Sure, but like to the person above, i tell you try out the latest official ready to install release image and look if there are problems 
The same what i will do if i have time (hope someone here give use a test info before 
) - I am not interested to patch and build by myself cause just it is funny normal way before 24.10.0 for me was using firmware selector - was enought for me, and maybe i can go back there with this release
@Zerogiven Just tried stock OpenWrt 24.10.1. It does seem to work for the most part, but wireless is limited still (and shows only 255 dBm selectable, which is obviously wrong):
Looking as the changelog the TX power patch is still not included (neither is it in current openwrt's git).
That's really something I don't understand.
Ok did not looked at the changelog at all @Rahzadan thanks for trying, i play a bit with/withput mediatek feed and the tx power patch asap
I checked 10.24.1 yesterday, unfortunately no changes in this problem.
Thanks @Zerogiven! Same as @PentiumB, I don't have the setup or capability to compile firmware with patches. I appreciate everyone's (@woziwrt, @RafalB82, etc) contributions to OpenWrt for the BPI-R4 + BE14 board. It's greatly appreciated!
Can you try a speedtest too please, cause with official release i also had a very poor bandwidth at all... cause if it is still poor i am pretty sure that i have to include mediatek feed
if we are still talking about the official 24.10.1 I can say that the speed is ok.
5ghz gives its legal 300 (the test was carried out almost close to the router).
but the range and penetration at 7db are disgusting. sitting 2 meters from the router, I can cover the phone (receiver) with part of my body, and the signal will drop from excellent to good, I won't even mention the penetration of the wall. Bruh.
It seems that not everyone has the same problems as us. This is probably some kind of revision of the board. For example, I was one of the first to get the 8GB version of the router, when officially (for now) only 4GB is supported
@Zerogiven Just did some more testing on 24.10.1 official. 5G wifi is slow. 6G won't even connect 
Ok good to know, i also have the 4gb board... however a build from official 24.10.1 with the mediatek feed should give all possibility to update packages and install kmods by yourself 
Wozi's new build script (some posts above) would be also a good starting point for "simple users" (to be honest, if using OpenWRT you are not a simple user 
- If you mostly not know what you do inside OpenWRT than you should think if it's right using it. Just for your security )
Don't know how long i need to build an image. Latest on friday, maybe another guy (@RafalB82 / @woziwrt ) can make a build from the "v24.10.1" tag with patch and mediatek feed earlier than me
Maybe tonight or tomorrow
@RafalB82 @Zerogiven
If it's not too much trouble, please send me a link when your image is ready. I'll be happy to test it.
So I figured out how to set up a build environment in an Ubuntu VM. Using @woziwrt's script with a slight modification to checkout commit d183d7bb7827a469f09bf77f2f22fd9d70ac0ed6 on the openwrt-24.10 branch. This commit is the official 24.10.1 release. It's building now. I'll report back if it works or not!
Well done! 
for getting a build environment! With basic config it should compile, i learned that i need some host libs for some packages.
The same way which i'll try later this night (using 24.10.1 and mtk commit from wozis script)
OK. So I was able to successfully build a 24.10.1 image - but the Wireless still has the TX power problem (Wifi blasts at full dBm with no way to lower it). I know there is a way to fix this, but I'm not sure how to implement the following:
The default power control from user space is disabled to follow the maximum power from eFuse. If you would like to enable power-relevant features (e.g., SingleSKU/iw set Tx Power)
make sure to set 'sku_idx' to zero for a single SKU table or to any positive number for the index of SKU tables you want in the hostapd configuration to enable it.
You can double-check whether the value under '/sys/kernel/debug/ieee80211/phy0/mt76/sku_disable' is 0.
The above is a quote from:
After searching, I'm not sure where 'sku_idx' is - and I'm not sure where to find the hostapd configuration.
You will find the final hostapd configuration here:
openwrt/build_dir/target-aarch64_cortex-a53_musl/hostapd-wpad-full-openssl/hostapd-2025.02.09~c8c7d56a/src/ap/ap_config.c
In the latest mtk commits sku_idx = -1 so only max tx_power from eFuse is available. This setting is here:
mtk-openwrt-feeds/autobuild/unified/filogic/mac80211/24.10/files/package/network/services/hostapd/patches/0184-fixup-mtk-hostapd-Add-txpower-vendor-command.patch
If you want to experiment with tx_power from userspace, you should remove or modify this patch and leave only patch
mtk-openwrt-feeds/autobuild/unified/filogic/mac80211/24.10/files/package/network/services/hostapd/patches/0115-mtk-hostapd-Add-txpower-vendor-command.patch
in which sku_idx is set to 0.
If I remember correctly, tx power was adjustable from userspace in some of my older alpha, beta builds that didn't have these patches.
Thank you @woziwrt! I modified your script slightly to add the following just underneath the wireless-regdb lines:
### remove patch to allow modifying txpower from userspace
rm -rf mtk-openwrt-feeds/autobuild/unified/filogic/mac80211/24.10/files/package/network/services/hostapd/patches/0184-fixup-mtk-hostapd-Add-txpower-vendor-command.patch
Compiling now... For some reason it takes around 2 hours (give or take) to build. Is that normal? These have been my first attemps at building/compiling code 
I will report back with my findings.
Yep I compile it on a virtual machine and it takes ages, the first full compilation 4h.
I would have to change the processor to one with more cores, at least 16 like AMD Ryzen 9 5950X
Maybe one of you has something?
With the Macbook M3 Pro it also needs about two hours by using docker 
but the docker mac implementation sucks especially for synchronized directories. So its a good idea to build openwrt inside a directory which is not synchronized to the host
One question about the tx power... any idea why mtk want to disallow user space settings? Isn't that the default way how OpenWRT should handle it? Cause especially if i want take more tx power for 5g than for 2g.
if your devices are picking 2.4ghz instead of 5ghz and both channels use the same SSID, lowering power on the 2.4 can help the devices pick the 5ghz band.
Too bad that i did not read it earlier, my test image is nearly done 
I think i'll try that too, cause it's not something i like if a setting is fixed inside OpenWRT and in austria if i remember right you should use a max of 23dbm for 5g and 30dbm for 6g (i am on 27 )
