[Beta] Asuswrt-merlin 380.68 Beta is now available

[RMerlin]

Ac88u on 380.68 beta 1
Noticed on network map page, system status area

The LAN port info is only showing WAN0 and LAN1-LAN4
The other LAN port info (realtek) not displayed

Any similar occurrence on other ac88u?

Works correctly on my own RT-AC88U. Did you scroll down? There's only enough room for the first WAN=LAN1-4 ports.

 

[Odkrys]

Interesting option on 382 beta firmware for AC68U.
It seems able to implement on merlin.

 

[Odkrys]

When I reverted to merlin from 382 beta, I noticed complete button didn't work in QIS_wizard.htm page.

Uncaught TypeError: Cannot set property 'onclick' of null
at checkIsNewFirmware (QIS_finish.htm:411)
at HTMLInputElement.onclick (QIS_finish.htm:1555)

 

[RMerlin]

Interesting option on 382 beta firmware for AC68U.
It seems able to implement on merlin.

The OpenVPN implementation in firmware 382 is now closed source.

 

[gpz1100]

@Odkrys Where are you getting the 382 beta for the ac68u?

Edit: Appears if you follow the snb link in post #102, it will eventually lead you to the downloads for the ac68u and ac86u.

 

[skeal]

Works correctly for me. Make sure you don't have a guest Network enabled.

Sent from my Nexus 5X using Tapatalk

Works with my AC3100 with guest network enabled.

I added a checkbox to disable the auto-refresh.

Thanks for the support. This is my first Beta testing experience with Asuswrt-merlin. Great fix in my opinion!!

 

[netware5]

The OpenVPN implementation in firmware 382 is now closed source.

Personally I don't like this. I'm not a developer, so I don't know what is easier for you, but I would prefer if in future releases you are able to keep the open source implementation. I'm using your FW since times when the stock FW had not OpenVPN at all and your implementation was the only possible option. So may be it is time now to revert to this approach.

 

[RMerlin]

Personally I don't like this. I'm not a developer, so I don't know what is easier for you, but I would prefer if in future releases you are able to keep the open source implementation. I'm using your FW since times when the stock FW had not OpenVPN at all and your implementation was the only possible option. So may be it is time now to revert to this approach.

Even if I end up switching to the 382 GPL code base, I will definitely retain my own OpenVPN implementation over Asus' new implementation, as mine is more advanced, has OpenVPN 2.4 support, and has a far more flexible Policy-based routing than what VPNFusion seems to promise (based on what portions of the code I've seen so far).

This will mean that for the first time my branch might no longer offer everything offered by the stock firmware. This is one of the problems related to the whole GPL 382 situation. It's... complicated.

 

[RMerlin]

@RMerlin, not sure if reported yet. On the VPN connections page; when refreshing, it floods the fields with the same connection.

Thanks, fixed locally. I need to test other related fixes before pushing it to Git, as the issue might have affected other sections of that page as well.

 

[kpratte]

A couple of bugs, both of these have been around for a while.. When in Traffic Analyzer, if you are in anything other than WAN like Wireless 5Ghz, about 1 second later the focus changes to WAN. So, click on Wireless 5Ghz, switches back to WAN in 1 second, same for LAN and Wireless 2.4Ghz. Makes anything other than WAN useless.

In QoS, if you enable Traditional QoS, and put in something like 12 Up and 200 Down, they are reversed. I did a speedtest afterwards and only got 12 and 12. If you put 200 Up and 12 Down, things work as expected. Doesn't happen with Adaptive QoS.

I'm on a RT-AC56U in Router Mode, latest beta1 release.

 

[RMerlin]

A couple of bugs, both of these have been around for a while.. When in Traffic Analyzer, if you are in anything other than WAN like Wireless 5Ghz, about 1 second later the focus changes to WAN. So, click on Wireless 5Ghz, switches back to WAN in 1 second, same for LAN and Wireless 2.4Ghz. Makes anything other than WAN useless.

Cannot reproduce here.

In QoS, if you enable Traditional QoS, and put in something like 12 Up and 200 Down, they are reversed. I did a speedtest afterwards and only got 12 and 12. If you put 200 Up and 12 Down, things work as expected. Doesn't happen with Adaptive QoS.

They are not reversed. This has been discussed many times in the past, the problem is at the internal traffic "routing" level, it's not a matter of just swapping values. This is an internal issue caused by Broadcom's architecture over which I have no control.

 

[RMerlin]

When I reverted to merlin from 382 beta, I noticed complete button didn't work in QIS_wizard.htm page.

Thanks, fixed.

 

[Mutzli]

I installed the beta firmware five days ago on an RT87U and everything worked great until last night when I started to get the following error in the sys log: kernel: TCP: time wait bucket table overflow. Any ideas what could cause this sudden error?
The other two routers, RT1900 and RT88U don't have any issues.

 

[Vexira]

Wait is the 382 code base on the same kernel version or a newer version of the kernel, makes me wonder if the gt 5300 has a newer kernel version of the Linux kernel or just a new version of the existing kernel.

 

[fookyuan]

Works correctly on my own RT-AC88U. Did you scroll down? There's only enough room for the first WAN=LAN1-4 ports.

View attachment 10144

Thanks mate !! Great job !!
It wasn't displayed with the scroll bar on Chrome for android. But swiping the table reveals the rest of the ports

Anyway this ver is quite solid, tested 3 says no issues so far

Only gripe is channel 13 for 2.4Ghz not avail
5Ghz signal is now stronger than before

Verified with inSSIDer on the signal strength

 

[RMerlin]

Wait is the 382 code base on the same kernel version or a newer version of the kernel, makes me wonder if the gt 5300 has a newer kernel version of the Linux kernel or just a new version of the existing kernel.

The kernel isn't tied to the codebase, it's tied to a device's SDK. GT-AC5300 uses a much newer Broadcom SDK because of it's newer CPU, therefore it also has a newer kernel (4.1).

 

[RMerlin]

5Ghz signal is now stronger than before

It's the exact same wireless driver as 380.67, with the exact same wireless configuration code.

 

[RMerlin]

380.68 final is now available. Thank you for the beta feedback.