S3Bubble Support Forums
To stay up to date with the latest S3Bubble updates and tutorials please make sure you sign up to our YouTube Channel.
Forum Replies Created
-
AuthorPosts
-
January 12, 2019 at 4:15 am in reply to: WordPress: Streaming is breaking for some users on Safari #32440
Patrick
ParticipantWe found the cause of this streaming problem. We needed to exclude caching of the S3Bubble plug-in folder of our WordPress install. This immediately fixed the problem.
We still have a problem with iOS streaming using Media Convert. But that’s another thread.
Patrick
ParticipantSam,
Does this mean I can tie in our membership platform via Roku, and they use our WordPress credentials to stream protected content?
– patrick
Patrick
ParticipantThanks Sam! Much appreciated.
-
This reply was modified 6 years, 12 months ago by
Patrick.
Patrick
ParticipantI’ve seen this and other strange behavior since the S3 Bubble update to v5.0, using the WordPress plugin.
A large number of our members are complaining that the overlays aren’t properly displaying for playback controls. They can universally fix the problem when going into Incognito or Private mode in their browsers. Most get the controls back if they do a full delete of their cache.
We use various levels of caching on our website – but we can’t seem to force our viewers to load up the latest code without forcing a complete emptying of their cache if they’re regular visitors (a solution I consider unacceptable to deal with an update to a plug-in). We’ve emptied every layer of our caching system, except for AmazonS3.
Do we need to force a purge of our Cloudfront edge servers to clear out the old code? If so, what’s the best way?
Thanks for your help!
– patrick, mixinglight.com
Patrick
ParticipantThanks Sam!
I just updated the plug-in and the pop-up is now disabled!
Much appreciated.
Patrick
ParticipantThis behavior definitely started with the most recent version of the WordPress self-hosted plug-in. Is that what you’re doing?
I’m starting to look into the various possibilities of minification & etc, but want to make sure this isn’t the intended behavior before I contact our web developer and start spending our money before knowing what should be happening vs what is happening.
But thanks for the ideas!
Patrick
ParticipantYup. It’s working again! Thanks.
– patrick
Patrick
ParticipantThanks Sam! This all sounds terrific.
One more question:
There’s code you’re recommending we add to our theme. Is this required to fully implement the self-hosted option?
If we don’t implement the code then what’s the downside?
I just want to be able to tell our developer what we should expect will happen after executing this code vs just updating the plugin?
– patrick
Patrick
ParticipantFrom another User who was very concerned since I was seeing these kinds of errors all during our development cycle… I was really really worried.
Once we got off our dev server and out from behind the server password protection we had in place – the serving of the videos has been terrific. We had a customer in Australia who complained that the videos weren’t loading or were stopping all the time. But with a Cloudfront distribution point not far from him, we agreed it was probably a local issue.
36 hours later he came back to the site and said it was working flawlessly.
So to sum up: Once our main DNS was up and pointing to our site and the server was in its normal configuration almost all the issues I was seeing along these lines have disappeared.
FWIW.
– pat
Patrick
ParticipantI’ve been looking at s3stat.
But it still drives me nuts seeing metrics in GA that are not human-readable. I’m hoping there’s an elegant solution.
Patrick
ParticipantThat works!
Thanks much.
Patrick
ParticipantYou’re right, it’s working again!
So bizarre. I’m curious, are you aware of this happening with S3, where some clients just can’t get access for a few hours and then the network clears up?
We’ve been delivering our videos embedded in the EVS player served off regular S3 buckets for years. And once a month a client complains they can’t access our videos. After doing the normal ‘clear the browser cache’ routine, we tell them to wait 36 hours and than reach out if it’s still a problem. In most cases the problem clears up.
I’ve run into this locally with Cloudfront on our Dev site a few times the last few week (where a video acts like I hadn’t attached our Cloudfront distribution to it, but I did – and it was playing just a few hours earlier)… I can’t quite explain it (and I don’t have the technical chops to deal with the Amazon engineers who speak an entirely different form of the English language).
Patrick
ParticipantHi Sam,
The original video is still in my system: ML0606-DIT-Part3-rolling-kit
There are two HLS versions: One encoded today (the good one) and one encoded two days ago (the one that stopped working). Let me know if you need any more details.
Patrick
ParticipantConfirmed on all counts! Thank you.
– patrick
Patrick
ParticipantOkay – I’ve followed what you did in the update video and watched the network stats while waiting for the video to update.
I can confirm switcher is working for me.
Thanks for getting it fixed!
– patrick
-
This reply was modified 6 years, 12 months ago by
-
AuthorPosts