I know that, DBJ and that's the whole problem for me : I was strictly unable to trigger the bug on simpler pages. So, basically, I have to strip down the current one which is full of generated html, initialization javascript & jsonp requests. Oh joy! Thing is I have no clue what "could" cause the problem. At all. I just posted the whole page here so that:1) you could all see there actually was a bug, 2) maybe something's so stupid in my page someone with more experience than I have will immediately intuite it.
I'm in the process of breaking down the page anyway (had to work on something else first but thought Brian or yourself would like to have a glance at what was wrong asap). Talking of the daunting task of breaking down the monster, is there any tool out there to extract the actual generated HTML and toy with it (especially when it comes to IE)? Again, I'm really sorry about the complexity of the page but I'm really stuck here. By all means, let me know if you see any mean/technique/approach that I could follow. That's the first time I have such an evasive & blocking bug jumping at my face in 24 years of programming (and that's counting a memory leak in a C/Prolog application with no proper memory inspection tools back in the days). Never thought I'd miss those days Oo -- Julian 2009/4/7 DBJDBJ <[email protected]> > > Julian , you have to present *much* simpler example ... We can not > debug your html + css, sorry. > > On Apr 6, 10:27 am, Julian Aubourg <[email protected]> wrote: > > Hi, > > > > http://icvbook.creative-area.net/ > > login: [email protected] / pass: jquerydev > > > > Just try with IE7 and you'll see the problem when drag&dropping widgets > > (easywidgets 2.0). It works perfectly as long as you keep the vertical > > scrollbar to the top but goes berserk if you scroll even a little before > > dragging. > > > > This IS a problem with .offset() since the popup lib uses the fix I > talked > > about earlier (that is to add the .scrollTop & .scrollLeft of the > $("html") > > element to get the right position for an element -- .scrollTop() and > > .scrollLeft() don't seem to give proper results on the $(window) object). > > > > I already tried to remove the xml prolog and switched to any doctype > > possible to no avail... This works on any other browser and worked on IE > > with 1.2.6. > > > > I guess there's something wrong with the html structure but I can't find > it. > > > > So, if anyone sees something jumping at his/her face, you'll be my hero ! > > > > -- Julian > > > > 2009/4/2 DBJDBJ <[email protected]> > > > > > > > > > Ok, have you tried removing the xml prolog AND having this doctype > > > > > <!DOCTYPE html> > > > > > only ? > > > > > Also :http://dbjsystems.com/jq/, take from it the very simple html > > > +jQ code, that displays the values of the jQuery.support and add it to > > > your problem page. I am curious what will be the boxModel result ? > > > With your doctype and with the one above. > > > > > On Apr 1, 4:30 pm, Julian Aubourg <[email protected]> wrote: > > > > I DID remove the prolog to no avail. > > > > > > 2009/4/1 DBJDBJ <[email protected]> > > > > > > > Remove this: <?xml version="1.0" encoding="utf-8"?> > > > > > > > An XML Declaration does trigger “Quirks Mode” in IE and unless the > > > > > document is sent as XML, e.g., application/xhtml+xml, in the > Content > > > > > Type there isn’t any benefit. > > > > > > > The HTML 5 DocType <!DOCTYPE html> triggers standards-compliant > mode > > > > > because it is not recognized. Please read the table in > > > > > > >http://msdn.microsoft.com/en-us/library/ms535242(VS.85).aspx > > > > > > > and note that an “Unrecognized DOCTYPE”, i.e., <!DOCTYPE html>, is > > > > > “On” (for standards-compliant mode). > > > > > > > Also in presence of <!DOCTYPE html > all other browsers will nicely > > > > > switch to "standards mode" ... Try. > > > > > > > On Apr 1, 12:54 pm, Julian Aubourg <[email protected]> > wrote: > > > > > > I seriously doubt it given:1) It works in every other browser (PC > and > > > MAC > > > > > > versions) > > > > > > 2) It worked with 1.2.7 > > > > > > > > But in case: > > > > > > > > <?xml version="1.0" encoding="utf-8"?><!DOCTYPE html > > > > > > PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" > > > > > > "http://www.w3c.org/TR/xhtml1/DTD/xhtml1-transitional.dtd";> > > > > > > > > <html xmlns="http://www.w3.org/1999/xhtml"; xml:lang="fr"> > > > > > > > > Now I'm trying to get a dumbed down example because it's actually > > > > > happening > > > > > > in the popup plugin I wrote and I doubt it would be really > usefull if > > > I > > > > > > dumped 1K lines of js code ;) > > > > > > > > I'm looking into the DOM structure with firebug and try & make a > > > minimal > > > > > > example. I'll get back to you asap Brian. > > > > > > > > 2009/4/1 DBJDBJ <[email protected]> > > > > > > > > > It seems you have a <!DOCTYPE >"issue" ? > > > > > > > That is if !DOCTYPE is changed to a proper one, things of this > sort > > > > > > > will start appearing, very soon ;o) > > > > > > > > > For my team I have prepared this "quick jQ 1.3.X reminder" > page: > > > > > > > > >http://dbjsystems.com/jq/ > > > > > > > > > PS: which is still called "quick test" ... ;o( > > > > > > > > > On Apr 1, 12:29 pm, Brandon Aaron <[email protected]> > wrote: > > > > > > > > Could you show me the markup of the page that you are having > > > these > > > > > issues > > > > > > > > on? > > > > > > > > -- > > > > > > > > Brandon Aaron > > > > > > > > > > On Wed, Apr 1, 2009 at 5:54 AM, Julian Aubourg < > > > > > [email protected] > > > > > > > >wrote: > > > > > > > > > > > I don't know if something changed with the 1.3.2 version > shared > > > on > > > > > > > google > > > > > > > > > but the viewport size is now correct Oo. > > > > > > > > > Anyway, I have a couple more problems (though I found > > > workarounds). > > > > > > > They > > > > > > > > > are all IE only (tested on IE7): > > > > > > > > > > > 1) $(window).scrollTop() and .scrollLeft() do not work. You > > > have to > > > > > do > > > > > > > > > $("html").scrollTop() & .scrollLeft() respectively. > > > > > > > > > 2) element.offset() doesn't take scrolls into account (I'm > > > pretty > > > > > > > confident > > > > > > > > > it is linked to bug #1). So you have to manually add > > > > > > > $("html").scrollTop() & > > > > > > > > > .scrollLeft() to have the correct coords. > > > > > > > > > > > -- Julian > > > > > > > > > > > 2009/3/30 Brandon Aaron <[email protected]> > > > > > > > > > > >> You could pull in the window logic from 1.2.6 ( > > > > > > > > >>http://dev.jquery.com/browser/tags/1.2.6/src/core.js#L1338 > ) > > > into > > > > > the > > > > > > > > >> 1.3.2 release ( > > > > > > > > >> > http://dev.jquery.com/browser/tags/1.3.2/src/dimensions.js#L25 > > > ). > > > > > > > > >> -- > > > > > > > > >> Brandon Aaron > > > > > > > > > > >> On Mon, Mar 30, 2009 at 3:48 PM, Julian Aubourg < > > > > > > > [email protected] > > > > > > > > >> > wrote: > > > > > > > > > > >>> Thanks for the quick answer Brandon. > > > > > > > > >>> And yes, I can live with a nightly build though, as you > can > > > > > imagine, > > > > > > > the > > > > > > > > >>> client is like not understanding what's taking so long > (go > > > > > explain > > > > > > > the joy > > > > > > > > >>> of version migration to a client!). Any timeframe? I > could > > > also > > > > > live > > > > > > > with an > > > > > > > > >>> ugly workaround until it's done if you happen to have > one. > > > > > > > > > > >>> -- Julian > > > > > > > > > > >>> 2009/3/30 Brandon Aaron <[email protected]> > > > > > > > > > > >>> There is a bug open on window dimensions in jQuery > 1.3.x... > > > > > > > Unfortunately > > > > > > > > >>>> it looks like we have a gap in unit tests that allowed > this > > > > > > > regression to > > > > > > > > >>>> occur. It looks like it happened in r5985. It will be > fixed > > > for > > > > > the > > > > > > > 1.3.3 > > > > > > > > >>>> release but I'm unsure when that release will happen > right > > > now. > > > > > I > > > > > > > should be > > > > > > > > >>>> able to allocate some time to fix this issue soon though > if > > > you > > > > > > > don't mind > > > > > > > > >>>> using a nightly build once it is fixed. > > > > > > > > >>>> -- > > > > > > > > >>>> Brandon Aaron > > > > > > > > > > >>>> On Mon, Mar 30, 2009 at 2:14 PM, Julian Aubourg < > > > > > > > > >>>> [email protected]> wrote: > > > > > > > > > > >>>>> Hi all, > > > > > > > > >>>>> I'm in the process of migrating a site from 1.2.7 to > 1.3.2. > > > > > > > > > > >>>>> So far, I just ran into a bug in webkit browsers that > > > forced me > > > > > to > > > > > > > > >>>>> rewrite "expr1, expr2, ... exprN" selectors into loops > (I > > > > > suppose > > > > > > > you're > > > > > > > > >>>>> aware of this one). > > > > > > > > >>>>> I also have trouble with attaching event handlers to > > > > > dynamically > > > > > > > > >>>>> created elements, but I guess it's all a question of > timing > > > > > (and > > > > > > > probably > > > > > > > > >>>>> the correct spot to look into $.live() ). > > > > > > > > > > >>>>> Anyway, I had some simple code to get the position for > a > > > div to > > > > > > > appear > > > > > > > > >>>>> at the center of the viewport: > > > > > > > > > > >>>>> getCenterPosition: > function(targetWidth,targetHeight) { > > > > > > > > >>>>> var win = $(window); > > > > > > > > >>>>> var target = { > > > > > > > > >>>>> top: > > > > > > > Math.round((win.height()-targetHeight)/2)+win.scrollTop(), > > > > > > > > >>>>> left: > > > > > > > Math.round((win.width()-targetWidth)/2)+win.scrollLeft() > > > > > > > > >>>>> } > > > > > > > > >>>>> if (target.top<0) target.top = 0; > > > > > > > > >>>>> if (target.left<0) target.left = 0; > > > > > > > > >>>>> return target; > > > > > > > > >>>>> }, > > > > > > > > > > >>>>> It was cross-browser and quite simple indeed. > > > > > > > > > > >>>>> Problem is, in 1.3.2, windows dimensions are those of > the > > > whole > > > > > > > body, > > > > > > > > >>>>> not just the viewport. I turned the code upside/down > but > > > just > > > > > can't > > > > > > > find the > > > > > > > > >>>>> proper mean to do the exact same thing with the changes > > > that > > > > > > > probably > > > > > > > > >>>>> occured at 1.3.0. > > > > > > > > > > >>>>> Since you all know the internals of jQuery, I guess > this is > > > as > > > > > good > > > > > > > a > > > > > > > > >>>>> place to ask for ideas or pointers. > > > > > > > > > > >>>>> Take care all, > > > > > > > > > > >>>>> -- Julian > > > --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "jQuery Development" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/jquery-dev?hl=en -~----------~----~----~----~------~----~------~--~---
