[request] allow manual re-download of dupe detected nzbs

Discuss newly added features or request new features.
dime
Posts: 18
Joined: 14 Aug 2013, 21:09

[request] allow manual re-download of dupe detected nzbs

Post by dime » 21 Jan 2016, 20:24

When a dupe is detected via RSS, it never enters the history and all there is is a warning in the error message notifications.

To bypass this you have to disable dupe detection, restart NZBGet, re-queue the nzb, re-enable dupe detection, and restart NZBGet once again. :evil: :evil: :evil:

Please enter the dupe into the History, in such a way that it can be manually re-downloaded, so you can easily bypass dupe detection. :idea:

hugbug
Developer & Admin
Posts: 7645
Joined: 09 Sep 2008, 11:58
Location: Germany

Re: [request] allow manual re-download of dupe detected nzbs

Post by hugbug » 21 Jan 2016, 21:10

They are added to history actually.
However in order to prevent the filling of history with incredible amounts of duplicates there are special conditions which disable adding of items to history:
- history contains an item marked as good (marked by user manually);
- history contains a hidden item with status success. For item to become hidden the history cleanup interval must expires (30 days by default). After that the item is considered good and all subsequent items are ignored.

Since the same titles are reposted again and again and again, without these measures your history would be regularly filled with titles you watched months or years ago.

When adding items via "add files dialog" you can always deactivate duplicate check for current files via a check box.

You can also remove items from history instead of changing global dupe check option and restarting. Click on button "history" to see all items, then use the search box.

dime
Posts: 18
Joined: 14 Aug 2013, 21:09

Re: [request] allow manual re-download of dupe detected nzbs

Post by dime » 21 Jan 2016, 21:12

hugbug wrote:When adding items via "add files dialog" you can always deactivate duplicate check for current files via a check box.
But not with RSS...

hugbug
Developer & Admin
Posts: 7645
Joined: 09 Sep 2008, 11:58
Location: Germany

Re: [request] allow manual re-download of dupe detected nzbs

Post by hugbug » 21 Jan 2016, 21:20

Do you mean the dialog where you see content of rss and can manually choose items to download?

dime
Posts: 18
Joined: 14 Aug 2013, 21:09

Re: [request] allow manual re-download of dupe detected nzbs

Post by dime » 21 Jan 2016, 22:59

hugbug wrote:Do you mean the dialog where you see content of rss and can manually choose items to download?
I'll have to check again but I don't think either method allows skipping dupe check.
1. newznab indexer > my cart > add to cart
2. its something already downloaded before
3. gets stuck in cart, nzbget wont download it
4. only notice is in the warnings, no way to override without steps mentioned in OP (unless I'm mistaken)

hugbug
Developer & Admin
Posts: 7645
Joined: 09 Sep 2008, 11:58
Location: Germany

Re: [request] allow manual re-download of dupe detected nzbs

Post by hugbug » 21 Jan 2016, 23:16

Cart is rss and the whole point of duplicate check is to prevent duplicates from rss feeds.

Can you explain why you download the same titles again?
no way to override without steps mentioned in OP (unless I'm mistaken)
I've offered an alternative with deleting from history.

hugbug
Developer & Admin
Posts: 7645
Joined: 09 Sep 2008, 11:58
Location: Germany

Re: [request] allow manual re-download of dupe detected nzbs

Post by hugbug » 21 Jan 2016, 23:43

In the rss filter for cart feed:

Code: Select all

Accept(dupemode:force): *

dime
Posts: 18
Joined: 14 Aug 2013, 21:09

Re: [request] allow manual re-download of dupe detected nzbs

Post by dime » 22 Jan 2016, 00:53

hugbug wrote:I've offered an alternative with deleting from history.
Instead of having to delete a perfectly valid record from history, why not add a new hidden record in the history marking the dupe?

I mean nzbget is an nzb processor is it not? What if I didn't delete from history before I tried to enqueue? Nzb and enqueue params are all lost? :cry:
hugbug wrote:They are added to history actually.
Nzb enqueue failures actually not added to history, in case the of a dupe... but dupe can be anything, different source different config params etc - lost data if you don't provide a way to retry and skip dupe check.
hugbug wrote:In the rss filter for cart feed:

Code: Select all

Accept(dupemode:force): *
I don't want to force, I want to manually re-download via web gui based on my own judgement.
Last edited by dime on 22 Jan 2016, 01:03, edited 4 times in total.

hugbug
Developer & Admin
Posts: 7645
Joined: 09 Sep 2008, 11:58
Location: Germany

Re: [request] allow manual re-download of dupe detected nzbs

Post by hugbug » 22 Jan 2016, 07:11

Duplicate check is designed for certain use cases. And I've explained in detail the reasons when duplicates are not added to history. If that doesn't fill into your usage patterns you have an option to disable duplicate check altogether and a make a dumb nzb processor out of it.
dime wrote:I don't want to force, I want to manually re-download via web gui based on my own judgement.
Rss is about automation. You can't talk about own judgment and rss in one sentence. When you add files via web-ui you pefectly can disable duplicate check for each nzb.

There is also no universal solution for all problems. That's why I'm asking about use cases. You talked about rss cart problem and I gave you a solution. If you have other use cases where the problem arises please explain them.

Doing what you ask (always add to history) isn't a solution for reasons I' ve explained in my first post.
Nzb enqueue failures actually not added to history, in case the of a dupe.
What do you mean with "enqueue failures"? Everything is meant to be added to history unless (see my first post).

dime
Posts: 18
Joined: 14 Aug 2013, 21:09

Re: [request] allow manual re-download of dupe detected nzbs

Post by dime » 22 Jan 2016, 20:48

hugbug wrote:
dime wrote:I don't want to force, I want to manually re-download via web gui based on my own judgement.
Rss is about automation. You can't talk about own judgment and rss in one sentence.
Rss is also about newznab's "my cart" functionality. At least that is the only way Nzbget can interface with these indexers. If deleting a hidden record is going to be the only means you provide to allow an override, then so be it. Instead of doing what it does now (deleting an nzb from indexer's rss feed, and not providing any way to override), I suggest adding a "dupe" record in the history. At least then that particular case of the nzb (and its associated category/priority/etc) is preserved.

The problem with the current way is that the nzb is lost. So unless I anticipate that I am downloading a dupe, and delete the original history item before manually adding the dupe to the cart, will this nzb be treated in an expected way. If I don't delete the original history item first the nzb is effectively lost. That is what I am asking for here, a way to not lose the dupe nzb. The current situation is that I have to go back to the indexer, find thee nzb, re-add it to the cart a second time, and then it queues up properly. I would rather Nzbget just provide a means for this case, rather than just completely ignoring the nzb and effectively deleting it.

Nzbget should never just delete an nzb that is handed to it, without a means to process it again. Just my opinion. The nzb in this case comes from rss, but since newznab cart has &delete=1 param attached to its rss url, the nzb is removed from the indexer cart and also never added to Nzbget at all. Nzbget is deleting the nzb and all the category/etc attached with it, simply because it was detected to be a dupe.

Post Reply

Who is online

Users browsing this forum: No registered users and 13 guests