shibboleth has quit [Remote host closed the connection]
emOne has joined #libreelec
egos has quit [Ping timeout: 260 seconds]
<emOne>
where exactly is balbes image published?
egos has joined #libreelec
<emOne>
I can only find the source
buzzmarshall has quit [Ping timeout: 245 seconds]
adhux0x0f0x3f has quit [Remote host closed the connection]
emOne has quit [Ping timeout: 246 seconds]
RedSoxFan07 has joined #libreelec
blackest_mamba has quit [Quit: ZNC 1.7.2+deb3 - https://znc.in]
blackest_mamba has joined #libreelec
RedSoxFan07 has quit [Quit: RedSoxFan07]
buzzmarshall has joined #libreelec
blackest_mamba has quit [Quit: ZNC 1.7.2+deb3 - https://znc.in]
blackest_mamba has joined #libreelec
mETz has quit [Ping timeout: 252 seconds]
mETz has joined #libreelec
Colty has quit [Excess Flood]
Colti has joined #libreelec
buzzmarshall has quit [Quit: Leaving]
RaphGro has joined #libreelec
RaphGro has quit [Quit: Please remember your own message. It'll be read as soon as possible.]
JohnDoe_71Rus has joined #libreelec
elektirnis has joined #libreelec
elektrinis has quit [Ping timeout: 272 seconds]
chbmb has quit [Ping timeout: 260 seconds]
andy-burns has joined #libreelec
_whitelogger has joined #libreelec
andy-burns has quit [Quit: gorn]
<tigermann>
i know Kodi/LE dont work with scheduled releases, but is there a time frame for the changes in the software stack HDR support for newer x86 platforms?
<tigermann>
Im wodering if its worth buying a Shield while I wait for progress
<chewitt>
there are people posting images with current "not fully baked but working" HDR support in the forums
<chewitt>
it's still work in progress .. but what works will be in K19 alpha1 when it arrives
<tigermann>
thanks chewitt. Do you have a guesstimate on when it alpha1 arrives and when LE10/Kodi19 final release arrives?
<chewitt>
alpha1 was supposed to ship around Christmas to beat the Python2 EOL date
<chewitt>
but we appear to be still pulling bits together
<chewitt>
final release .. no idea :)
<tigermann>
more or less then a year?
<chewitt>
"Yes"
<tigermann>
:)
<tigermann>
the images you mentioned posted on forums, is that LE-forum or Kodi-forum?
<tigermann>
cant seem to find any
<tigermann>
only milhouse nightlies
<chewitt>
I think we'd all be disappointed if K19 overran 2019, but I learned to stop guessing release dates many years ago
<chewitt>
... 2020 .. I have a headcold and am brain-dead today
<chewitt>
LE forum .. there's a thread where someone posted images
<tigermann>
you seem very coherent for a brain dead :)
<tigermann>
thanks, i'll loog firther in LE forum to find the thread
<DaRock>
why does libreelec ignore time servers set in the dhcp config
<DaRock>
?
RaphGro has joined #libreelec
chewitt has quit [Ping timeout: 268 seconds]
<DaRock>
it would seem that according to this https://forum.libreelec.tv/thread/13619-solved-custom-ntp-server/ that network protocols are deliberately broken. I fail to understand the logic here - if a NTP server is provided by the dhcp why not use it? Isn't that simpler?
<DaRock>
regardless of whether the service seems inconsequential?
<buzzmarshall>
normally his stuff is at the top of the amlogic directory on the LE forums site
<buzzmarshall>
under i think its called the Test LibreELEC images post
<buzzmarshall>
might be on his git site as well
<buzzmarshall>
those are the normal places i see them but he's probably got stuff links to stuff over on Armbians forums and others as well
<emOne>
hello buzzmarshall
<buzzmarshall>
not seen you for a bit but wanted to tell you that the mt7668 stuff after looking into some of the builds is definately not in any of the current builds
<emOne>
oh
<emOne>
that is a bit of a shame
<buzzmarshall>
but i have now found at least a couple of source sets that should let me create one
<emOne>
ooohh now that is more exciting
<emOne>
buzzmarshall: did you hear that I found the relevant dts file?
<buzzmarshall>
before that tho i am trying to get set back up to started building on LE build system again
<emOne>
fair enough
<buzzmarshall>
chewy gave me a bit of Git help the other day which is now got me building their master branch while pulling in a few other sources
<emOne>
you might want to ask chewy for a build system
<emOne>
fantastic
<emOne>
and yes I havn't been around for a few days. I promised I would test balbes image when I wasn't around
<emOne>
I idn
<buzzmarshall>
so far ive built their masters for the rockpro64 and the amlogic gxl based boxes
<emOne>
I didnt find the correct image in the end and real life got in the way of things
<emOne>
buzzmarshall: let me msg you the DTS file
<emOne>
It also includes the loader code for the wifi driver afaik
<buzzmarshall>
theres been a number of changes to their build since i last used it so i am still working my way thru it to learn the finer points to how it now works
<emOne>
can you not just make , make install ? :)
<buzzmarshall>
its nice just gonna take me a bit
<emOne>
I have never cross compiled successfully
<emOne>
it takes a lot of patience too
<emOne>
buzzmarshall: would it be easier to compile on something like a raspberry pi?
<buzzmarshall>
well with their transitioning to mainstream and my normal way being different i need to get back into their system
<buzzmarshall>
compiling on the machine your building for is something i normally don't do as it takes for ever
<buzzmarshall>
usually its much faster and easier to cross compile on a linux box
<buzzmarshall>
i've got someone looking into getting me a couple x3 f3 boxes but its gonna take a bit for them to get here as they gotta come the slow boat way
<buzzmarshall>
so i am hoping by the time they get here i will be further along it building and have something ready to start testing
<buzzmarshall>
those dumps i got from you are just to messy to work with and i needed to mess with the dtc tool and try and recompile to get it stop erroring out on the bad magic number
<emOne>
did it decompile in the end without the magic number error?
<emOne>
I wonder why it was done the way it was
<emOne>
the link above has the dts bit for wifi
<emOne>
the comments for the code are in chinese btw. you can put it through google translate to see what it says
<emOne>
basically it talks about the first part of the code being the HAL
<emOne>
it initialises the wifi from within the UI
<emOne>
I didnt say that
<emOne>
but the chinese text did , and that is not verbatim
<emOne>
what it actually does is just pull a gpio to high
<emOne>
from looking at the code
<emOne>
buzzmarshall: so if I do get this compiled . Is there a way to turn on all video gpu acceleration? :)
<emOne>
or should I just use a 4.9 kernel?
<buzzmarshall>
sorry was just off looking at what you posted
<emOne>
np
<buzzmarshall>
not sure if thats complete but i can see he was working or collecting some info
<buzzmarshall>
right now its probably best just to stick with whatever is out there that you have working and wait to see how this pans out
<buzzmarshall>
i just no that between LE current work load and the lack of interests of the other forks on this its probably gonna take a bit
<emOne>
did you find the source code for this?
<emOne>
for the wifi driver I mean
iNDi has left #libreelec ["Leaving"]
<buzzmarshall>
my plan is to get LE master branch up and running on the S912's and S905s with whatever i need to add or edit without breaking to far from LE's direction
<buzzmarshall>
and while doing so i will try and construct a more distinct x3 project as i go
<buzzmarshall>
i have a couple of sources now for the mt7668 but need to go thru them as they are for other devices
<buzzmarshall>
but the bins and all the code appears to be their
<emOne>
ooh that is cool
<buzzmarshall>
ive been kinda reading and following mediateks maillogs to see whats going on there as well
<emOne>
that is super kind of you
<emOne>
did you understand what exactly the issue is?
<buzzmarshall>
but the good news is there are other devices around using that chipset
<emOne>
are the chipsets based on ralink chipsets by any chance?
<buzzmarshall>
my quess is from what i see sofar its not in the linux sources at all, but i may have missed something on the newest kernels which is why i though i would just move into building LE sources which are working up on the newer kernel
<emOne>
I am really confused as to what the chipset actually is
<buzzmarshall>
but from what i have seen in the source tree sofar it appears they skipped over the mt7668 as the mailogs seem to indicate it was a break away from their normal old base code on the ralinks
<emOne>
is it possible that it isn't a mt7668 at all?
<buzzmarshall>
for the sake of arguement think of the wifi-bluetooth kinda like its own small little soc
<emOne>
when I type lsmod into my box it says it is using the wifi module "wlan_mt76x8_sdio"
<emOne>
buzzmarshall: ok
<buzzmarshall>
is a small asic with discrete code setup to run and present a programming api to the main soc to handle the processed its designed to do in hardware
<buzzmarshall>
in this case they handle all the wifi and bluetooth functions
mrpelotazo has quit [Ping timeout: 265 seconds]
<emOne>
so that is good
<emOne>
it means the driver is already in there somewhere
<buzzmarshall>
normally with chipsets like this they just expose a programming api to run and control them as to off load all the functions they handle in hardware away from in this case the S005x3
<buzzmarshall>
sofar from what i have seen in the 5.3 kernel sources used in LE master branch i don't specifically see the mt7668
<buzzmarshall>
i see the lower mt7603 and then the jump up and over to the mt7615's
<emOne>
how exactly is it working in android?
<buzzmarshall>
but maybe the 7668 is mixed into the base mt76 code
<buzzmarshall>
if those functions work in android then there will be a driver in the androids kernel
<buzzmarshall>
but it would need to be fished out and messed with
<buzzmarshall>
its not that i can't be done its really more of a matter of someone wanting to invest the time and energy to do so
<emOne>
got it
<emOne>
that will be me :)
<buzzmarshall>
in LE's case seeing the direction they are taking that responsibility would lie with the manufacturer releasing whats needed and doing what it takes to get it into the upstream sources
<buzzmarshall>
thats the way it really needs to be done as to off load all that work back to where it should be which is the maker
<emOne>
it doesn't look like the manufacturer is budging
<buzzmarshall>
thats kinda the whole point in getting the ducks all lined up and how mainstream is meant to work
<emOne>
are the ducks waiting to be shot?
<buzzmarshall>
well who knows if the maker will or won't as its a time consuming thing for any vendor to produce and get things setup properly in a manner that will gain acceptance by the upstream over lookers
<buzzmarshall>
its just up till know it appears they haven't made that info readily available
<emOne>
that is fair enough
<emOne>
are the ducks there as a sacrifice lol?
<emOne>
do the ducks even know? :P
<emOne>
hehhh jokes
<buzzmarshall>
i probably started a mess over on FT in the last couple of days with the CE mininions as they are trying to defend what someone i will leave nameless had said about it not being doable
<emOne>
haha yes
<emOne>
I was actually thinking about that the other day
<buzzmarshall>
personally i just don't think they want to go down that road as nothing they are trying to sell uses that chipset
<emOne>
As soon as that someone wrote it wasn't doable I knew that there was something very fishy with CEE
<emOne>
yes
<buzzmarshall>
plus outside a few good coders that are there not many would have the skillset to take it on
<emOne>
at least they can point out what is needed to support the chipseet
<buzzmarshall>
the good guys they got are busy doing other stuff i am sure
<emOne>
so is it really not a matter of something quite simple?
<emOne>
I am pretty surprised actually
<emOne>
I would have thought that it is just a matter of dropping the right dts in
<buzzmarshall>
personally i should have probably not made the posts but it irks me when people that know they can't do something try and pass off their inabilities as factual crap about it not being able to be done
<emOne>
and maybe adding something to the kernel
<emOne>
I didn't even see those posts yet
<buzzmarshall>
the kernel needs a driver and then the device needs to have the dtb to pass the proper device settings to the kernel when it starts up
<buzzmarshall>
so its kinda a 2 part thing
<emOne>
maybe the driver is already in the kernetl ?
<buzzmarshall>
so far on the newer 5 i am seeing i don't see it specifically but its base maybe combined but the more i build and play i will see
<emOne>
cool
<emOne>
I think the chinese guy is onto something
<emOne>
it might be a gal
<emOne>
:)
<emOne>
"The interface provided above is mainly used in wifi.c to dynamically load the WiFi driver."
<buzzmarshall>
as for the older 4.9 kernel that the devices are running on im not sure i want to go backwards as i really would like to follow LE lead and keep moving up and forward
<emOne>
yes I know
<buzzmarshall>
i can see he was working on a tree file but don't see any of the source for the driver but haven't gone thru everything on that site yet
<emOne>
the chinese guy has a lot of "includes" in his code
<buzzmarshall>
but theres a bunch of stuff involved in the driver such as bin files and other control stuff that all needs to be compiled and moved into place
<emOne>
it is unfortunate that the other files are not in his githun
<buzzmarshall>
i have all of that currently or at least what appears to be it from some sources of other devices
<emOne>
so can it be compiled?
<buzzmarshall>
ive not tried moving it into linux and building any of it yet
<emOne>
he was really thorough with his research from what I can see
<emOne>
fair enough
<emOne>
building and compiling on linux for arm is scary stuff
<buzzmarshall>
ya i will read some more once i get this next compile cycle running here shortly
<buzzmarshall>
building kernels isn't all that difficult
<buzzmarshall>
its more in understanding all the setting in it as they change all the time
<emOne>
I remember I was compiling something android related a while ago
<buzzmarshall>
ive been building linux kernels since linus 1st realeased them years ago and now the way they work is much much better
<emOne>
and I had to download what ended up being nearly 100 gigs of source material
<emOne>
fun times
<buzzmarshall>
thats kinda why i can appreciate what everyones trying to do by seeing everything up and into the mainstream
<buzzmarshall>
the whole device tree thing is so so much better then the old ways
<buzzmarshall>
using this scene as a example
<buzzmarshall>
ive been doing linux on android boxes pretty much since the boxes appeared
<buzzmarshall>
and for years watched as everyone struggled while a handful of good coders got things working
<buzzmarshall>
but it was all stuck down in the 3's for years
<buzzmarshall>
all the makers bsp held everyone down there
<chewitt>
he appears to have changed the image names (different to LE master)
<emOne>
good thank you!!!!
<emOne>
amlg12?
RaphGro has quit [Quit: Please remember your own message. It'll be read as soon as possible.]
<emOne>
or x?
<emOne>
12 it is
<emOne>
nice to see that he uses yandex heh
<emOne>
downloading now
<emOne>
going to test it with vim3l or sei610
<emOne>
I will read up on the differences for those two boxes first
<chewitt>
they're both S905X3 .. that's all that's relevant
<emOne>
cool
iNDi has joined #libreelec
RaphGro has joined #libreelec
<emOne>
chewitt: if I had CE installed on the system before will I need to restore first?
<emOne>
can I just restore from within android ?
<emOne>
I had CE installed on the sd card
<chewitt>
no idea
<chewitt>
I've never run CE on anything and I don't track their codebase at all .. so have zero idea what changes they inflict on the system
hijacker has quit [Quit: Leaving]
<emOne>
buzzmarshall said that it is best to restore the box first
<emOne>
so I will do that from within android ... I hope that will be enough
<emOne>
I havn't flashed anything to the internal flash
<emOne>
sooo I hope that it will be just fine
RedSoxFan07 has joined #libreelec
<buzzmarshall>
sorry been off messing around trying to get this compiled LE master on a T95U Pro box
<buzzmarshall>
as far as the CE and restore thing goes, i had originally said that in response to others saying they had issues with bables images initially installing if CE had altered the boot environmet and that he was saying to restore the box so that his newer bootloader scripts would work
<buzzmarshall>
balbes himself has a note on it on the LE thread right at the beginning about what the issue concerns
lacrimosa has joined #libreelec
<emOne>
chewitt: it isn't booting
<emOne>
I tried restoring the image from within android
<emOne>
but it still isn't booting
<emOne>
this is balbes image that isn't booting
<emOne>
I am going to try yours now
<emOne>
I am using the toothpick reset method
<emOne>
I was using the vim3l
<emOne>
I will try the s610 with yours
<buzzmarshall>
hey you gonna be here for a bit?
prahal has quit [Quit: prahal]
<emOne>
I am
<emOne>
buzzmarshall:
<emOne>
I am just burning chewitts image
<buzzmarshall>
i will be back in a couple of mins.... i gotta change my vpn and go check a oversees server for a bud
<buzzmarshall>
shouldnt be to long.
buzzmarshall has quit [Quit: Leaving]
<emOne>
enjox
RaphGro has quit [Quit: Please remember your own message. It'll be read as soon as possible.]
|Jeroen| has joined #libreelec
<emOne>
chewitt: it isn't booting either
<emOne>
with balbes image + vim3l dtb it was stuck on the logo of the box
<emOne>
(the first thing that shows up when android boots)
<emOne>
with your image + SE610 it goes to a black screen and that is all
<emOne>
I renamed the dtb to dtb.img and have put it into the root of the sd card
<emOne>
I am not sure if that is what is neeed
<emOne>
needed
elektrinis has joined #libreelec
elektirnis has quit [Ping timeout: 268 seconds]
shibboleth has joined #libreelec
ghostcube has quit [Quit: Verlassend]
prahal has joined #libreelec
_abbenormal has joined #libreelec
_abbenormal has quit [Read error: Connection reset by peer]
_abbenormal has joined #libreelec
buzzmarshall has joined #libreelec
_abbenormal has quit [Read error: Connection reset by peer]