nzbhydra2
nzbhydra2 copied to clipboard
theotherp
Automatic re-try if "incomplete"
-
What do you want to happen? [No idea if this is possible with integration, but it never hurts to ask] If possible, the application somehow is linked to SABNZBD [it already sends to it] and if it notices that download X is "incomplete" when processed it would try with a similar file from a different indexer [if configured]. Whether some basic AI could be done if all outlets are exhausted, e.g. all 1080 has been tried, go to 720 I don't know.
-
If not clear, why do you want it?
-
Do you think it's something only you need or something that might be popular? I dare say IF such a wild idea could be implemented, it can only benefit people in general. I do not underestimate the possible work involved, even if it is technically possible.
-
If not clear, explain the use case.
The hard part is determining what a "similar result" is. Without deeper knowledge of the results and how they relate to each other this is basically impossible. Automation tools know this. I think I could use results with the same name which are not duplicates. And I think there's some way of referencing them in the call or the NZB but not sure about that. I'll look into it but it will take a while.
Sure, it was just a thought.
For example, I could have downloaded a file today and it was marked "incomplete", so I copied the name back to NZBHYDRA, got more results and clicked on other "versions" returned by different indexers. In some cases it took several goes to get a working one.
Even if (and I do not know how you may parse it) it looks at the file name up to some SxxExx text and then looks for 720/1080/DVD etc? That may be a more advanced version.
Anyway, just throwing it out just in case even elements of it might be possible one day.
Thanks!
On Tue, 11 Oct 2022 at 19:51, theotherp @.***> wrote:
The hard part is determining what a "similar result" is. Without deeper knowledge of the results and how they relate to each other this is basically impossible. Automation tools know this. I think I could use results with the same name which are not duplicates. And I think there's some way of referencing them in the call or the NZB but not sure about that. I'll look into it but it will take a while.
— Reply to this email directly, view it on GitHub https://github.com/theotherp/nzbhydra2/issues/801#issuecomment-1274992763, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABV5MABZUEYSFPIM2WFOC73WCWLHBANCNFSM6AAAAAARCOKXOI . You are receiving this because you authored the thread.Message ID: @.***>