nested descendant width exceeds first cell descendant width)
- CORE-34305 (svgz files saved as empty)
- CORE-35190 (CSS: border-radius ignored on fieldsets that have a legend)
- CORE-35547 (Content of http://www.good.is not displayed, @font-face with floating point number and comma in local())
- CORE-36266 (Can't post link in orkut message)
- CORE-36816 (DOM updates during composition causes doubled characters and oddities): unable to use ime on qzone.qq.com
- CORE-37191 (computedStyle returns an empty string for box-shadow property)
- CORE-37325 (rx, ry calculations for rect are incorrect according to svg spec)
- CORE-37340 (Improve GetObjectProperties with regards to complicated properties)
- CORE-37422 (Recognizing meeting invites .ics)
- CORE-37452 (When JavaScript moves focus from input field, it gets empty (elm.focus() will abort ongoing IME))
- CORE-38132 (0 returned for XHR.status when readyState == 2 (HEADERS_RECEIVED))
- CORE-38210 (Invalid Link header causes Internal communication error)
- CORE-38229 (opera:cache is not well laid out)
- CORE-38771 (Unnecessary bottleneck in certain cases with "overflow: hidden" and CSS transforms)
- CORE-39072 (Orkut profile picture change still not working)
- CORE-39098 (Make Math.log() faster)
- CORE-39112 (Border and overflow affects offset of events)
- CORE-39380 (Opera doesn't fire events for empty scripts)
- CORE-39391 (IME position not updated immidiately when text input is changed on getting focus)
- CORE-39456 (Overflow container with anchor inside table introduces unwanted scrollbars)
- CORE-39485 (Back button does not load image in reddit pages)
- CORE-39490 (User initiated click event not handled on file input)
- CORE-39516 (Drag-to-pan in image viewer only for left mouse button with no modifier keys)
- CORE-39525 (Set window.event when events occur even if it's rewritten by page scripts): Attachment opening fails in new Hotmail
- CORE-39547 (DATA_CLONE_ERR uses wrong code (24 instead of 25))
- CORE-39559 (No data stream sent to invisible windowless plugins): Lytro picture page with focus demos doesn't work
- CORE-39653 (Support event.isTrusted)
- CORE-39654 (Native validation messages pops up everywhere (on top of other tabs))
- CORE-39660 (CSS attribute selections and removeAttribute() regressed - removeAttribute() no longer triggers reevaluation of selectors?)
- CORE-39722 (JSON.stringify number with single digit string produced not equal to equiv. string literal)
- CORE-39755 (Radio button group with attribute "form" doesn't work properly)
- CORE-39803 (Crash with VOIPlay plugin)
- CORE-39852 (Minimum font size applied on computed style in svg while not rendered)
- CORE-39862 (Canvas 2D sometimes draws shadows even when it shouldn't)
- CORE-39871 (Remove HTMLHeadElement.profile): http://www.whatwg.org/specs/web-apps/current-work/complete/obsolete.html#dom-head-profile
- CORE-39902 (Array.prototype.slice() doesn't propagate exceptions)
- CORE-39912 (Extra suspend event fired for some preload=metadata tests that use ogg videos)
- CORE-39917 (HTMLEmbedElement src and HTMLFormElement action IDL attributes reflected as strings, not URLs (i.e. is not resolved))
- CORE-39923 (Updated Google Maps: My Places pop-up can't be closed preventing use of Google Map Directions (something with EcmaScript eval))
- CORE-40033 (Empty string values not returned from web storage after being set)
- CORE-40040 (Response with content-encoding:gzip decompressed automatically without changing the .gz file extension)
- CORE-40047 (inserting options in SELECT fails with DOMNodeInserted listener active)
- CORE-40073 (Missing horizontal scrollbar when a multicol container contains more columns than specified)
- CORE-40126 (Clicking on Google+ tab (top left corner) results in page that is never finished loading (history.replaceState))
- CORE-40146 (Column-spanned element in RTL multicol is incorrectly positioned)
- CORE-40155 (The title of the extension that appears in the debugging context dropdown should be the extension name from config.xml)
- CORE-40189 (image.width/naturalWidth and image.height/naturalHeight returns 0 after switching to fullscreen if the image is only in DOM)
- CORE-40210 (Anchors added to the URL changed by pushState reloads page)
- CORE-40229 (List items are added outside of contenteditable element)
- CORE-40230 (List is deleted when pressing backspace in a contenteditable element within a list item.)
- CORE-40247 (Weird results of HTML select options)
- CORE-40281 (Images sometimes displayed via onload events)
- CORE-7544 (Nested inline relatively positioned elements with non-auto z-index are not rendered)
- CORE-7597 (XMLHttpRequest can not POST data that consists of plain literal numbers only.)
- CORE-9459 (Page reloaded when refreshed to fragment URI)
- CORE-39763 (Crash when reloading page with Dragonfly open)
- Updates to some translations
- Security improvements from 11.51
- Further crash and memory leak fixes
Desktop
- DSK-232055 (Unable to re-download file to same download location if an entry for the previous download still exists in transfers manager)
- DSK-343630 (Crash when exiting Opera after changing "Preferred languages for webpages")
- DSK-344289 (Using mouse wheel over windowed plugin can also cause the page to scroll)
Speed Dial
- DSK-329998 (Thumbnails are omitted while cycling Speed Dials with the Tab key)
- DSK-330661 (Add an animation when resizing window to change number of columns)
- DSK-337427 (Add an animation when undoing Speed Dial move/swap)
- DSK-344603 (Crash when dragging a Speed Dial while Edit menu is opened)
- DSK-344606 (Crash when deleting a Speed Dial set to Reload Every after dragging it from the original position)
- DSK-344912 (Hover effect remains on unhovered Speed Dials)
Windows
- DSK-344870 (Can't go fullscreen by double clicking the VLC plugin)
Linux/FreeBSD
- DSK-318433 (Excel file not attached if open in OpenOffice)



1 2 3 4 5 6 Next »
each # Friday, September 2, 2011 1:52:16 PM
Mağruf ÇolakoğluZAHEK # Friday, September 2, 2011 1:52:41 PM
alekksander # Friday, September 2, 2011 1:53:45 PM
Muh. Caesar Novaldyxzzz # Friday, September 2, 2011 1:53:46 PM
ep01 # Friday, September 2, 2011 1:55:30 PM
Christophchristoph142 # Friday, September 2, 2011 1:55:42 PM
EDIT: First impression: Damn, it's fast!
Stephen Turrellstve10 # Friday, September 2, 2011 1:55:51 PM
Ar1Pear1pe # Friday, September 2, 2011 2:01:01 PM
dengwenwei # Friday, September 2, 2011 2:01:32 PM
Weliton RaulinoNoTiLeW # Friday, September 2, 2011 2:01:44 PM
Pramod Ghugegeeneeyes # Friday, September 2, 2011 2:02:16 PM
Ruarí Ødegaardruario # Friday, September 2, 2011 2:05:53 PM
Originally posted by geeneeyes:
we have quite a bit still to do yet.David Tsunamydavidtsunamy # Friday, September 2, 2011 2:06:00 PM
Mac OS X Lion BUG: In fullscreen mode, the Opera window can move on click and drag tab-bar!
http://files.myopera.com/davidtsunamy/blog/Captura%20de%20pantalla%202011-09-02%20a%20la(s)%2016.13.41.png -
Spadar ShutSpShut # Friday, September 2, 2011 2:07:05 PM
nimu_channimuchan # Friday, September 2, 2011 2:11:11 PM
minho # Friday, September 2, 2011 2:13:59 PM
Testing.
earth01 # Friday, September 2, 2011 2:18:51 PM
Gonna test this build deeper now
Michael A. Puls IIburnout426 # Friday, September 2, 2011 2:19:14 PM
That's a weird-looking effect. Hope I can get used to that.
MartinMastermind1 # Friday, September 2, 2011 2:29:03 PM
At this rate I'm going to need my 8GB just to browse the web.
Surferz Worldsurferzworld # Friday, September 2, 2011 2:30:33 PM
MagicM # Friday, September 2, 2011 2:31:07 PM
Szőts Ákosszotsaki2 # Friday, September 2, 2011 2:31:30 PM
Francisco Herrerafherrerav # Friday, September 2, 2011 2:35:12 PM
User Agent:Opera/9.80 (Windows NT 5.1; U; Edition Next; en) Presto/2.9.186 Version/12.00
Build number:1033
Popups:Block unwanted
Java:Off
Plug-ins:On
JavaScript:On
Referrer logging:On
Proxy:Off
Encoding setting:
Display mode:Strict
Load images:All
Normal cookies:Accept
Third party cookies:Accept
EnricoEnrico89 # Friday, September 2, 2011 2:35:49 PM
Stephen Turrellstve10 # Friday, September 2, 2011 2:36:06 PM
http://files.myopera.com/stve10/albums/5142082/Opera%2012%20build1054%20Peacekeeper.JPG -
m910q # Friday, September 2, 2011 2:36:10 PM
earth01 # Friday, September 2, 2011 2:40:00 PM
Originally posted by fherrerav:
That's because the build you're using does have a bug with downloading files.Download this snapshot with another browser and you'll never have this bug again.
Francisco Herrerafherrerav # Friday, September 2, 2011 2:41:43 PM
Originally posted by earth01:
Rafael Novaski rnovaski # Friday, September 2, 2011 2:49:36 PM
http://www.inovacaotecnologica.com.br/noticias/noticia.php?artigo=laser-criar-chuva&id=010125110902
Win XP.
Anyone confirm?
mulB # Friday, September 2, 2011 2:50:55 PM
I still have sometimes horizontal lines in pictures in "Page zoom" mode (150%) ... is this a rendering bug, a bad installation problem or a config mistake?
Thanx
Crabman # Friday, September 2, 2011 2:51:33 PM
Originally posted by rnovaski:
Yeah, crash. Win7 64x
metude # Friday, September 2, 2011 2:54:28 PM
EricJH # Friday, September 2, 2011 2:55:56 PM
minho # Friday, September 2, 2011 2:56:32 PM
All ramdonly crashes.
win 7 x64
minho # Friday, September 2, 2011 2:59:21 PM
Originally posted by rnovaski:
Confirmed.
win 7 x64
Dandan7960 # Friday, September 2, 2011 3:08:24 PM
screen shot --> http://goo.gl/c8Ru5
(win 7)
Stilezy # Friday, September 2, 2011 3:19:26 PM
I'm using Memguard as requested to track down a hard-to-trace bug. It worked fine with 11.50 and 11.51, and with the last version of 12.00.1047. No system change of any kind, but Memguard now fails with latest snapshot 12.00.1054.
Opera asks whether to use Memguard (memguard.ini contains the correct snapshot build), on clicking YES gives a dialog:
"Couldn't initialize Memguard, function signature not found in Opera.dll!"
And then on clicking OK the message
"Opera failed to start because:
The operation completed successfully"
Confirmed as a regression - install 12.00.1047 over 12.00.1054 and change memguard.ini contents to 12.00.1047, it doesn't have this problem. Install 1054 back over this, and it does.
Ruarí Ødegaardruario # Friday, September 2, 2011 3:20:27 PM
Originally posted by surferzworld:
fixed, thanksRuarí Ødegaardruario # Friday, September 2, 2011 3:23:35 PM
Originally posted by Enrico89:
Numbers would be meaningless I am afraid. One fix might take a developer 10 minutes and another fix might involve so many behind the scene changes it takes weeks.Blaz(ž) Pristavitalianjob44 # Friday, September 2, 2011 3:38:31 PM
Peacekeeper is faster about 400, 9222 vs 8851 (10.51), Chrome 9302. Data is getting faster is important thing, then we can beat Chrome, as the scores shows below.
http://www.shrani.si/f/s/128/1om9UzwV/bench.png
Wilbundel # Friday, September 2, 2011 3:40:13 PM
Click on all photos (Все фото).
Pull the tab for the picture in the lower boundary of the parent window. Opera freezes.
berend ytsmaytsmabeer # Friday, September 2, 2011 3:42:07 PM
GwenDragon # Friday, September 2, 2011 3:46:43 PM
Testcase: http://labs.gwendragon.de/bugs/opera/12B1054-rendering_fieldsets-legend.html
Is this a known bug or should i report this as a bug?
Crabman # Friday, September 2, 2011 3:50:00 PM
Hm, maybe related to this but now a site stops scrolling via mousewheel when the mouse hovers over plugin content like a YT-video after it is activated (once clicked on). eg http://www.youtube.com/watch?v=wt3ubzl7JWs&feature=feedu
This was a problem a while ago and then fixed, so looks like a regression.
Jordan TrillesCristallix # Friday, September 2, 2011 3:50:44 PM
abdur92 # Friday, September 2, 2011 3:54:09 PM
tisazalay # Friday, September 2, 2011 3:54:21 PM
Originally posted by mulB:
That is probably the longest living bug...But most probably the end is near with the rendering changes that are coming...
Stilezy # Friday, September 2, 2011 3:59:10 PM
Originally posted by Swapnil99pro:
This isn't a general discussion forum, but I have to say I prefer it this way. Opera 12 has a lot of changes, and the release/changelog of 11.51 makes it sound like they are more willing to update current releases, only issuing new releases when they are feature complete and debugged. Personally I am very glad if the slow growth of Opera 12 is a sign they are taking code quality and stability and bug fixing really seriously. That's worth the wait.
BinbiK-Gergo # Friday, September 2, 2011 4:03:09 PM
Originally posted by ruario:
Hy!!!
Opera 12 final Hw and Webgl????
Jordan TrillesCristallix # Friday, September 2, 2011 4:04:34 PM