Superb quality and spec AB-Com PULSe 4K SE. Crazy offer! Only £129! FREE UK DELIVERY! 4K UHD, Enigma 2, Multiboot 4 images & more!...
Superb quality and spec AB-Com PULSe 4K Rev II Twin Satellite tuner only £179! FREE UK DELIVERY! 4K UHD, Enigma 2, SATA HDD facility, Multiboot 4 images & more!...

Build my own Vix image

Well I followed birdman's example and have eventually added a spare m2 drive in a usb caddy and everything seems to be working fine :thumbsup:
I've built 3 images on the usb drive and all succeeded.

Just one query though, I noticed the following warnings during the dinobot build. Anyone know what they mean and is it something I've done wrong?

WARNING: enigma2-plugin-skincomponents-kiddac-shared-skin-1.01+gitAUTOINC+12fe907be3-r1 do_package_qa: QA Issue: enigma2-plugin-skincomponents-kiddac-shared-skin: /usr/lib/enigma2/python/Components/Converter/Slyk18ClockToText.py is owned by uid 1000, which is the same as the user running bitbake. This may be due to host contamination
enigma2-plugin-skincomponents-kiddac-shared-skin: /usr/lib/enigma2/python/Components/Converter/Slyk18MovieInfo.py is owned by uid 1000, which is the same as the user running bitbake. This may be due to host contamination
enigma2-plugin-skincomponents-kiddac-shared-skin: /usr/lib/enigma2/python/Components/Converter/Slyk18RemainingToText.py is owned by uid 1000, which is the same as the user running bitbake. This may be due to host contamination
enigma2-plugin-skincomponents-kiddac-shared-skin: /usr/lib/enigma2/python/Components/Converter/Slyk18ServicePosition.py is owned by uid 1000, which is the same as the user running bitbake. This may be due to host contamination
enigma2-plugin-skincomponents-kiddac-shared-skin: /usr/lib/enigma2/python/Components/Converter/SlykServiceInfo.py is owned by uid 1000, which is the same as the user running bitbake. This may be due to host contamination
enigma2-plugin-skincomponents-kiddac-shared-skin: /usr/lib/enigma2/python/Components/Renderer/OnyxNextEvents.py is owned by uid 1000, which is the same as the user running bitbake. This may be due to host contamination
enigma2-plugin-skincomponents-kiddac-shared-skin: /usr/lib/enigma2/python/Components/Renderer/OnyxNextEvents2.py is owned by uid 1000, which is the same as the user running bitbake. This may be due to host contamination
enigma2-plugin-skincomponents-kiddac-shared-skin: /usr/lib/enigma2/python/Components/Renderer/Slyk18Picon.py is owned by uid 1000, which is the same as the user running bitbake. This may be due to host contamination
enigma2-plugin-skincomponents-kiddac-shared-skin: /usr/lib/enigma2/python/Components/Renderer/Slyk18XPicon.py is owned by uid 1000, which is the same as the user running bitbake. This may be due to host contamination
enigma2-plugin-skincomponents-kiddac-shared-skin: /usr/lib/enigma2/python/Components/Renderer/SlykNextEvents.py is owned by uid 1000, which is the same as the user running bitbake. This may be due to host contamination
enigma2-plugin-skincomponents-kiddac-shared-skin: /usr/lib/enigma2/python/Components/Renderer/SlykNextEventsTime.py is owned by uid 1000, which is the same as the user running bitbake. This may be due to host contamination [host-user-contaminated]
WARNING: enigma2-plugin-skincomponents-kiddac-shared-skin-1.01+gitAUTOINC+12fe907be3-r1 do_package_qa: QA Issue: enigma2-plugin-skincomponents-kiddac-shared-skin: /usr/share/enigma2/slyk-common/timeformat.txt is owned by uid 1000, which is the same as the user running bitbake. This may be due to host contamination [host-user-contaminated]
 
I seem to remember those Kiddac warnings ages ago, I don't think you need to worry about them, but Kiddac needs to sort them out sometime.
 
Kiddac did not create the bitbake files, so nothing to do with him.

You have skins building at the same time that depend on the same shared component. If someone knows how to tweak them, we can improve the bitbake files to remove the notification.
 
Last edited:
Just one query though, I noticed the following warnings during the dinobot build. Anyone know what they mean and is it something I've done wrong?
No.
It's the way the enigma2-plugin-skincomponents-kiddac-shared-skin_1.*.ipk file has been packaged.
The files in it should be owned by root, but are actually owned by the producing id.
So there's a bug somewhere in the packaging code.

EDIT:
And I suspect it is kiddac's "fault".

His do_install() macro includes:
Code:
    cp -rp ${S}${libdir}/* ${D}${libdir}/ 
    cp -rp ${S}${datadir}/* ${D}${datadir}/
so the files get copied in under their original id, which will be the running user.

Might need something like:

Code:
cp -R --no-dereference --preserve=mode,timestamps,links,xattr ......

which is used elsewhere.
Mind you, none of this should be running as root, so this could well be wrong.
 
Last edited:
And I suspect it is kiddac's "fault".

His do_install() macro includes:
Code:
    cp -rp ${S}${libdir}/* ${D}${libdir}/ 
    cp -rp ${S}${datadir}/* ${D}${datadir}/
so the files get copied in under their original id, which will be the running user.
I was right.
If you remove the -p option to the cp commands the warnings go away.
And checking the uid/gid set in the .ipk file shows that it is now root.
(His other 3 skin bb files all use cp -rf, not cp -rp.)
There's not much point trying to preserve timestamps, as the files have just come out of git, so will be set to "now" anyway.
 
I see Abu has fixed this on github.
Thanks - I was about to start searching for which repository it was in.
 
Are these "Taskhash mismatch" errors I keep getting recently when building something I can ignore?
I get 17 of them but the build still seems to complete and it produces image files.
 
Are these "Taskhash mismatch" errors I keep getting recently when building something I can ignore?
I get 17 of them but the build still seems to complete and it produces image files.

I have always ignored, but it is worth just checking why they have occurred - there are numerous instances where the message is valid but can be ignored
 
I have always ignored [taskhash mismatch], but it is worth just checking why they have occurred - there are numerous instances where the message is valid but can be ignored

I can't work out how.

Also I always get feeds with a totally broken samba plugin at the moment that won't even install.
Is that me or something on github?
 
I can't work out how.

Also I always get feeds with a totally broken samba plugin at the moment that won't even install.
Is that me or something on github?

Samba has always been problematic - often not problem after a new Vix update but then sometimes it will not install or start..

The work around is usually...

Install the Vix image
On reboot set-up video and basic network connection - do NOT restore anything, no scan for channels etc.

Go to network -> utilities -> samba

You will get the message
Are you ready to install packagegroup-base-smbf-server?" with the option Yes and no.

It probably now will install

After a reboot you can restore your settings.

Check that samba is running and autostart is enabled.

If you are doing something different with your own build just try and install samba before a restore of any settings.
 
.... you'll almost certainly need an opkg update before installing samba, but I reckon it's his plugin build that's got a problem with samba.
 
Also I always get feeds with a totally broken samba plugin at the moment that won't even install.
Is that me or something on github?
It was all me doing stupid things.
I fixed it just in time to not need a developer build because 5.4.004 is released.
 
There is no need to build the image each time - you can do an online software update on the dev build
 
There is no need to build the image each time - you can do an online software update on the dev build
I do make update before I do make image
I think I may have deleted something I thought was just a cache because the virtual machine for the build environment was filling up.
So I restored an old backup of the virtual machine.
But then I found the batch file I thought was correctly copying stuff from the virtual machine to my new separate web server I wanted to use for my feeds wasn't copying everything across, the logic to avoid re-copying files that were already present and up to date was wrong. It worked the first time so I wrongly assumed it was correct. But when I updated and did another build some stuff in my feeds was left from the previous build.
That was why samba wouldn't install. The packages.gz did not match what was available in the feeds. The .ipk files seem to come out slightly different sizes even if, as far as I can see, they contain exactly the same version of a package.
 
Can I set up a separate build environment for ViX 5.5 but still be able to build ViX 5.4 on the same PC or does it mess up any core files?
 
you can create multiple build environments on a PC. I tend to keep them on different HDD's so its easier to control, but you don't have to do that.
What you cannot do is run both at the same time as Bitbake always complains! - I guess there may be a way (?), but builds tend to be so CPU intensive it would be counter productive
 
Can I set up a separate build environment for ViX 5.5 but still be able to build ViX 5.4 on the same PC or does it mess up any core files?

I am presuming that you followed the OpenATV guide where they say


mkdir -p ~/openatv
cd openatv
git clone git://github.com/oe-alliance/build-enviroment.git -b 5.0


You would then be cloning 5.0 branch into the openatv directory. The instructions were the same for older branches and that is why I presume you are asking the question as you are cloning into an existing directory

What LraiZer advised/suggested a few years ago was to use a directory named by the branch that you are cloning.
Code:
[B][COLOR="#FF0000"]Link removed.

Forum rules,
6. No links to private forums, satellite forums or links/requests about porn sites to be posted[/COLOR][/B]

So, instructions would be

mkdir -p ~/oe-a-5.0
cd oe-a-5.0
git clone git://github.com/oe-alliance/build-enviroment.git -b 5.0


As you are cloning the branch, to me it makes sense to use branch names.
 
Last edited by a moderator:
I haven't actually done it yet. I wanted to make sure I wasn't going to cock up my 5.4 environment which, touch wood, has been trouble free lately.
 
So basically log in and you will be in you user folder. i.e.
Code:
/home/myuser

Now create 5.4...
Code:
git clone https://github.com/oe-alliance/build-enviroment.git -b 4.4 5.4

That will give you your 5.4 build env in /home/myuser/5.4 ...

Now create 5.5...
Code:
git clone https://github.com/oe-alliance/build-enviroment.git -b 5.0 5.5

That will give you your 5.5 build env in /home/myuser/5.5 ...

-----------------------------------------------------------------------

Now move your "sources" folder to /home/myuser/sources.

Then modify "site.conf" in both build folders to point at the sources folder.
Code:
DL_DIR = "/home/myuser/sources"
 
I am presuming that you followed the OpenATV guide where they say


mkdir -p ~/openatv
cd openatv
git clone git://github.com/oe-alliance/build-enviroment.git -b 5.0


You would then be cloning 5.0 branch into the openatv directory. The instructions were the same for older branches and that is why I presume you are asking the question as you are cloning into an existing directory

What LraiZer advised/suggested a few years ago was to use a directory named by the branch that you are cloning.
Code:
[B][COLOR="#FF0000"]Link removed.

Forum rules,
6. No links to private forums, satellite forums or links/requests about porn sites to be posted[/COLOR][/B]
So, instructions would be

mkdir -p ~/oe-a-5.0
cd oe-a-5.0
git clone git://github.com/oe-alliance/build-enviroment.git -b 5.0


As you are cloning the branch, to me it makes sense to use branch names.
Look at bottom of page:
https://github.com/OpenViX/enigma2/tree/Dev-python3-compatible
 
Last edited by a moderator:

OpenViX Feeds Status

Back
Top