What's new

[384.16_Alpha Builds] Testing all variants

  • SNBForums Code of Conduct

    SNBForums is a community for everyone, no matter what their level of experience.

    Please be tolerant and patient of others, especially newcomers. We are all here to share and learn!

    The rules are simple: Be patient, be nice, be helpful or be gone!

Hello friends.
I’m sorry, I’m using a translator.
I did a "dirty" upgrade (384.15 -> 384.16 alpha 1) to the alpha version of the firmware, but after a successful firmware, I lost the Internet.:mad:
I am using the AX88 as the main router, a static connecting to the Internet.
After flashing or rebooting the router, the Internet status is not connected and the Internet does not actually work.
If you disconnect in the web ui and then connect to the Internet, then everything works, but until the next reboot.
Returned to the latest stable firmware.:(
If you have ssh access to the router issue "service restart_ntpd" at the prompt and see if the connection comes up. Is the connection you are using to test wireless 5G, if so select a channel like 36 and test.
 
no, but tomorrow I’ll try to experiment.
Get winscp for windows if that's your OS on your computer. Set it up and install with it notepad++ these two tools will help you a great deal, putty wouldn't hurt as well.
 
@RMerlin the change log shows “Merged GPL 384_8137“ this is AX58 firmware yes?

The GPL code was for the RT-AX58U, but that code is still compatible with the 7968 binary blobs the RT-AX88U, so it allows me to update that router's base code. 8137 brings L2TP fixes among other things.
 
Don't update because of that fix, it's still broken. WPA3 still reports weak password in AI-Protection scan.

I already fixed that quite a while ago (and Asus also fixed it on their end).
 
The GPL code was for the RT-AX58U

So you use the code intended for the AX58 then whats the issue with supporting it ? I understand this is a hobby and your busy but maybe it's time to drop some of the legacy AC devices like the 66 and 68 and move on with the future. Just saying, know this will be a flame. O well.. o_O
 
Last edited by a moderator:
I already fixed that quite a while ago (and Asus also fixed it on their end).
The fix is not working. Not on my router with the new alpha. o_O I too have a complex password.

Edit: I was wrong one of my guest networks had a weak password.
 
Edit: I was wrong one of my guest networks had a weak password.

If it's fixed then a complex password should not be required. Am I wrong ? A complex password on a guest network is not practical.
 
Dirty upgrade to the Alpha without a hitch!
Thanks for the smooooth landing...

Update: 2 days now without a hitch!
 
Last edited:
If it's fixed then a complex password should not be required. Am I wrong ? A complex password on a guest network is not practical.
I was wrong again it was a 2.4G non-guest network that had a weak password. You're right the guest network should be no problem. I noticed this because I have 2.4 disabled on this router, when I changed the password on the disabled 2.4G network it started passing the test again. Sorry for the confusion.
 
I was wrong again it was a 2.4G non-guest network that had a weak password. You're right the guest network should be no problem. I noticed this because I have 2.4 disabled on this router, when I changed the password on the disabled 2.4G network it started passing the test again. Sorry for the confusion.

I changed all my wireless passwords to longer/more complex ones, I still get the “weak password” warning. (I used a password generator to try & get genuinely random/difficult suggestions).
I’m confused......:(
 
So you use the code intended for the AX58 then whats the issue with supporting it ? I understand this is a hobby and your busy but maybe it's time to drop some of the legacy AC devices like the 66 and 68 and move on with the future. Just saying, know this will be a flame. O well..

Why would I drop the RT-AC68U? It's the most popular model among those I currently support...

upload_2020-2-22_1-40-0.png


I've been working on a customer's server until well past midnight today, and will be on-site to deploy that new server this Saturday. Spending the 1-2 hours of spare time I have today (my lunch today consisted of a granola bar and a cup of tea while answering customer emails) before heading to bed and running into posts like yours - not going to give me any motivation to find some time. I stopped commenting on which model I may or may not support because the vast majority of the time, I can't even answer the damned question. So let me enlighten you on this specific case, since you seem so invested in it.

First thing, I got the RT-AX58U GPL only a week ago (on the 14th to be precise). Due to my work schedule, it's only two nights ago that I had the time to actually unpack, and look at it. So first thing: no, I won't be supporting a new model only 48 hours after having looked at the actual source code.

Second: this model is based on a new SDK, and new hardware platform as well. So on top of merging the GPL code itself (which was what I spent my spare time doing yesterday, as I needed first to determine if that code would be compatible with the other AX model that I have to support), I also need to review the SDK that is specific to this model. At a first glance (which I did last night before heading to bed): I'm looking at between one and two weeks of work just to CLEANUP the SDK, and another week or so to integrate it into my code. So, the code portion of the project is at minimum three weeks of work.

Third: I actually need an RT-AX58U to even be able to try anything that I compile (after the weeks of work required to cleanup and merge that code into mine). I don't have one, due to production shortage in China (take a guess why...). I can't tell if I will be able to support it until I can actually try it myself. So I need the code, I need to analyse it, I need to clean it up, I need to integrate it into my existing code. And after that, IF the code experiment was succesful, I need an actual router to debug it, and then to start analyzing the actual issues involved by running my code on this specific model.

So bottom line: can I support it? I don't know, for all the reasons I enumerated. Will I support it? Likewise, I won't know until I've had the time to actually try. Do I want to support it? Yes. But until I've spent a couple of weeks actually trying, and until I receive a development unit, I have no way to answer this question. So no, I won't make vague "Yes, I might, perhaps, I don't know" answers. The only proper answer I can give you right now is: I don't know.

And no, I won't send blindly compiled test images for people to try. New SDK and new platform = high chance that the first firmware images will fail to boot, and will require serial-attached debugging, something only myself, or another developer can do.

So, after having the code for only a single week, very little free time to look at it during that week, and no device to actually experiment, you will excuse me if I don't have an answer to make you happy RIGHT NOW. And supporting the RT-AC68U (which in itself requires very little time due to its mature platform) has nothing to do with my current inability to answer your question.

If all of this is too slow or unacceptable for your need, then by all means, download the GPL code, install the toolchain, and have a try yourself. If GNUton can support the DSL-AC68U, then it means I'm not the only person in the world with the technical know-how to do this kind of work.

Now if you will excuse me, I'd like to catch at least a few hours of sleep before heading out for that server deployment tomorrow morning.
 
Updated AX88U.
Found a problem (it's PPPoE specific in my case):
After rebooting the router, and the Web UI complete loading, it shows the "Internet is disconnected" on "Network Map" page. But in the log I found this: "custom_script: Running /jffs/scripts/wan-event (args: 0 connected)" as the last log corresponding to "wan-event" and in the log I found dnsmasq still querying but getting invalid reply because there's actually no Internet connection. This seems to mean the kernel or whatever module thought the Internet is ALIVE but actually it's OFFLINE, so it's never get reconnected (Actually it's never connected since booted) until you manually switch the button from "ON" to "OFF" and then "ON" or did a "service restart_wan".
The occurrence of this is random (not every time) but the probability of this is around 30% ~ 50%. On 384.15, I tried a lot of times and it never happened.
 
Last edited:
If you have ssh access to the router issue "service restart_ntpd" at the prompt and see if the connection comes up. Is the connection you are using to test wireless 5G, if so select a channel like 36 and test.
No, unfortunately, this did not help.
I configured ssh, the command completed successfully, but this does not help to return the Internet connection.
I waited a few minutes after running the command, it is clear that the time is not synchronized.
 
Last edited:
There is still such a question.
I am using a mesh network.
The main router is AX88U, and the nodes are AC88U and AC68U.
Everywhere is the latest firmware from Merlin. Everything works perfectly. But there is one drawback.
I use silent mode (when disable LEDs). On the main router, everything is in order, on other routers, after rebooting or updating the firmware, they turn back on.
Can this be fixed?
 

Sign Up For SNBForums Daily Digest

Get an update of what's new every day delivered to your mailbox. Sign up here!
Top