David, Erik, Sorry for my post being a little unclear.
svgdoc.createElementNS(svgNS,"script") works for Opera and Firefox. One question was if there's an alternative that works for Safari and ASV as well. The other issues are: 1) In an XMLHttpRequest context, where one file lambda.svg is downloaded by main.svg, the script(s) of lambda don't execute, if even they are in the DOM (normal?). 2) The workaround for above: load lambda; fetch the scripts strings and the onload attribute string; create script nodes in main; invoke lambda's init function as the onload event doesn't fire in any implementation (normal?). This architecture theoretically works, only, in Firefox the whole thing doesn't work because lambda is not loaded. At this point I must say that this idea comes directly from my autozoom widget (BTW, Geoff Whale has made a nice extension to it) where Firefox wouldn't load images but at least it would load svg's. After some headaches I have the brilliant idea of checking autozoom online, untouched. Result: now version 2.0.0.10 doesn't load svg's anymore. Safari doesn't do it either (unless SVG carbon). Finally, only Opera is up to this type of evolution that I want to implement in my gui, unless some of you guys can suggest another portable method. Thanks. OT Some remarks come to my mind about the state of the situation. At this point I'm frustrated with the fact that my creativity that was stimulated by the advent of SVG must come to a halt now, and I'd rather take an overdose than being condemned to do the same thing over and over again. When I read "Why is SVG development hampered by resources while MozCorp is sitting with 70 million pounds in th bank? That is what I read in a report a couple of months ago. Kinda missing the point of donations, ain't it!" (http://weblogs.mozillazine.org/tor/archives/2007/03/svg_priorities_in _firefox_3.html), when I know that the (tiny) svg team has to constantly postpone their deadlines fault of finding more volunteers, I say this is wrong. Microsoft. I actually know the real reason why they don't implement svg natively. They don't know how to do it. To do a native svg implementation is apparently difficult and for that, one needs to have some smart cookies in his sleeve, and Bill doesn't have any. He probably doesn't want any, either. I'm sure they secretly evaluated the task in-house and they got discouraged. Silverlight was safer (because blah blah mistakes blah blah comparison blah blah). I mean, they are really scared; the world should have by now places where to put away whole corporations that have gone too far out with their psychotic episodes. The huge squads of drones that they managed to produce over time should also be cured in some way. To cut it short (trust me, I'm cutting it short!), SVG has been out for almost a decade and yet now we are in the paradoxal position of having to choke our creativity; we have to wait now, see? We are also living another paradox, rediscovering the joys of having to spend loads of time with forks, switches and workarounds like in the old days of html, or else wait until the next miracle. Opera? I wish they could get 50% of the market but how can they? Domenico --- In [email protected], Erik Dahlström <[EMAIL PROTECTED]> wrote: > > On Wed, 28 Nov 2007 04:11:41 +0100, ddailey <[EMAIL PROTECTED]> > wrote: > > > Hi Nico, > > > > What a fun problem! > > > > I fiddled around with it just enough to confirm what seems like odd > > behavior in the browsers: > > > > The enclosed code (with a simple .js file of some sort added on ) shows > > very different behavior in Opera 9.5alpha, FF1.5 and IE > > with FF able to run the code from the <script> built through DOM. FF > > behaves the way I would expect it to. > > Ok, that is contrary to my experiments :) > See below for source of that. > > > I am not sure where within DOM one would expect the text of a .js (or > > .es) file to actually appear, which is why I was experimenting with > > various DOM explorations here. If the commented out code (both SVG and > > JavaScript) is commented back in then all three browsers see three > > script tags, but only in the first case, when the code is provided > > in-line, can I figure out how to interrogate the source code > > programmatically. Only in FF is the script, thusly constructed, > > executable. > > This is another problem, no? You're looking for a way of getting the text > content of a script element that has external content? > > > Since FF works the way I expect it should, it is probably the one which > > is buggy. Erik Dahlstrom has pointed out that some of the CDATA hoopla > > that I typically use is more extravagant than it needs to be, so in this > > context my extraneous hoopla could prove problematic since it does pop > > up as the nodeValue of the script that is commented out. Erik and maybe > > Martin H. have also mentioned the use of href.baseVal in setting xlinks, > > but this is on my list of things to ask about sometime, so am not sure > > why or when one would do that. > > > > cheers, > > David > > ------------------------- > > <svg xmlns="http://www.w3.org/2000/svg"; width="100%" > > xmlns:xlink="http://www.w3.org/1999/xlink"; onload="start()" > > viewBox="0 0 100 100"> > > <script><![CDATA[ > > svgRoot=document.documentElement > > svgNS="http://www.w3.org/2000/svg"; > > xlinkNS="http://www.w3.org/1999/xlink"; > > svgRoot.setAttribute("onclick","runit()") > > function start(){ > > var s = document.createElementNS(svgNS,"script"); > > svgRoot.appendChild(s); > > s.setAttributeNS(xlinkNS,"xlink:href","xxxx.js"); > > SCR=document.getElementsByTagName("script") > > alert(SCR.length+SCR.item(0).nodeName+SCR.item (0).nodeName+SCR.item(0).firstChild.nodeValue) > > alert(SCR.length+SCR.item(1).nodeName) > > alert(SCR.item(1).getAttribute("xlink:href")) > > Replacing the line above with: > > alert(SCR.item(1).getAttributeNS(xlinkNS, "xlink:href")) > > gives me the correct xlink:href back in Opera as well. > The question is whether this is correct in Firefox or not, is the name > given to getAttribute a localName or a qualifiedName? The DOM specs are > not very clear about that subject. And the same thing goes for > setAttribute. Now logged as bug 300509 in Opera's bugtracker. > > > //alert(SCR.item(2).nodeName) > > //alert(SCR.item(2).getAttribute("xlink:href")) > > runit() > > } > > //]]></script> > > <!--<script xlink:href="xxxx.js"></script>--> > > > > <rect x="0" y="0" id="R" height="100" width="100" fill="red" /> > > > > </svg> > > So, here is the simple variant that I made up which works fine in the > browsers I tested: > > <svg xmlns="http://www.w3.org/2000/svg"; > xmlns:xlink="http://www.w3.org/1999/xlink"; viewBox="0 0 1024 768" > onload="init()"> > <script> > var svgNS="http://www.w3.org/2000/svg";; > var xlinkNS="http://www.w3.org/1999/xlink";; > > function init() > { > var s = document.createElementNS(svgNS,"script"); > s.setAttributeNS(xlinkNS,"xlink:href","myscript.es"); > document.documentElement.appendChild(s); > } > > </script> > </svg> > > The file myscript.es: > > alert("hello"); > > Cheers > /Erik > > -- > http://my.opera.com/MacDev_ed/blog > ----- To unsubscribe send a message to: [EMAIL PROTECTED] -or- visit http://groups.yahoo.com/group/svg-developers and click "edit my membership" ---- Yahoo! Groups Links <*> To visit your group on the web, go to: http://groups.yahoo.com/group/svg-developers/ <*> Your email settings: Individual Email | Traditional <*> To change settings online go to: http://groups.yahoo.com/group/svg-developers/join (Yahoo! ID required) <*> To change settings via email: mailto:[EMAIL PROTECTED] mailto:[EMAIL PROTECTED] <*> To unsubscribe from this group, send an email to: [EMAIL PROTECTED] <*> Your use of Yahoo! Groups is subject to: http://docs.yahoo.com/info/terms/
