5.8 & 6.3 beta firmware screws up web interface when accessing device via name rather than IP address - bug in JavaScript regarding CORS

Problem: the web interface is hosed in 5.8 on my Altair. I can access it through the Lightning DS app, but not through any desktop web browser.

Reboots (even using the hardware switch) don’t help. Playback is fine.

How can we prevent automatic updates to firmware, since software QA still seems to be a common problem with Auralic?

Before jumping to conclusions, i’d recommend you have a look at your routers DCHP table. You will most likely find your Aries on another IP Address.
And, when having changed FW, you might want to power cycle your Aries by disconnecting from mains, and reattaching.

Still no web gui? Well, actually it IS a webinterface you are looking at through Lightning DS! :slight_smile:

That is not the case. (FYI I am an IT professional.) The web shell is present, but the content in the right-hand side is blank. At first blush, this appears to be a JavaScript-related error. And of course I am aware that it is what I am looking at in the app; that I why I mentioned it as distinct from desktop web browsers.

What I might need some time to look into is whether or not the new 5.8 firmware employs JavaScript on the desktop web GUI that is calling external tracking servers that are blocked by my firewall. That seems a distinct possibility, and would be unfortunate if that were something that was changed “under the hood” in this release.

Update: I did that, and I unblocked two rubiconproject-related tracking sites, and simultaneously, everything started working. However, to test things, I went back and blocked the rubiconproject sites once more, yet everything else remained working. This is all likely a coincidence. But now everything is working just fine.

My Aries Mini just updated to 5.8 FW, so I thought I would check the web interface, which I always pull up via Lightning DS, and as per normal, it came up…

Great! :slight_smile:
The screen shot you posted looks exactly like when you are trying to access a Lightning Server-page which is no longer available on that address. (Such as accessing the WLAN address when you have switched to LAN)

It’s back hosed this morning. Doesn’t look to be my firewall this time. FYI - I am on a wired network and have a DHCP reservation for my Altair - have accessed this way for 2 years. It’s obviously communicating and getting the shell, since I can get it even on browsers that have never accessed the device before (and therefore have nothing cached.)

What does this mean? Have you had “help” from auto correct or is it me? :blush:

Maybe if I had said “it is back to being hosed this morning” it might have made some more sense. Sorry if that caused confusion.

Despite all this, I still want automatic firmware updates to stop. It’s simply unprofessional to force firmware updates.

I assume “hosed” would translate into “broken” then?
I think you might have a configuration issue somewhere…

My Femto on FW 5.8 displays the Lightning Config pages just fine, either accessed from the Lightning DS app or directly in Safari on its IP-address.

Like you, i have my both Aries hard wired, and on reserved address leases from my basically configured Asus RT-N66U router.

Yes, broken. It appears it’s just a bad firmware update, It could be a fluke. I do have an Altair, and not an Aries.

Basically, the dark background behind the currently-selected item in the lefthand navigation pulses all the while that the page sits there. I don’t know if that detail will help Auralic better understand what is going on…

Ah, sorry… I assume the firmware is basically the same but i cant verify that.

I have lost my web interface on my Aries Mini FW 5.8 now also! What gives?

Have you guys just please just try to power cycle the unit?


As I noted above, I had already done this, to no avail. However, for no apparent reason, things seemed better for the last couple of days. I will keep an eye on it. But my bigger request is that you provide a function to disable automatic updates. I can ultimately just create a firewall rule that prevents my Altair from reaching out of the local network, I suppose. I haven’t done that yet because a) I didn’t know what other aspects of operation that might interfere with and b) I don’t generally like altering my firewall rules once they are set (and I would have to do that if I decided I wanted to accept a new update). Manufacturers should make this a setting, and allowing end users to decide is a good practice overall.

It’s back to being inaccessible after a reboot that was caused by a crash of Qobuz.

For what it is worth, 6.0 beta 3 exhibits the same behavior.

The JavaScript debugger on Safari repeatedly reports (replaced my local device name with [name]):

[Error] XMLHttpRequest cannot load http://[name]:45935/lightningServer-1d-29-a0-00-34-bc/av.openhome.org-ServerConfig-1/control due to access control checks.


[Error] XMLHttpRequest cannot load http://[name]:45935/lightningServer-1d-29-a0-00-34-bc/openhome.org-SubscriptionLongPoll-1/control due to access control checks.

On Google Chrome:

Access to XMLHttpRequest at ‘http://[name]:45935/lightningServer-1d-29-a0-00-34-bc/av.openhome.org-ServerConfig-1/control’ from origin ‘[name]’ has been blocked by CORS policy: Response to preflight request doesn’t pass access control check: No ‘Access-Control-Allow-Origin’ header is present on the requested resource.

Access to XMLHttpRequest at ‘http://[name]:45935/lightningServer-1d-29-a0-00-34-bc/openhome.org-SubscriptionLongPoll-1/control’ from origin ‘[name]’ has been blocked by CORS policy: Response to preflight request doesn’t pass access control check: No ‘Access-Control-Allow-Origin’ header is present on the requested resource.

Workaround: Access the device by IP address until this is fixed. But this needs to get fixed! It was not an issue in 5.7 and earlier. I prefer to access my devices by name, within my network, not by IP address. I run my own domain and have DHCP reservations for all permanently installed devices.

@Xuanqian - I updated the title of the thread for clarity. Please let us know that you have seen this and will work on fixing this. Thanks!


Software team confirmed there is no binary change of web control function in either 5.8 or 6.0. I suggest you check the other settings in your network and find out the reason. We have more than 10,000 users on these two firmware version now, if it is a firmware problem, we would expect much more report.

@Xuanqian - Thanks for checking, but it appears to be related to an updated version of openhome.org that is being used - this possibly coincides with work done in 6.0 beta 1, as it is noted in the release notes: https://support.auralic.com/hc/en-us/articles/221252547

I am betting that most users are not adding the name of their device to local DNS like I am, but are instead accessing via IP address, which works just fine.