... and by Brian I meant Brandon :( Yeah, I'm that good with first names, sorry :/
2009/4/7 Julian Aubourg <[email protected]> > 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 -~----------~----~----~----~------~----~------~--~---
