I have a couple of video-requests to an s3-bucket. the last two are cached.
Every not cached request had the expected cors-headers:
Once the video is cached, the cors-headers are missing and it comes to the well known cors-policy error:
It's only reproduceable in windows-chrome, everything works fine on my mac.
Does anyone have the same problem? Is it a browser-bug or am I able to set anything in the bucket to prevent this problem?
Related
Earlier this week the video content we use in our Test and Production environments, which is using https://player.vimeo.com/external/... links, has been throwing a bunch CORS errors in the console after loading the very first part of the video.
Sample error:
The 'Access-Control-Allow-Origin' header has a value 'null' that is not equal to the supplied origin.
The initial request we make out to player-vimeo.com does have Access-Control-Allow-Origin:*, as expected from the configuration of our application. However, that player-vimeo returns a 302 which redirects to the CDN https://vod-progressive.akamaized.net and the response we get from there has Access-Control-Allow-Origin: null.
We have not changed any of the code around this video content in months and we've been using Vimeo this way for a long time. It seems to happen in Chrome, but doesn't happen, or happens very rarely, in FireFox or on Mac. The only solution we have found so far is to remove the attribute crossorigin="anonymous" from the video tag, but that is not ideal as it breaks the getting of the caption files associated with the videos.
Weirdly, if you let the video throw the CORS error many times in a row, it actually starts working after it has thrown the error enough times.
As far as we can tell nothing we've done would have led to this breaking, it seems as if Vimeo isn't passing on the Access-Control-Allow-Origin param anymore, or somehow akamaized.net is losing it.
UPDATE
We heard back from Vimeo support 6/10/2021:
“This is an issue that our developers are aware of and are working to fix. We apologize for any inconvenience caused and appreciate your patience.
We’ll reach back out once this issue is resolved.”
The same application that I am developing works fine on Firefox but not on Chrome.
My application a day ago printed some things on the console, but I removed them today, which means that the actualized version of the application does not print anything on the console.
On Firefox it works, everything renders and there is not prints on the console.
However on Chrome, the app still prints on the console. It means that it is retaining some old state for some reason.
Does that even make sense? The code is already gone (erased), so the only way Chrome is printing the old state on the console is if he is storing it.
On Chrome I also get errors of undefined functions, but they are defined. On Firefox that does not happen.
Have you tried reloading the page? Another thing to try to really get rid of the state that the browser is obviously holding onto is to clear the browser cache.
I am a google chrome user, almost 90% of the time this browser works best for me, but most oftenly i face an issue and that is while downloading the videos, which are big in size like 400 mb or above, what happens is download proceeds till 80-90% then some error message appears and everything downloaded just gets vanished, this is while downloading videos from MVA, so i wanted to know if it is an issue with the browser or with the server, but i face it almost everytime, according to me its not the server issue.
If anyone can explain.
Yeah and I forgot to specify the error, actually the error is unknown.
The Image of the error is shown here:
And this happens very time i try to download any big video, and i have even encountered this error while downloading big application, well anything big in size lands to this problem.
I have an issue that is occurring if you try and seek a few times to different points in the video. Eventually you end up getting an error saying it's a network issue. If I download the file from the server it's fine and if I don't seek to much it's fine as well.
Wondering if anyone knows what the issue might be.
Here is an example: https://www.la1tv.co.uk/player/107/195
Just jump around a bit (in chrome) and you should get the error. I have attached a screenshot.
Thanks!
error screenshot
The issue was that I was not sending a header with an etag in the response containing the video. Adding the etag header (which in my case is the file id) solved the problem.
Here is some information about etags: http://en.wikipedia.org/wiki/HTTP_ETag
I have a website with a manifest that validates (http://manifest-validator.com/), and files are clearly downloaded since the "progress" event is fired for each file - but, window.applicationCache.status is always 0 (which mean the website does not have a cache manifest). This is in Firefox 11.
This is at first load. When i refresh the page after the manifest is donwloaded, the applicationcache status if first checking, and the idle - which is correct.
Any ideas?
I'm seeing similar results — Firefox still has major issues with applicationCache status and eventing.
I've filed bug 825618 to hopefully get this a little bit more attention again, although I'm seeing a few other similar bugs that should have flagged the problem a long time ago but are still basically unresolved.