Thanks for your reply.! Would check again and get back.!! Gireesh
On Sat, Sep 18, 2010 at 9:52 PM, Henry Minsky <[email protected]>wrote: > So I believe the issue here is a cross-domain security error from the > browser > > Is it the case that your PHP script test.php is running on a different web > server than > your Laszlo app is running on? If so, then if the app is running SOLO, > there will be a cross-domain security issue with the XMLHTTPRequest that is > being used; the browser will not allow you to load XML data from a different > host URL than the one which the app was loaded from. The error is not being > printed out by the LFC, but I am pretty sure that is what is happening. > > Note that even if the data URL you are fetching from is the same host as > your app was served from, if it has a different TCP port number, you will > not be able to load it directly from the browser. So if your app is coming > from localhost:8080 and you're trying to fetch data from > localhost/test/foo.php, you will get a security error. > > I was able to deploy your test app as a SOLO application, and put it onto a > web server > which runs php apps (i.e. my apache server, not the tomcat server I use for > LZX development), and it works in that case, where both the app and the php > script are > on at same server host URL. > > [Note, I found a bug in the DHTML solo deployer tool in trunk, it left > some absolute pathnames in the HTML wrapper that it generates for the solo > app in index.html, I am fixing that now. ] > > > + What is really weird is that the error I am seeing shows up as the name > of the HTTP request that goes out being "OPTIONS" instead of "POST", and I > am seeing that in both Safari and Firefox. > > > > > > On Sat, Sep 18, 2010 at 6:30 AM, Gireesh Subramaniam > <[email protected]>wrote: > >> *All, >> >> Thanks for the reply.This is a sample of the XML that is generated by my >> PHP server, >> Sample code where an OL file gets XML from a php.!! >> >> It works fine when proxy="true" in canvas but dataset throws error >> >> 'client could not parse XML from server' >> >> Could some one let me know what am i doing wrong.!! >> >> OL FILE: >> >> <canvas width="100%" height="100%" proxied="false" debug="true"> >> <dataset name="dsSample" type="http" querytype="post" request="false" >> ontimeout="Debug.write('timeout')" >> src="http://localhost/Test/test.php"; oninit="" >> ondata="Debug.write(this.rawdata);" >> onerror="Debug.write('goterror'+this.getErrorString())"/> >> <view> >> <simplelayout axis="y" spacing="5"/> >> <button width="100" name="click" x="10" text="click"> >> <handler name="onclick"> >> canvas.dsSample.doRequest(); >> </handler> >> </button> >> <text datapath="dsSample:/table" text="$path{'admin/@value'}" >> x="10"/> >> </view> >> >> </canvas> >> >> PHP FILE (Test.php): >> >> >> <?php >> header("Content-type: application/xml"); >> $xmldata ='<?xml version="1.0" encoding="ISO-8859-1"?>'; >> $xmldata .='<table>'; >> $xmldata .='<admin value="SOLO"/>'; >> $xmldata .='</table>'; >> >> echo $xmldata; >> ?> >> >> Content-type: text/xml was also tried. >> >> I validated the generated XML with a XML validator from the below link >> >> http://www.xmlvalidation.com/index.php?id=1&L=0 >> >> >> >> It reports that there are no errors.The OL server delivers this XML from the >> PHP to application and it works well in Proxied mode.So was wondering if I >> am doing something wrong when it comes to solo deployment and something >> special needs to be changed apart from changing the variable proxied to >> false in canvas.! >> >> >> >> On Fri, Sep 17, 2010 at 10:15 PM, Raju Bitter < >> [email protected]> wrote: >> >> <?php >>> header("Content-type: application/xml"); >>> ... >>> >>> should work >>> >>> On Fri, Sep 17, 2010 at 3:58 PM, Henry Minsky <[email protected]> >>> wrote: >>> > One more thing, if you're using DHTML runtime in solo mode, the browser >>> may >>> > require that your server send the proper mime type in the HTTP >>> response, >>> > e.g., xml/data or something like that. I'm not sure about XML, I do >>> know >>> > that some browsers care about >>> > the mime type for java script includes. >>> > >>> > On Fri, Sep 17, 2010 at 9:02 AM, Gireesh Subramaniam < >>> [email protected]> >>> > wrote: >>> >> >>> >> Hi , >>> >> I tried your first suggestion and upgraded to 4.8.1 and think I have >>> >> cornered the issue. >>> >> I use php code to interact with database backend.My application passes >>> the >>> >> data from the backend to the OL via XML in the following format >>> >> <table> >>> >> <admin admin msg = "....." >>> >> user id= >>> >> user_level id >>> >> etc > >>> >> Now ,both flash and DHTML solo runtimes do not work for me. >>> >> On seeing the error code returned by the OnError function of the >>> datapath >>> >> ,I see that the error is >>> >> >>> >> 'client could not parse xml from server' >>> >> The same application works for Proxied deployment on both OL 4.7.1 and >>> OL >>> >> 4.8.1 ,seems to me that in the solo runtime of both OL 4.7.1 and OL >>> 4.8.1, >>> >> the XML parser does some strict type-checking or is there any other >>> known >>> >> issues related to this ? >>> >> Can anyone help me on this ? >>> >> On Thu, Sep 16, 2010 at 6:57 PM, P T Withington <[email protected]> >>> wrote: >>> >>> >>> >>> Suggestions: >>> >>> >>> >>> 1) Update to 4.8 >>> >>> >>> >>> 2) Turn on debugging and backtracing, see if you can generate a bug >>> >>> report. >>> >>> >>> >>> (See http://www.openlaszlo.org/misc/OpenLaszloView021210.pdf) >>> >>> >>> >>> On 2010-09-16, at 08:55, Gireesh Subramaniam wrote: >>> >>> >>> >>> > All, >>> >>> > >>> >>> > I am encountering this error when i run DHTML run time on OL 4.7.1. >>> >>> > >>> >>> > This error is not encountered on compiling but i see it on the >>> Error >>> >>> > console >>> >>> > of my browser when i try to run the DHTML application. >>> >>> > >>> >>> > Hence DHTML application does not start >>> >>> > >>> >>> > >>> >>> > *Error: this.vip.subviews is undefined* >>> >>> > *Source File: >>> >>> > http://localhost:8084/DHTMLApp/lps/includes/lfc/LFCdhtml.js* >>> >>> > *Line: 5110* >>> >>> > * >>> >>> > * >>> >>> > Can someone help me understand if it is a known bug ? >>> >>> > >>> >>> > Thanks, >>> >>> > Gireesh >>> >>> >>> >> >>> > >>> > >>> > >>> > -- >>> > Henry Minsky >>> > Software Architect >>> > [email protected] >>> > >>> > >>> > >>> >>> * >> > > > > -- > Henry Minsky > Software Architect > [email protected] > > >
