To some extent we all love to see new features and be up to date with all the latest thing and it can be all to easy to be persuaded with higher numbers promising the very cutting edge of development.
And so, we will often test in order to provide feedback and hope that something new will wow us or an old bug has been fixed.
But we all also know that the development process is always ongoing and so it is very likely that something that worked before may not do so with a new version.
TVheadend 4.3 is a good example, as some, including myself have found has one strange behaviour, that of having fewer channels available than with 4.2.
So we report these changes and regressions, as well as confirming progressions so that developers can take things further and make them better.
Ultimately though it comes down to what works best for you as one persons priority is different from another, so opinions will differ.
Myself, I do tend to come down to using that which shows greater stability but that’s just me, which when I compare advantages and disadvantages of 4.3 against 4.2, choose 4.2 because I want access to the same amount of channels that I had before I tested an update.
You’re right. @Compent
The issue here is the reason for every problem and why I support the development of Libreelec, just like coreelec in small heads there is a division.
I realized that everything here became a joke.
There is nothing serious here to follow, team members with this type of attitude. @Shoog
It is good that this attitude comes from amateurs and not from serious people as developers.
I’m certainly not a person who sells himself for crumbs, I don’t want him here because I support the community.
Good luck to everyone.
The simple fact here is that by definition 4.2 is a stable fully mature release.
4.3 is a development release and by definition no single release is fully tested and mature. Hence the version offered in the CE repo has issues not present in the stable 4.2 version offered by CE.
Make your choice.
The topic about tvheadend 4.3
And not about which is better or worse.
But his attitude was very clear, as I am a libreelec tester, I cannot participate in the coreelec forum.
I am grateful to all the developers here, who help and little appears.
Adam and his team.
People like you purposeless amateurs get in the way.
I dropped it here.
I’m still trying to understand what the problem here is.
This is a Tvheadend 4.3 discussion thread, so what? CoreELEC team’s position is to have a stable software for people to use, LibreELEC is on the bleeding edge with its mainline efforts. We don’t have to follow them, and they don’t have to follow us, as our teams seem to have different points of focus.
If you’re a LibreELEC “tester”, there’s nothing that prevents you from being a CoreELEC user or vice versa. I would probably say that your temper and attitude need a little adjusting.
We don’t hold anyone here against their will, so don’t let the door hit you on your way out.
My account has been limited for months,
changes were made on behalf of someone with these permissions, for no reason.
It has been changed from a regular user to a user who needs permissions to post.
They changed the registration date on the forum.
I don’t know why you’re accusing me.
Feel free to block user and ip would be a favor to exclude the user.
Erbas, your first post is I guess ok? Not sure what you want to tell… There are really nice fixes since last 4.3 TVH in CE? Should we make bump? Or you already tested it, compile, and could sent PR with version bump?
The folowing posts about LE/Team members ets are really offtopic, its garbage for me as I intent to get news about TVheadend 4.3 addon
Please come back on topic track. Describe new fixes, PR would be welcome. I guess there wouldnt be problem with compiling newer version, but we always welcome testers (I compiled 4.3, working 100% for me, but some have some problems, which I didnt hit. And could be because HW, used channels, or just setup - as I am use some direct specification and other use Auto setting for the same thing).
And thats purpose of this topic. Feel welcome here, but we are interested in experiences according topic… 4.3 is here, 4.2 is here, you can find it on github I am sure, we dont need to go through whole architecture and why there is that plugin or that
I created PR with tvheadend43 based on 2af3b9e (master 2020.05.22). Feel free to test, due to lack of computing resources, so far I compiled only Amlogic platform, Amlogicng will take another day…
@Compent - would you be willing to open issue on tvheadend, and provide them logs/examples, so they could try to recreate issue? I think there best chance, how to adress this problem. Maybe its already there, but that needs to be checked by affected users.
I am not personally affected here in CZ, but it could be there will be another users who could help identify problem and give devs helping hand that way…
Thanks… I am never 100% sure, when it comes to binary addons, so i followed safest path available, to prevent any (as not likelly and hidden they can be ) possible problems always compile for the platform / ie. project.
I’ve only just spotted your post as was laid up for a time when you posted.
For the past week I have been working on comparing 4.2 to 4.3 to get at the root of the issues and first I noticed that the dvr scan file for my transmitter Sutton Coldfield was out of date, so updating that allowed me to go from 60 to 70% scanning before stalling.
I then took apart the file to make a single manual entry and tested each one in turn to see if one would not get to 100% (ignoring COM7 and COM8 entries as these will disappear from 21st June) and found that each reached 100% but oddly found that there was an overlap of channels found that one mux should not really have picked up based upon it’s settings.
However, each one individually completed successfully.
It was only when I combined them all again that scanning stalled at 70%, which defies logic unless the overlapping is the cause.
I’m still checking though differences to see what I can highlight before considering making any report to the tvh devs.
Good troubleshooting. Cross fingers. I suspected mux tables, due to that I included also dvb-tables commit, to have latest ones (from tvheadend).
This overlapping could be definitivelly problem, and I would believe that stalling could be because some racing conditions, although I would guess that scanning wouldnt be paralel. Hopefully you will be able to catch some circumstances, which are different for specific mux, and if it wont be something trivial in actual scan table (like bad characters, etc.) but really something with tvh processing them, I believe tvh devs could pick this up, if they will have chance to reproduce it, or you provide some debug logs to it.
Strangely for me I fired up 4.3 a few weeks after my initial experiments and the 30 muxes stuck queue that I encountered went away. To me this suggests that the signal was stronger the second time out and that 4.3 has a different set of filters built in to 4.2
I still went back to 4.2 because for me this sort of randomness is a pointer to problems ahead and 4.3 offers my use case absolutely nothing over 4.2
Hm… I always defined all muxes by myself (delete all others), scan it once, and then disable any sort of automatic scanning (there are more grabbers and levels, where scanning could be fired as I recall vaguelly), and I dont need to touch config again (ofc after setup for programs, their paring among different local versions (or IPTV + DVBT sources) of the same TV station, EPG setup etc etc.). When new TV station pops up (it doesnt happens often), I just rescan particular mux.
That way I am independent on scan-tables, weather, and not necessary occupied tuner by tuning
It was also interesting that even with a file with no mux references at all in the file that a scan still picks up two thirds of available channels,so there must be a default scan level of some sort, perhaps just using the hardware itself to scan every possible channel.
Thanks for the 101 version, it worked much better than the previous one. My xbox dvb tuner scanned all 33 muxes and found 155 (all / most) services. It didn’t get stuck this time after 3 muxes. Now I just need to figure out why TVH reports “[ ERROR] linuxdvb: unable to query /dev/dvb/adapter0/ca0”.
I was asked by @Portisch to test latest TVH. As I didnt watch much TV (although my relatives using 103 last half year without issue), I am offering to test latest version to whoever benefits from tvh43. Feel free to test 104 4 days old tvh commit:
) possible problems always compile for the platform / ie. project.