The window size should be stored in session file. Make sure you have autosave on or save the session before closing. It absolute does this, just not for all connections / remote devices. I wasn't able to isolate what actually causes this yet. Never happened before 3.22. Please test around a little ...

Looking good so far! 3.22 seems to be the version 3.21 was supposed to be. DPI problems all but vanished, as far as I've seen yet.

Add: Well except the fact that if you reconnect, the window is always reset as tiny size and everything is smashed together. That's not so great a thing.

owsugde - Is it possible that you are using a certificate verification feature which was introduced in this release? "ovpn - added "verify-server-certificate" parameter for OVPN client (CVE-2018-10066)" I don't think this is it, because a cold reboot done by the customer "fixed" it. At this time I ...

All I can say is be very careful with this update. I have rolled it out on some remote devices, and now, one isn't having ethernet connectivity at all (at the least) and the other isn't coming back up over OVPN. Can't say more about what has caused this in particular, yet. Probably will have to be o...

It appears that UBNT is fighting the same problem (how to detect RADAR and not detect other pulses). The before-latest UBNT firmware introduces the same problem as MikroTik has had for years.... I had read that too. With the latest firmwares, my DFS detects went from zero to zero :D . Even if it ac...

My biggest problem with MT wireless right now is the sensitivity of DFS detection. Even in relatively clear environments, I get far too many DFS changes each day. In urban settings with close device proximities and generally full spectrum, it just becomes ridiculous, with devices continually "detect...

Changes in this release: *) routerboot - fixed RouterOS booting on devices with particular NAND memory (introduced in v6.43); I suspect this fix to be incomplete, to say the least. Several of my wAP ac's in the field have become quite unstable, especially on reboots. Not sure if by .43.x or specifi...

How could you think that 60g will work without clear line of sight? That's just wireless PHY basics. The smaller the wavelength gets, the penetrative wave properties decrease rapidly but the reflective ones increase. Think low/medium frequency ≈ broadcast radio (goes literally straight through the ...

*) w60g - improved link stability;
*) w60g - improved maximum link distance;

Can anyone already quantify this in some way?

I'd like to know, because two NLoS pathways I tried were ever so slightly too far / too faint as of rc21/23. Maybe after some improvements they could actually work.

Also got enticed to install rc17 because of features and fixes but then couldn't login. Old Winbox did work though. Thanks for the tip!

Oh OK, I was under the impression that the problem was with updating from rc20 to rc24. Didn't know it was even worse than that.

Installing 6.42rc30 from rc20 makes the RB3011 unbootable! See other descriptions of update bug.

Updated from rc20 to rc30, didn't boot back up. Resetting didn't do anything, so I had to netinstall back the stable. And that even only worked after three tries.

6.42rc28 wrecked my CHR on libvirtd / Qemu KVM setup. Virtio network interfaces (or any interfaces at all) aren't recognized anymore. Do not install 6.42rc28 on KVM CHR! Still trying to salvage this thing, will write more later... Add: Just managed to get 6.41.2 installed. Had to get the stable pack...

Just tried on current LEDE (former OpenWRT) release 17.01.4. The Linux kernel in there is of version 4.4.92. I have set up separate networks to route through, so that firewall gets active. WMM is enabled on WiFi and the following firewall rule is running on all packets: iptables -t mangle -A POSTROU...

Well who am I to talk to Apple? I wouldn't even know where to write to. On the other hand, if MT could identify a problem here, they could either fix a potential bug if it's on their side, or they could maybe interact with Apple in order to get something going there. That said, I'm looking to try th...

Did some more tests on a different device in a whole separate network. Same thing happens generally. I have some new observations to share, though. There seems to be some variability on which specific priorities cause problems. When AMPDU functionality is disabled , prios 1 and 3 seem to not work at...

I've come across a very specific, but possibly quite critical problem with MacBooks connected as clients to Mikrotik APs with WMM enabled/required, running the latest ROS 6.41 (stable). The problem seems to be located somewhere in the interplay of the Mikrotik WMM implementation using priorities and...

Nothing new really, still doing the ugly workaround. Haven't come upon a method to influence the DHCP's behavior yet.

Yes, the hotspot controller is set as DHCP authorative without delay.

Not sure if this has been fixed yet in RC, but anyway, I need to report a (somewhat) critical bug with 6.39(.1): My Netmetal 5 (921UAGS-5SHPacT, the three chainer without PCIe) isn't able to use two transmit antennas anymore (say, chain 1 and 2, and not 0). All three is ok, only one is ok. With two,...

I am having trouble with a specific detail of Mikrotik routers used as hotspot gateways. First some context: I'm running hotspot networks with a Mikrotik device as the main gateway. In a roaming environment with several access points bridged together, which is rather the norm, client devices sadly t...

Yeah well, all mobile communications with moving terminals are severely affected (give or take, depends only on speed really) by Doppler shift, in particular Doppler spread. At least OFDM subchannels are flat-fading, so something will get through often enough. Even if the "orthogonal" part won't wor...

For the receive part, I'm certain Mikrotik devices do feature spatial diversity gains for every receive antenna chain you activate, by making use of maximal ratio combining (MRC). You can check this happening quite easily in Winbox, Wireless -> Registration -> {radio name} -> Signal. Here, receive s...