Re: [WSG] please help me understand how Opera (9.5) deals with em, % and pixel
tee wrote:
http://lotusseedsdesign.com/opera-test/opera.html [1] input element
or class has background color borders - no checkmark (or tiny
checkmark depending on the width/height of the class for input
element) in checkbox.
Padding seems to cover the area, and does so until the padding becomes
too large so padding from all sides starts overlapping inwards. Once
that happens the checkbox presents its status properly again.
In your test-page, with 'input {border:1px solid #555;padding: 5px;...
}' there's 1 red pixel visible when checked.
With 'input {border:1px solid #555;padding: 6px;... }' there's no
visible check-mark when checked.
With 'input {border:1px solid #555;padding: 7px;... }' the checkbox and
check-mark appears normal. Continue to do so on larger padding, but (of
course) eventually the padding becomes so large that it breaks the
checkbox apart.
Conclusion: one or more nonsensical bugs for styled checkbox in Opera.
[2] input element or class has background color, no borders -
checkmark shows up alright.
[3] input element has no background color, no borders and no class
added - see second checkbox/radio button example.
http://lotusseedsdesign.com/opera-test/opera-nobg-in-input.html
So, [1] seems to be a bug. What about [2] and [3]? Which one is the
correct rendering? It looks like none of them are correct because
they are all inconsistent.
If by "inconsistent" you mean "don't appear identical across
browser-land and OEes", then you won't find a "correct" one in Opera.
Opera seems to use its own across OSes, while for instance Firefox seems
to change with (rely on) OS. They will never look the same.
[2] is probably the best shot at "consistency" based on what I see on my
win-OSes, but it definitely doesn't look the same in Opera 9.52 and
Firefox 3.0.1 on win2K.
regards
Georg
--
http://www.gunlaug.no
***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***
Re: [WSG] please help me understand how Opera (9.5) deals with em, % and pixel
Georg, Thanks for the detailed reply. Also, in the above page, do you know which checkbox/radio button is the correct rendering in Opera? When 'Enable Styling On Forms' is checked (default) in 'opera:config', all instances of checkbox/radio button in your page look like in your image in Opera 9.20 <---> 9.60beta version on win2K. In Opera on winXP the colored area of checkbox is smaller - the size of unstyled checkbox, and therefore bordered area wider than colored area with white space in the "gap". On Mac version, it reads: 'Enable Styling of forms' and it's on by default, so I am presuming most user will see the checkbox isn't working on [1] just as I did. I realized I wasn't being clear in my checkbox question. First, in a site I was working, I have the borders and background image declared for input element and use class to overwrite t borders for radio button. I first found out the checkbox was not clickable and no 'tick - checkmark' shown in Opera. See the first first checkbox [1] in my example page (in this on you can actually see a tiny line, but in the site it has nothing.). Upon setting up the page, I decided to test a bit on checkbox and radio button, I then realized if borders are removed from input element, the 'checkmark' shows up (second checkbox without borders) [2]. At some point I accidentally deleted the borders and background color in input element, and I am seeing Opera renders the checkbox and radio button as Safari does (FF3 too) [3] http://lotusseedsdesign.com/opera-test/opera.html [1] input element or class has background color borders - no checkmark (or tiny checkmark depending on the width/height of the class for input element) in checkbox. [2] input element or class has background color, no borders - checkmark shows up alright. [3] input element has no background color, no borders and no class added - see second checkbox/radio button example. http://lotusseedsdesign.com/opera-test/opera-nobg-in-input.html So, [1] seems to be a bug. What about [2] and [3]? Which one is the correct rendering? It looks like none of them are correct because they are all inconsistent. tee *** List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm Help: [EMAIL PROTECTED] ***
Re: [WSG] please help me understand how Opera (9.5) deals with em, % and pixel
tee wrote: Hi Georg, very nice to 'see' you :) :-) http://lotusseedsdesign.com/opera-test/opera-has-issue.png I am narrowing down to the % and em that are causing many problems I have encountered. The site is in my localhost therefor I can't post it, but I will move it to a webserver maybe tomorrow or in the weekend, and email you the url (can't show the site to public) as I have stumbled 3 more issues after my post. Some of Opera's '%' and 'em' issues are old as , and should have been fixed half a decade ago. I'll see if any of those issues are creating problems for you when I get the page for debugging. Problem in debugging is that I can't recreate the exact conditions of OS and parallels you use. Should be able to improve things for the average end-user though - to the degree Opera is willing to cooperate. I see problems in that page, but even after side-by-side comparison with Fx3.0.1 I can't see opacity problems. Care to explain? You can mail me off-list if it becomes long-ish. http://marinersq.com Sorry, the opacity bug is in the homepage, the large slide show. I looked at the correct page first time around, but couldn't decide what intended rendering-effects should be. Looked a bit "challenged" in all my browsers. When the image pan out, it turns to semi-transparent and fading, but Opera shows a solid image. Indeed. Opera starts fading too late and behaves as if it it "hangs" on fading at cross-points. I reported this before and talked to the author who wrote the slideshow script, he couldn't figure out even though he was very keen to fix it. Something for those at Opera, me thinks. Looks like an excellent real-world test case. Re the unclickable and unselectable issue, here is the page. http://lotusseedsdesign.com/opera-test/opera.html Bruno wrote he has no problem selecting the content and clicking the link 'go there'. No problems with select or functions in Opera 9.20 <---> 9.60beta versions on win2K or winXP at my end. I trashed my Opera from both Mac and PC version (via Parallels), re-installed it but the problem doesn't go away. I am also seeing the same behavior (links not clickable) in the site I was working in my local server. I figure it must be something to do with %, pixel and em units I have for font size, margins, padding, width and line-height and Opera isn't smart enough to do the math. Sounds like a problem in the relationship between Opera and MacOSX, and I've quit updating/using Mac so can't test. Though very unlikely, I have started thinking maybe the problem occurs because I have more than 1 classes declared in a div. Never observed any "multiple classes" problems in Opera. Don't think there are any. Also, in the above page, do you know which checkbox/radio button is the correct rendering in Opera? When 'Enable Styling On Forms' is checked (default) in 'opera:config', all instances of checkbox/radio button in your page look like in your image in Opera 9.20 <---> 9.60beta version on win2K. In Opera on winXP the colored area of checkbox is smaller - the size of unstyled checkbox, and therefore bordered area wider than colored area with white space in the "gap". How "far" you style those elements should have no effect on function. regards Georg -- http://www.gunlaug.no *** List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm Help: [EMAIL PROTECTED] ***
Re: [WSG] please help me understand how Opera (9.5) deals with em, % and pixel
Hi Georg, very nice to 'see' you :) http://lotusseedsdesign.com/opera-test/opera-has-issue.png I am narrowing down to the % and em that are causing many problems I have encountered. The site is in my localhost therefor I can't post it, but I will move it to a webserver maybe tomorrow or in the weekend, and email you the url (can't show the site to public) as I have stumbled 3 more issues after my post. I see problems in that page, but even after side-by-side comparison with Fx3.0.1 I can't see opacity problems. Care to explain? You can mail me off-list if it becomes long-ish. http://marinersq.com Sorry, the opacity bug is in the homepage, the large slide show. When the image pan out, it turns to semi-transparent and fading, but Opera shows a solid image. I reported this before and talked to the author who wrote the slideshow script, he couldn't figure out even though he was very keen to fix it. Re the unclickable and unselectable issue, here is the page. http://lotusseedsdesign.com/opera-test/opera.html Bruno wrote he has no problem selecting the content and clicking the link 'go there'. I trashed my Opera from both Mac and PC version (via Parallels), re- installed it but the problem doesn't go away. I am also seeing the same behavior (links not clickable) in the site I was working in my local server. I figure it must be something to do with %, pixel and em units I have for font size, margins, padding, width and line-height and Opera isn't smart enough to do the math. Though very unlikely, I have started thinking maybe the problem occurs because I have more than 1 classes declared in a div. http://lotusseedsdesign.com/opera-test/opera.html http://lotusseedsdesign.com/opera-test/ss1.png (this screenshot just show that I cannot select page content in Opera) Version:9.52 Build: 4916 Platform Mac OS X System: 10.5.4 Also, in the above page, do you know which checkbox/radio button is the correct rendering in Opera? tee *** List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm Help: [EMAIL PROTECTED] ***
Re: [WSG] please help me understand how Opera (9.5) deals with em, % and pixel
tee wrote:
Is it just me kept running into issues with Opera? I really don't
remember having these problems so obvious with pre-version 9.5 that I
can't tolerate. I am getting this impression that the evolution of Opera
has come to an end and now it's rapidly reversing back to the buggy
primitive state that it almost mimics IE 6's behavior and the worse, I
don't have a conditional comment to make it works better. From some
articles I have read, I got a sense that Opera folks are very proud of
the CSS3 spec achievement, but my take is, if I have to make double
effort to have my layout render just slightly closer to FF and Safari, I
can't afford to care all those shiny CSS3 attributes and Selectors. What
is more aggravating is that it's less than 3% population that needs to
support.
Not the first time Opera, or any browser, goes into regression - or
worse. Lack of extensive real-world testing, I guess.
I can see that Opera is weakening on a few points that matters to me,
but generally the latest versions (9.5x & 9.60b) haven't created any
real problems.
Good thing Opera is hard to hack, or else those regressions might end up
being pretty permanent - as in IE. Might make it easier for some, but
not for progress.
Better to ignore those 3% or whatever if you can't make it work,
although I'm not sure Opera's performance is to blame for all your problems.
I still have not fixed the problem I have had from my previous post (and
still can't figure why the link and text are not clickable/selectable),
now I am seeing another two little problem.
In this link, it illustrates two issues: the screen shots on the right
are taken from Opera, as you can see, the background image drops to the
bottom in the breadcrumbs.
http://lotusseedsdesign.com/opera-test/opera-has-issue.png
And the button, it works likes IE does, the difference is that IE
doubles the width and Opera doubles the height.
My font size declared in the body tag:
font: normal 100.1%/1.5em
Obviously Opera got it wrong again somewhere with em and pixel
calculation. Is this another new bug or something so old that never got
fixed ?
Have no thread/post to look at for debugging. What page(s) (links)
present the problem?
And here is another Opera problem I stumbled few days ago when I was
doing Opera bug hunting:
The page uses Prototype Glider script:
http://marinersq.com/index.php?id=50
In Opera, you will see clicking the snapshot on the left changes nothing.
Turn out I goofed, I forgot to load the script in the page (time wasted:
over 24 hours):
var my_glider = new Glider('glider', {duration:0.5, frequency: 4});
But the fact that page still works (except without glider effect) in all
other browsers make me think this is just another bug of Opera.
Confirming that. Opera before 9.50 beta supported effects like that,
without scripting. No go in later Opera versions.
Example:
Here is the working page with glider script loaded.
http://marinersq.com/meet_trainers.html
if you go to homepage of the site, you will see another Opera opacity bug.
I see problems in that page, but even after side-by-side comparison with
Fx3.0.1 I can't see opacity problems. Care to explain? You can mail me
off-list if it becomes long-ish.
regards
Georg
--
http://www.gunlaug.no
***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***
[WSG] please help me understand how Opera (9.5) deals with em, % and pixel
Is it just me kept running into issues with Opera? I really don't
remember having these problems so obvious with pre-version 9.5 that I
can't tolerate. I am getting this impression that the evolution of
Opera has come to an end and now it's rapidly reversing back to the
buggy primitive state that it almost mimics IE 6's behavior and the
worse, I don't have a conditional comment to make it works better.
From some articles I have read, I got a sense that Opera folks are
very proud of the CSS3 spec achievement, but my take is, if I have to
make double effort to have my layout render just slightly closer to FF
and Safari, I can't afford to care all those shiny CSS3 attributes and
Selectors. What is more aggravating is that it's less than 3%
population that needs to support.
I still have not fixed the problem I have had from my previous post
(and still can't figure why the link and text are not clickable/
selectable), now I am seeing another two little problem.
In this link, it illustrates two issues: the screen shots on the right
are taken from Opera, as you can see, the background image drops to
the bottom in the breadcrumbs.
http://lotusseedsdesign.com/opera-test/opera-has-issue.png
And the button, it works likes IE does, the difference is that IE
doubles the width and Opera doubles the height.
My font size declared in the body tag:
font: normal 100.1%/1.5em
Obviously Opera got it wrong again somewhere with em and pixel
calculation. Is this another new bug or something so old that never
got fixed ?
And here is another Opera problem I stumbled few days ago when I was
doing Opera bug hunting:
The page uses Prototype Glider script:
http://marinersq.com/index.php?id=50
In Opera, you will see clicking the snapshot on the left changes
nothing.
Turn out I goofed, I forgot to load the script in the page (time
wasted: over 24 hours):
var my_glider = new Glider('glider', {duration:0.5, frequency: 4});
But the fact that page still works (except without glider effect) in
all other browsers make me think this is just another bug of Opera.
Here is the working page with glider script loaded.
http://marinersq.com/meet_trainers.html
if you go to homepage of the site, you will see another Opera opacity
bug.
tee
***
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
***
