* I had sent this email weeks ago and I just copy & pasted it here to start a discussion and see if others have same idea with me. Also to keep on touch with the AMS team 😁
Hello team, I'm Jinhyeok Lee from MyMusicTaste. We stream a few K-POP concerts live and I was evaluating your solution to ease some of the pain points I encountered.
I've successfully gone through a live event using your software and I was very grateful that I could control the flow of RTMP traffic right at my fingertips.
I used to rely on nginx-rtmp server and had numerous nerve-racking moments when the configuration file gets malformed by mistake.
However, I was a little bit frustrated about a few things using your solution and decided to share my experience with you.
1. There could've been more links between pages. One example could be the "Applications" view in the dashboard. Clicking an application title could let me jump right into the app (into the WebRtcEE page for example) without using hamburger menu on the right.
2. I wish I could select HLS/WebRTC/DASH for the preview popup box.
3. I wish I could resize/popup the preview player.
4. I wish I could rename the stream-name on the fly. Something other than [no name]🙂
5. I wish I could stop ever-growing storage usage when I stop watching a HLS preview. I was wondering why a stream with no recording was gaining disk usage over time. (It went
over 40GBs within an hour due to the fact that I was using a quite high bitrate, then it almost immediately went away when I stopped broadcasting.)
6. When I casually launched the AMS-EE docker image on my t3.micro EC2 instance using docker-compose without knowing its memory requirments, the instance went straight into brain-dead state due to the memory deprivation. I wish there was some kind of basic safety check(?) before booting up the server... (had some trouble rebooting the instance back to normal)
7. I wish I could keep my "Copy RTMP url" context menu even if the stream went online. (might add a little warning alongside that button though.) When I was running a test drive it slightly baffled me that the RTMP url was nowhere to be found. (again, I understand why you want to hide it while the stream is online but I still needed the url again for a different configuration 😄 )
8. When I entered an invalid outbound RTMP URL(aka restreaming RTMP), it kind of stuck and I couldn't delete it until I stopped sending RTMP traffic to that one.
* To be more specific, I had a few "stopped" AWS medialive pipelines and I mistakenly entered the 'not-yet-alive' RTMP URLs into the AMS console while broadcasting. Then I tried to "start" the AWS medialive pipelines and AWS medialive got also stuck at "starting..." state! In other words, AMS and AWS deadlocked each other until I stopped broadcasting!
9. I wish I could see what the current recording state is like. Duration, file size, bitrate... etc. Fortunately it served me well in the end but I was very curious(and a bit concerned at) how it would turn out while I was broadcasting.
10. I know this might sound very impolite but I wish the server was... more optimized for smaller instances, possibly using different languages like Go/Node/Rust etc... (just an idea)
As a fellow front-end developer I am very glad that there are some talented engineers like your team who went for innovations like AMS. And had the server-app was written in some frameworks that I'm familiar with, I would've definitely sent those features as PRs instead of one-line feedback! I'm sincerely sorry that I'm not familiar with Angular...
Question
Jinhyeok Lee
* I had sent this email weeks ago and I just copy & pasted it here to start a discussion and see if others have same idea with me. Also to keep on touch with the AMS team 😁
Hello team, I'm Jinhyeok Lee from MyMusicTaste. We stream a few K-POP concerts live and I was evaluating your solution to ease some of the pain points I encountered.
I've successfully gone through a live event using your software and I was very grateful that I could control the flow of RTMP traffic right at my fingertips.
I used to rely on nginx-rtmp server and had numerous nerve-racking moments when the configuration file gets malformed by mistake.
However, I was a little bit frustrated about a few things using your solution and decided to share my experience with you.
1. There could've been more links between pages. One example could be the "Applications" view in the dashboard. Clicking an application title could let me jump right into the app (into the WebRtcEE page for example) without using hamburger menu on the right.
2. I wish I could select HLS/WebRTC/DASH for the preview popup box.
3. I wish I could resize/popup the preview player.
4. I wish I could rename the stream-name on the fly. Something other than [no name] 🙂
5. I wish I could stop ever-growing storage usage when I stop watching a HLS preview. I was wondering why a stream with no recording was gaining disk usage over time. (It went
over 40GBs within an hour due to the fact that I was using a quite high bitrate, then it almost immediately went away when I stopped broadcasting.)
=> This No. 5 might have been fixed at the latest snapshot
6. When I casually launched the AMS-EE docker image on my t3.micro EC2 instance using docker-compose without knowing its memory requirments, the instance went straight into brain-dead state due to the memory deprivation. I wish there was some kind of basic safety check(?) before booting up the server... (had some trouble rebooting the instance back to normal)
7. I wish I could keep my "Copy RTMP url" context menu even if the stream went online. (might add a little warning alongside that button though.) When I was running a test drive it slightly baffled me that the RTMP url was nowhere to be found. (again, I understand why you want to hide it while the stream is online but I still needed the url again for a different configuration 😄 )
8. When I entered an invalid outbound RTMP URL(aka restreaming RTMP), it kind of stuck and I couldn't delete it until I stopped sending RTMP traffic to that one.
* To be more specific, I had a few "stopped" AWS medialive pipelines and I mistakenly entered the 'not-yet-alive' RTMP URLs into the AMS console while broadcasting. Then I tried to "start" the AWS medialive pipelines and AWS medialive got also stuck at "starting..." state! In other words, AMS and AWS deadlocked each other until I stopped broadcasting!
9. I wish I could see what the current recording state is like. Duration, file size, bitrate... etc. Fortunately it served me well in the end but I was very curious(and a bit concerned at) how it would turn out while I was broadcasting.
10. I know this might sound very impolite but I wish the server was... more optimized for smaller instances, possibly using different languages like Go/Node/Rust etc... (just an idea)
As a fellow front-end developer I am very glad that there are some talented engineers like your team who went for innovations like AMS. And had the server-app was written in some frameworks that I'm familiar with, I would've definitely sent those features as PRs instead of one-line feedback! I'm sincerely sorry that I'm not familiar with Angular...
Link to comment
Share on other sites
Top Posters For This Question
2
2
Popular Days
Jul 1
1
Jul 6
1
Jul 11
1
Aug 1
1
Top Posters For This Question
mekya 2 posts
Jinhyeok Lee 2 posts
Popular Days
Jul 1 2022
1 post
Jul 6 2022
1 post
Jul 11 2022
1 post
Aug 1 2022
1 post
3 answers to this question
Recommended Posts