Re: [WSG] 「Opera」 Percent with css
ピエール・アンリ・ラヴィン wrote: > > I don't understand the following issue with Opera: > > Let's set a container to 4000px, and children elements to 12.5% > 4000px / 12.5 = 320px. > But for Opera, > * 12.5px => 12px : I can understand > * 12.5% => 12% : I don't understand I can only say that this is an old issue with Opera, still present in beta versions of Opera 10: it simply ignores the fractional part in percentages used as width. In the past Opera had even worse "approximation" problems, see [1]. Now most of these are fixed, but percentage widths are still "stepped" [2], and borders expressed in em still have a crazy behavior [3]. Best regards, Bruno [1] http://brunildo.org/test/emarg.pl [2] http://brunildo.org/test/percwidth2.pl [3] http://brunildo.org/test/embord.pl -- Bruno Fassino http://www.brunildo.org/test *** List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm Help: memberh...@webstandardsgroup.org ***
Re: [WSG] encounter the first IE8 bug
On Fri, May 29, 2009 at 8:37 AM, tee wrote: > I see that Prototype has a version 1.6.1_rc2 that is compatible with IE8, I > tried using this version but the problem is still there. Yes. I created a minimal test case [1], and it seems that this is really an IE8 bug. In the test case a relative box contains two boxes: a red and a green one. The latter has position:absolute and a simple script (using the prototype function Element.clonePosition) should put it exactly over the red one. In IE8 the green box is displaced by the offsets of the relative box with respect to body. The problem is in the green box being "display:none" while the computations are done, and made visible only after (apparently, when display is 'none', in IE8 offsetParent is not the expected one...) If the box is made visible before the computation (like here [2]) the problem does not show up. Best regards, Bruno [1] http://brunildo.org/test/test/IE8_proto_clonePosition.html [2] http://brunildo.org/test/test/IE8_proto_clonePosition_2.html -- Bruno Fassino http://www.brunildo.org/test *** List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm Help: memberh...@webstandardsgroup.org ***
Re: [WSG] encounter the first IE8 bug
On 2009/5/29 tee wrote: > > Thank you so much!!! Never look into Prototype scripts, as a matter of fact, > I'd been working on getting rid of most Prototype scripts from Magento. One > more to get rid of. > > Will fix it for my projects and file a bug report to Varien instead of > Prototype community :-) > > Can you tell me exactly which line the code is in prototype.js? I tried look > into but couldn't find. Well, my "fix" is just a patch for this specific case, I would not really use it without further testing. It is in the viewportOffset and clonePosition functions of your prototype script, around line 2096. Also please note that I don't think you can fix the problem by setting left: 0!important in CSS because "search_autocomplete" needs a (non zero) left value. You could try something like #search_autocomplete { left: 730px! important; } assuming that widths in your layout won't change. Another thing that could help, since the script computed left value seems relative to the viewport instead of the first positioned ancestor, is to remove (reset to static) the position:relative on #container. But of course, since this "relative" is probably there for some good reasons, removing it may create other problems. What really should be done is to investigate why the prototype clonePosition does not seem to work correctly in IE8 (with absolute/relative elements). Best regards, Bruno -- Bruno Fassino http://www.brunildo.org/test *** List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm Help: memberh...@webstandardsgroup.org ***
Re: [WSG] encounter the first IE8 bug
On Tue, May 26, 2009 at 11:52 PM, tee wrote: > http://www.fengshuifabulous.com/jewelry.html > In the search field, if you enter "beads" it triggers the ajax > auto-complete, in IE8, the dropdown is off the search div and get pulled to > the right. As Сергей pointed out, the problem is in the scripts that add a left value for the absolutely positioned element with the suggestions. This left value is computed based on properties of other elements and the result is not correct for IE8. I cannot say if this is a bug or simply a different behavior. The computation is inside the prototype.js part of your script, that already has "corrections" for some browsers. Apparently IE8 needs a similar correction, but it is not there. If you have a look here: http://brunildo.org/test/test/jewelry.html I have copied your page and your scripts and added a correction line in the latter (correction at the moment fed to all browsers). This seems to correct the IE8 issue (and probably breaks someone else, but it should be easy to send it just to IE8). The best thing would probably to check this with the "prototype" community. Best regards, Bruno PS: In order to have the ajax work in the copied page no more on the original domain, I've used a "fixed" suggestion set. -- Bruno Fassino http://www.brunildo.org/test *** List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm Help: memberh...@webstandardsgroup.org ***
Re: [WSG] IEv8 support for Data URIs?
On Fri, Jan 23, 2009 at 11:55 AM, Foskett, Mike wrote: > I was under an impression that IEv8 was to support Data URI format for > images. > > Yet preliminary testing with IETester shows a lack of support. > > Can anyone confirm if IEv8 is to support the format? There are some limitations documented here [1]. In simple cases it seems to work, at least in IE8 build 18344 (see for example [2]). Bruno [1] http://code.msdn.microsoft.com/Release/ProjectReleases.aspx?ProjectName=ie8whitepapers&ReleaseId=575 [2] http://brunildo.org/test/datauri-from-file.pl?bg100A.gif+gif -- Bruno Fassino http://www.brunildo.org/test *** List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm Help: memberh...@webstandardsgroup.org ***
Re: [WSG] couple of bugs in Opera v9.5
On Sun, Sep 14, 2008 at 9:19 AM, tee wrote: > Thanks for asking this. No, it's not tabbing.What I meant is the whole page > is not selectable when I place the cursor anywhere in the content area. > Using 'Select All' also not working, and the link 'go here' is not > clickable. The only clickable areas in the pages are the three checkboxes > and radio buttons. > My system is Mac OS 10.5.4, Opera version 9.50 4870 build. Thanks for clarifying this. I cannot reproduce those problems. I'm using 9.52 build 4916 on OS X 10.5.4 (and 9.52 build 10108 on Windows XP). (and Dragonfly does not seem to interfere.) Bruno PS: > http://lotusseedsdesign.com/opera-test/ss1.html This is a 404 not found. -- Bruno Fassino http://www.brunildo.org/test *** List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm Help: [EMAIL PROTECTED] ***
RE: [WSG] couple of bugs in Opera v9.5
tee wrote: >>> >>> http://lotusseedsdesign.com/opera-test/opera.html >> > > And the unclickable, unselectable behavior in Opera is > driving me nut. > I still can't figure what is causing it. Anybody figure it? I'm not sure to understand what do you mean here, what exactly is unclickable/unselectable? (the "go there" link in that page works normally in my Opera 5.2). Do you mean "when tabbing with the keyboard" ? -- Bruno Fassino http://www.brunildo.org/test *** List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm Help: [EMAIL PROTECTED] ***
RE: [WSG] couple of bugs in Opera v9.5
tee wrote: > As mentioned in my Opera not playing nice with checkbox post, I have > collected a list of bugs that I stumbled doing just one > website. > > http://lotusseedsdesign.com/opera-test/opera.html The problem of the space at the right of your two columns (about 17px) is caused by a long-lived issue in Opera. With percentage width it simply ignores all decimals, so width: 49.9% is computed like width: 49%. Your 'box' is about 1000px wide, so you loose about 0.9% * 2 * 1000px = 18px. Older version of Opera had even more such "precision" problems, 9.5 corrected many of them, but this one with percentage width and another one with em borders are still alive. I have some test pages here [1], [2]. Of the current browsers Opera is the only one not displaying a smooth triangle there (few pixels rounding errors are normal, but Opera deliberately ignores all percentage fractions.) Bruno [1] http://brunildo.org/test/percwidth2.pl [2] http://brunildo.org/test/embord.pl -- Bruno Fassino http://www.brunildo.org/test *** List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm Help: [EMAIL PROTECTED] ***