Hi,
there is a new minor update for version 1.16:
krpano 1.16.2It fixes several bugs - updating is strongly recommended - all details about the update here:
Release Notes 1.16.2Best regards,
Klaus
Great update, the new CORS setting fixed my bug with Chrome.
In the next release I will add a setting for either disabling the CORS usage for images or for setting a custom crossOrigin value to the images (e.g. 'use-credentials') instead the currently default 'anonymous'.
Great news, thank you.
First thing, thank you for the help regarding this issue.
And when I compare the http request headers between 4.0.0 and 4.0.1 in your link (with the Chrome Developer Tools) then I see only differences are this 'Origin' line and this cookie:
Secondly, yes, the cookie seems to be the problem. When the request is made with the new standards, the request header does not contain the cookie required for access to specific content, which in turn receives an error page (hence the text/html content type). So my question would be: how can I limit access without a cookie? or can the request be changed to contain the cookie needed? (or more specifically, make the CORS with user authentication enabled / disabled omit credentials flag ; or something like the withCredentials flag).
Your server returns - Content-Type:text/html - for the images - this might be eventually a problem for Chrome.
Try to configurate your server/server-script this way that it returns the correct content/mime type for the images.Best regards,
Klaus
This is weirder than expected. On Firefox it returns the right content type, but on Chrome for some reason when loading the images by the viewer it says they are indeed "text/html", but when I access them directly by link in Chrome they have the correct content type.
Thanks for the help, I think this is the cause of the problem. I'll try to find a solution to it.
Later Edit: It seems that the problem is still 1.16.1 specific. If I use the 1.0.8.15 version, the content type is correct for all of the images. (the server script is the same for all versions)
To see the difference between versions, all you have to do is edit the URL after the redirect and put "4.0.0" instead of "4.0.1" (4.0.0 has version 1.0.8.15) . If I find out anything else, I'll post it.
Another curious issue is that desktop Safari works (content type is OK), while iPad Safari doesn't (content type incorrect).
One problem might be the HTML4 doctype - try using a HTML5 doctype instead.
I changed it to "<!DOCTYPE html>" and the results are the same.
I've updated to the new 1.16.1 and a strange issue came up. I only tested the HTML5 viewer since that's what I'm interested in for this version and it all works well in Firefox and Safari, but in Chrome I get loading errors that I cannot explain. And for some weird reason I get xml parsing failed only in IE and by looking at the dev console the xml isn't being loaded properly (it looks fragmented).
Test link http://panoramatec.com/html5test (the link has a redirect with a timer, so in case you you are inactive for a longer period of time, you have to access this link again and not refresh the page)
Another thing I noticed is that I get a license error if I add the "www." to the link which I didn't get before.
Please let me know if it's something on the implementation part or if it's a bug.
Any help would be greatly appreciated.
