I replied in your other thread. Note that we do not support 512MB RAM devices and faking more RAM with swap is not the solution to that.
Posts by chewitt
-
-
I want to try to create an external swap because randomly the system freezes and reboots by itself, the probable symptom is that the RAM memory is full. Trying the free command I see that about 100MB is left available and I think this is causing the random reboots
If the system randomly freezes something like power or disk corruption or an actual code bug are more likely causes. The kernel will actively manage RAM use and if there is ever an Out Of Memory (OOM) scenario the kernel oom-killer function will proactively kill idl/active processes to free space. If that happens it's fairly obvious from logs.
Users are forever seeking some magic cure for "my cheap board doesn't have enough RAM" and the ONLY solution to that problem is buying a less cheap board with more actual RAM. Swap and Zram and such can fake more; but at the cost of reduced I/O performance and high wear on the SD cards that most LE systems run from and/or reduced CPU performance (due to compression) on an already low-end device.
NB: RPi Zero was on the edge of acceptable performance with the older RPi codebase (LE 9.x) which uses highly optimised media drivers. The newer standards-based codebase (LE10/LE11) is in our opinion too heavy for acceptable performance on 512MB devices which is why we have formally dropped support for all boards/devices with 512MB.
-
You need to manually connect and cache the credential once, then you can connect/disconnect the service in scripts.
connmanctl
agent on
scan wifi
services
connect <service>
<enter pw>
-
NTFS and exFAT both use native kernel drivers now for better performance. The edges of Hell froze over a bit in the last year or so.
-
I'd guess through the same methods that Android uses, because that's the only way. It's only firmware that's in the secure world.
-
It's in one of our tools packages .. I forget which (not near the GUI to look) but "media tools" is probably a good guess.
-
and doesn't exist yet.
^ and won't exist, we have better things to be doing with our limited time
-
If you're able to play other media with audio? .. it's something specific to that file and I'd suggest using "mediainfo" to find out what codecs are needed to play the audio back (maybe some exotic thing or Kodi binary-addon is needed). If you're unable to get audio with other files, then it's something to do with audio configuration (wrong output device or such).
-
Doh.. I missed the destination folder off .. copy to /storage/.update/ and reboot
-
-
Capture the projector's EDID data to a static edid.bin file and configure the RPi3 to use this. It's normally used to address issues where the HTPC device is powered on before the TV (and thus HTPC doesn't detect audio and resolutions correctly) but it means the RPi always sees the TV (or Projector) as present and all you do is turn the projector off/on as needed. Instructions should be in the wiki.
-
I've not seen issues with stuttering and jumping in HEVC videos but the HEVC codec is unfinished so it's not impossible. I'd normally associate that kind of description with bad encodings or TV's without correct modelines (e.g. 24Hz not 23.976Hz, or perhaps 3:2 pulldown; 23.976Hz at 60Hz) where occasional corrections are required.
-
I have a VIM4 and currently see no sensible route to an LE image that runs on it due to zero upstream kernel support and a long (and not guaranteed to ever happen) time-gap until that happens. It's not simply a new iteration of the A311D (VIM3) chip, it's an all new SoC with some non-trivial IP changes (new USB, new HDMI driver, etc.) that need all-new drivers. CE can probably handle this device better since they work with vendor kernels that bundle those drivers, but even for them it's not so straightforward as Amlogic has dropped amcodec and most traces of older hardware from their latest iteration of codebase.
-
deffffz If you update using the .tar file from my share I've bumped the broadcom firmware package and it should be in the image now.
-
-
Install the current nighly test image (so latest code). Then run the SSH command and share the URL so we can see the boot log and what issues or errors might be there.
-
I'm not aware of any major issues with newer boxes, but you'd probably need to use a nightly to get newer kernel with drivers that support the latest Intel hardware. Current 'release' LE versions are probably too old.
-
There are more NUC variants than anyone cares to keep tabs upon these days so rather than ask if there are any known issues with a specific model, it's much easier to just stick the latest nightly LE image on a USB stick and try it - then come back if there are problems.