• 0 Posts
  • 27 Comments
Joined 7 months ago
cake
Cake day: April 13th, 2024

help-circle

  • I’ll just quote the OpenWRT Wiki here, because I think half the comments here confuse mesh and roaming:

    Are you sure you want a mesh?

    If you are looking for a solution to enable your user devices to seamlessly roam from one access point to another in your home, you need 802.11r (roaming), not 802.11s.

    It is unfortunate that some manufacturers have used the word “Mesh” for marketing purposes to describe their non-standard, closed source, proprietary “roaming” functionality and this causes great confusion to many people when they enter the world of international standards and open source firmware for their network infrastructure.

    • The accepted standard for mesh networks is ieee802.11s.
    • The accepted standard for fast roaming of user devices is ieee802.11r.

    These are two completely unrelated standards.

    Source: https://openwrt.org/docs/guide-user/network/wifi/mesh/802-11s#are_you_sure_you_want_a_mesh







  • Is it OK to simply dd the 128GB disk to the 32GB disk using count to stop after the 16GB partition was cloned?

    I think it would work, but it seems a little overcomplicated, you can just use the partition paths as if and of of dd directly, as long as the output partition is not smaller than the input partition. For example dd if=/dev/sdc1 of=/dev/sdd1 bs=4M status=progress

    Your method would also copy the partition table I suppose, which might be something you want under specific circumstances, but then it would be a little harder to get the count right, just taking the size of partition 1 would be wrong, because there is some space before it (where the partition table lives) and dd would start at 0. You’d need to add up the start position and the size of partition 1 instead.

    Personally I would prefer making a new partition table on the new eMCC, and create a target partition on it. Then you clone the content of the partition (i.e. the file system). This way the file system UUID will still be the same, and the fstab should still work because these days it usually refers to mounts by filesystem UUID in my experience.

    If you make the target partition larger than the source partition, and you intend to use the full partition going forward you will additionally need to resize the filesystem to fit the new larger partition, for example with resize2fs.












  • I wrote a script to turn the power of the the Wifi+Bluetooth chip off, then enumerate the PCIe bus again to start it back up.

    The chip sometimes hung itself when using both. I looked for the bug and even found an Intel engineer on some mailing list admitting that they had issues with coexistance mode.

    Just turning the wireless off and back on wasn’t enough I needed to reeinitialize the hardware and that was the best way I knew.


  • Programming in C and C++ just seemed way easier on Linux at the time.

    The assistants at university would frequently distribute virtualbox images with Ubuntu within which we were supposed to do the homework. At some point I decided that just putting Ubuntu on my laptop directly would be easier because GCC is just right there in the repos, plus I was a little interested anyway.

    Then it just kept being easy, for Java, Haskell, Scala, Python, everything was just supported nicely. The network simulators we used were Linux native, the course where we were reverse engineering binaries used GDB, Android development was simple with the tools and simulator being in the repos.

    That said for gaming I still use Windows. And my workplace forces me to use macOS.