Asuswrt-Merlin 376.47 is out

[D._Manley]

another "crappy wifi" here...

.47 suddenly became really bad on the 2.4Ghz side one morning. Pinging 192.168.1.1 was in 100+ms range with packet drops although the device was like sitting right next to it (however no problem on the 5Ghz side). Rebooted the router and everything immediately went fine.

I'm reverting it back to .43 for the time being. Was running .45 for a while and it had no problem either.

Couple of simple questions from a guy who feels less than competent in router matters.

1) Is it safe to *assume* that it is generally agreed amongst the brain trust here that .43 provides the most stable & best throughput on both 2.4 & 5G?

2) Having only flashed to the .47 build after updating the original OEM version once (upon original installation), can I simply flash it normally or, perform extra steps?

On the latter question I ask both because I see lots of conflicting information about factory reset before/after and/or, clearing the nvram. I've searched enough I *think* I know how to do it if it's indeed necessary.

Thanks for any guidance if .43 is the way to go for best WiFi. FWIW, I see no performance difference for my purposes in anything but the WiFi on .47.

 

[keenism]

A couple of updates:

I cannot re-enable threading support to sqlite, as it causes other parts of the firmware to completely crash now. Only Asus can resolve this, since the code tied to this is closed source.

Various recent changes by Asus broke things such as port forwarding and the NAT loopback. This seems to be particularly visible on the AC87U. Reverting the NAT loopback code to Asus's own code partly fixes it - now the loopback works half of the time instead of never working at all.

The increasing amount of closed-source code getting added by Asus is starting to make my job more difficult than I am willing to deal with, to be honest...

Your .47 release (on an RT-AC87) has been a rock since I upgraded...5 days, 17:58 ago, with around 14 wireless clients(2.4 & 5G from PC to Macs to Phones), streaming off a 4TB drive, back ups, OpenVPN, even sharing a 10 year old HP laser printer..

I'm a bit put off that asus doesn't appear to see you as the asset you are and have someone ready to help alleviate any problems you may be having as quickly as possible...if you do have good internal communications just ignore that as obviously I'm watching from a forum...

..much thanks for your work Merlin

 

[JoeSchmoe007]

A couple of updates:
...
Various recent changes by Asus broke things such as port forwarding and the NAT loopback. This seems to be particularly visible on the AC87U. Reverting the NAT loopback code to Asus's own code partly fixes it - now the loopback works half of the time instead of never working at all.
...

Are both port forwarding and NAT loopback already broken in your .47 release? Or only in ASUS code not yet incorporated in any of your releases?

Can someone clarify on significance of NAT loopback being broken (port forwarding I understand well enough)?

 

[bxwen]

config_base.6a is the correct place to change kernel options. Be warned however that due to the amount of pre-compiled, closed source modules used by the firmware, doing kernel settings change carries a high chance of breaking things up, as those precompiled modules will no longer match the kernel build. This can lead to breaking the NTFS driver, as one example of such precompiled module.

My plan was to recompile the kernel with some modules set to "M", so that I can insert them in to memory without reflash the firmware. However, once I looked into config_base.6a, it looks so much different from the PC linux kernel I have compiled before, that I don't even know where to start.

 

[RMerlin]

Are both port forwarding and NAT loopback already broken in your .47 release? Or only in ASUS code not yet incorporated in any of your releases?

I'm not entirely sure, as I didn't try going back to an older release. Note that it's possibly specific to the AC87 - again, I did only limited testing, as I got tired and gave up after a few hours of digging into it.

Can someone clarify on significance of NAT loopback being broken (port forwarding I understand well enough)?

NAT loopback is when you try to access a port forward from within your LAN by using the WAN IP. In essence, going from LAN, to WAN, and back to LAN through a forwarded port.

 

[RMerlin]

My plan was to recompile the kernel with some modules set to "M", so that I can insert them in to memory without reflash the firmware. However, once I looked into config_base.6a, it looks so much different from the PC linux kernel I have compiled before, that I don't even know where to start.

config_base.6a is the same syntax as any standard Linux kernel .config. If unsure about which config line to add/change, keep a snapshot of the existing .config file, do a "make menuconfig" from inside your kernel tree to adjust your settings, then do a diff between the resulting .config and the backed up one. Then, apply any resulting changes to config_base.6a.

 

[RMerlin]

I'm a bit put off that asus doesn't appear to see you as the asset you are and have someone ready to help alleviate any problems you may be having as quickly as possible...if you do have good internal communications just ignore that as obviously I'm watching from a forum...

To be honest, I don't come running to them every time I run into that kind of issue. I did in the past where I knew it was something they could easily take care of (for example in case of missing GPL code). I consider contacts to be a resource best used in moderation, especially in cases such as this one where it's not really because they did anything wrong - it's just the consequence of implementing more proprietary technologies into the firmware. And the fact I don't have the patience nor the motivation anymore to keep debugging un-commented source code that I only partly understand, not having written it myself.

 

[4Leaf]

Couple of simple questions from a guy who feels less than competent in router matters.

1) Is it safe to *assume* that it is generally agreed amongst the brain trust here that .43 provides the most stable & best throughput on both 2.4 & 5G?

2) Having only flashed to the .47 build after updating the original OEM version once (upon original installation), can I simply flash it normally or, perform extra steps?

On the latter question I ask both because I see lots of conflicting information about factory reset before/after and/or, clearing the nvram. I've searched enough I *think* I know how to do it if it's indeed necessary.

Thanks for any guidance if .43 is the way to go for best WiFi. FWIW, I see no performance difference for my purposes in anything but the WiFi on .47.

Clearing your nvram and factory resetting your router are the same thing. I just recommend clearing your nvram by turning off your router and then wait 30 seconds. Next, While holding the wps button down, turn your router on and wait 10 seconds until letting go of the wps button.

Some people prefer to hold the wps button down for longer then 10 seconds just to make sure the router clears out all of the settings correctly. I personally hold the wps button for 30 seconds until I let go of it.

 

[pilotboy72]

Clearing your nvram and factory resetting your router are the same thing. I just recommend clearing your nvram by turning off your router and then wait 30 seconds. Next, While holding the wps button down, turn your router on and wait 10 seconds until letting go of the wps button.

Some people prefer to hold the wps button down for longer then 10 seconds just to make sure the router clears out all of the settings correctly. I personally hold the wps button for 30 seconds until I let go of it.

Two questions about this. First, is it the WPS button or the reset button. I have the RT-AC87 and I think it has both. Second, does this differ from going to the Administration tab and doing a factory reset from there?

Thanks,
Brian

 

[4Leaf]

Two questions about this. First, is it the WPS button or the reset button. I have the RT-AC87 and I think it has both. Second, does this differ from going to the Administration tab and doing a factory reset from there?

Thanks,
Brian

You can use both buttons, but I believe that the wps button was intended to clear the nvram and the reset button was intended to send the device into recovery mode.

You can also clear the nvram with the reset button by holding it down for 5 seconds while the router is already on. I'm not sure which button is actually better to use or if it doesn't matter. Ask Asus.

If you factory reset from the router page, that would be considered a soft reset. If you use the wps or reset button, that would be considered a hard reset. Hard resets are always considered to be more reliable.

 

[bxwen]

config_base.6a is the same syntax as any standard Linux kernel .config. If unsure about which config line to add/change, keep a snapshot of the existing .config file, do a "make menuconfig" from inside your kernel tree to adjust your settings, then do a diff between the resulting .config and the backed up one. Then, apply any resulting changes to config_base.6a.

Than you. It is hard to get such detailed teach from someone.

ASUS really should hire you, or have some contract with you. It such a good deal for them.

 

[ocram59]

Hello all
I upgraded to 376.47 and all works fine on RT-N66u
There is a small "cosmetic" regression compared to previous version

Minidlna does not have upnp icon anymore and this results in an undefined icon seen by clients like Samsung TV

This dis not happen with previous firmware versions and there was a nice icon showing an Asus Router

I noticed that favicon.icon is missing and results as a 404 error in minidlna.log

==============
[2014/10/05 14:29:18] upnphttp.c:1008: warn: /favicon.ico not found, responding ERROR 404
==============

So I guess that this explains why Samsung TV does not show the right icon anymore but a generic "?"

Of course it is just a little cosmetic "bug" and it's fix is a "nice to have"

Best Regards
Ocram59

Anybody else with this little problem?

 

[Grievous Angel]

Just here to say I have a week of uptime on this build, and it would have been longer had the power not gone out.

This is the most feature rich and most stable version of Merlin to date (for me) on the AC66 . . I know there are many variables but I am reluctant to change anything at this point.

With four people we average 10-20 connections at any given time, last month we consumed 500gb give or take of internet data ..

 

[IAAI]

They should just hire you.

I wish not . Because if they did we are all screwed .

He may not be able to update his build due to the contract/company staff regulations .

I think a talented guy like him cannot take orders from someone else.

 

[RMerlin]

I wish not . Because if they did we are all screwed .

He may not be able to update his build due to the contract/company staff regulations .

I think a talented guy like him cannot take orders from someone else.

This is probably correct. As a business, Asus works with timetables, marketing projections, and so on. Features only get added if they estimate that the time required to implement them will justify the potential of sale increases. If as a developer you are allowed "10 hours to make this code work", then you better make it fit in that 10 hours, otherwise you might be forced to move on to something else, or leave the code partly broken, and left as something "that we'll patch later". If marketing says that the product ships next month, the code HAS to be ready by then.

This is a big reason why third party firmwares are always more feature-rich than OEM firmwares. Not because the OEM developers lack the skill to develop them, but because they have to work within certain parameters, as a paid employee.

As I've often said, Asus's FW developers are definitely well skilled and know what they are doing. Some of the stuff they write is definitely beyond the skill level of an amateur like me (professionally speaking I'm an IT tech and consultant, not a programmer).

Tweaking existing code? That, I can do. Writing something from scratch? It will take me 5x more time than it would take a professional programmer, and I probably wouldn't get it right the first time.

 

[Markmaster]

Time Machine problem

Time by time got this Time Machine problem http://note.io/1xmjo9O http://note.io/1uxVZTi . If I plug out the disk and put in my Mac the Disk Utility will found nothing:
http://pastebin.com/Un4Han2V

 

[Vandergraff]

So I had been running .45 on an RT-N66 and was seeing a lot of WiFi disconnects and low throughput.

When I came here I saw there was a new release so I updated (without reading this thread). Anyway WiFi is great again with devices connecting at long range and good speed.

I am on an acre - so no interference from other networks - but also need good range for mobile devices when around the property.

So at least for me .47 seems to have fixed the WiFi issues I was seeing with .45

 

[cs987]

.47 introduced Wi-Fi slowdown problems here on my 2.4ghz network. Reverting back to .45 without resetting the configuration fixed the issues.

 

[IAAI]

This is probably correct. As a business, Asus works with timetables, marketing projections, and so on. Features only get added if they estimate that the time required to implement them will justify the potential of sale increases. If as a developer you are allowed "10 hours to make this code work", then you better make it fit in that 10 hours, otherwise you might be forced to move on to something else, or leave the code partly broken, and left as something "that we'll patch later". If marketing says that the product ships next month, the code HAS to be ready by then.



This is a big reason why third party firmwares are always more feature-rich than OEM firmwares. Not because the OEM developers lack the skill to develop them, but because they have to work within certain parameters, as a paid employee.



As I've often said, Asus's FW developers are definitely well skilled and know what they are doing. Some of the stuff they write is definitely beyond the skill level of an amateur like me (professionally speaking I'm an IT tech and consultant, not a programmer).



Tweaking existing code? That, I can do. Writing something from scratch? It will take me 5x more time than it would take a professional programmer, and I probably wouldn't get it right the first time.

I m thinking of a solution could make everyone happy ... Wait for my new thread [emoji1]

 

[Gitsum]

"Starting in August 2014, ASUS’s RT-AC87U will begin to be embedded with Trend Micro’s Smart Home Network software suite, followed by two other popular models: RT-AC56U and RT-AC68U, as well as additional new and existing models. The updates will be completed by the end of 2014. Customers who have already purchased those products can also experience the latest firmware on the Internet by upgrade"

Merlin, when will we see the "Adaptive QOS" in some of the older models??