Yep, the someone was me. And as Daniel points out, the problem with sandboxing is that it is permeable but not transparent. You could go around onstart manipulations by hacking the script (i.e. point to window.top rather than window) but you'd also have to hack into all of the script's functionalities (as soon as it involves document manipulation without any given context -- i.e. a parameter that is the group of nodes to act on). And, beside, you'd need to hack jQuery itself too.
But from an engine point of view, I think what could be interesting would be to have a way to make jQuery sandboxed in an iframe and have its internal window reference being redirected. Something like a jQuery.setContextWindow() function. But I'm pretty confident this would mean quite some refactoring within jQuery itself, wouldn't it? Unless redefining the window variable within the jQuery main function closure would be enough? Anyway, this would make it possible to have an iframe content like this: <script src="JQUERY_URL" /> <script>jQuery.setContextWindow(window.top);</script> <script src="OTHER_SCRIPT_URL" /> The icing on the cake would be to have access to this context window with something like jQuery.getContextWindow() so that plugins could use it and be as transparently sandboxed as jQuery itself. 2009/4/22 Daniel Friesen <[email protected]> > > There was someone who did some work on a jQuery JSONP plugin which did > more than the simple stuff jQuery does, including onerror callbacks. It > was done using an iframe trick as well. > > An iframe is a bit of a possibility. > > if( window.jQuery !=== jQuery) > var iframe = create iframe > iframe.contentWindow.jQuery = jQuery; > create script tag and insert > jQuery(iframe).remove(); > > > However the issue with the iframe sandbox is that the document and > whatnot isn't properly available. > > ~Daniel Friesen (Dantman, Nadir-Seen-Fire) > > Andrea Giammarchi wrote: > > I feel "called in the cause" ... globalEval just evaulate code, if it > > causes problems, you need a sandbox to run multiple version of jQuery > > or whatever library (a sandbox for each library) and a manager to > > delegate tasks to the right sandbox when you need it. > > > > What's a sandbox? ... poorly speaking, an iframe and its window object > > with its global variables (libraries included) > > > > Hope this help. > > > > Best Regards > > > > On Wed, Apr 22, 2009 at 6:22 PM, John Resig <[email protected] > > <mailto:[email protected]>> wrote: > > > > > > Hmm... I do understand your predicament (wanting the > > dynamically-loaded script to use the right version of jQuery). I'm > > trying to imagine situations where your change could cause potential > > problems, but I'm not sure, yet. > > > > FWIW, this problem also affects getScript when pulling scripts from > an > > outside domain (and it can't be fixed used the technique described > > here since there's no window of opportunity where the global jQuery > > variable could be changed). > > > > I also note that you only change 'jQuery' and not '$' - it seems like > > both would be necessary, in this case. > > > > But, again, I'm worried about the external affects of manipulating > the > > global variables. > > > > --John > > > > > > > > On Wed, Apr 22, 2009 at 1:04 PM, Mark Gibson <[email protected] > > <mailto:[email protected]>> wrote: > > > > > > Hi, I'm using jQuery in a high conflict situation, ie. multiple > > > versions of jQuery in one page. > > > Unfortunately I have no way around this other than not using the > > latest jQuery. > > > > > > So first off, jQuery 1.2 is loaded and noConflict() called. > > > Later, jQuery 1.3.2 is loaded and fed into a closure as such: > > > > > > (function($) { > > > ... > > > })( jQuery.noConflict(true) ); > > > > > > Within the closure if I try to dynamically load some html that > > > contains script then the script will be executed with jQuery 1.2 as > > > the global jQuery object. > > > I believe it would be more appropriate to temporarily set the > global > > > jQuery to the one loading the script, ie. 1.3.2 in this case. > > > > > > I've got two demo's, first without the patch, second with the > patch: > > > > > > > > http://test3.internal.adaptavist.net/~mgibson/versions/load1.html > > <http://test3.internal.adaptavist.net/%7Emgibson/versions/load1.html > > > > > > > http://test3.internal.adaptavist.net/~mgibson/versions/load2.html > > <http://test3.internal.adaptavist.net/%7Emgibson/versions/load2.html > > > > > > > > It this worth considering, shall I raise a ticket? > > > > > > > > > Here's the patch: > > > > > > Index: src/core.js > > > =================================================================== > > > --- src/core.js (revision 6319) > > > +++ src/core.js (working copy) > > > @@ -290,7 +290,10 @@ > > > // Inspired by code by Andrea Giammarchi > > > // > > > http://webreflection.blogspot.com/2007/08/global-scope-evaluation-and-dom.html > > > var head = > > document.getElementsByTagName("head")[0] || > > > document.documentElement, > > > - script = > > document.createElement("script"); > > > + script = > > document.createElement("script"), > > > + globaljQuery = window.jQuery; > > > + > > > + window.jQuery = jQuery; > > > > > > script.type = "text/javascript"; > > > if ( jQuery.support.scriptEval ) { > > > @@ -303,6 +306,8 @@ > > > // This arises when a base node is used > > (#2709). > > > head.insertBefore( script, head.firstChild > ); > > > head.removeChild( script ); > > > + > > > + window.jQuery = globaljQuery; > > > } > > > }, > > > > > > > > > > > > > > > > > > > > > > > > > --~--~---------~--~----~------------~-------~--~----~ 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 -~----------~----~----~----~------~----~------~--~---
