NZBget, Feedtime, usenet indexers... and my own solution.

[lordy]

Hi there is no 'problem' , just requirements definition. Why does a user download into a category?
The answer used to be "Because they are doing their own categorisation?", so Oversight shouldn't then move the result?
(but maybe it should index it without moving it?)

But maybe the answer to the question is now different? I would like to change anything until I understand all the uses of categories?

Esp now with SickBeard etc, I think that uses categories too?

[jasch]

As an Indexer, I push the NZB's category for every file uplodad. It's up to the user if they want to use it or not.

You can disable NZBget from using different directories with the "AppendCategoryDir" config option, and this fixed the "files being extracted in different directories". But this 'category' altough not used by NZBget is still passed on to the unpak script (which apparently makes Oversigh ignore the file).

I am not saying it's Oversights fault obviously, but categories are included in the http headers of most popular indexers (nzbs, nzbmatrix, newznab, etc), which I guess nzbget will be taking advantage off at some point in the near future (hugbug is working on adding 'fetch by URL' options to the API, and possibly RSS feeds).

On the other side of the coin, you have SABnzb/Sickbeard, but since they have their own unpak and sorting scripts, they appears to have no conflicts.

[lordy]

Ah I see this is actually useful, as it makes searching easier.

Eg "Room 101.avi" is it a tv show 'Room s01e01.avi' or a movie. Oversight unpak script does some horrible contortions to try to figure that out.

But it always made more sense to work with any metadata associated with the original NZB in the first place.

I'll make Oversight use this in the near future to decide between TV and Movie search.

At the moment, for files like this, (assuming no nfo file) it starts with a TV search (to mop up simple 1x1, s01e01 etc), then failing that it tries a movie search (movies with numbers) , and then back to TV if IMDB says or no movie found.

[chris57]

Thanks for the prefer SD x264 and prefer HD WEB-DL options in preferences.

Anyone tried using this with the new CSI nzbget.r406:407?

[jasch]

That appears to be only a bugfix release.

Quote:nzbget 0.8.0-testing-r407

Changes since nzbget 0.8.0-testing-r394
improved the parsing of filename from article subject;
fixed incorrect displaying of group sizes bigger than 4GB on many 64-bit OSes;
fixed a bug causing error on decoding of input data in JSON-RPC;
fixed a compilation error on some windows versions;
fixed: par-repair could fail when the filenames were not correctly parsed from article subjects;
fixed a compatibility issue with OpenBSD (and possibly other BSD based systems); added the automatic configuring of required signal handling logic to better support BSD without breaking the compatibility with certain Linux systems;
corrected the address of Free Software Foundation in copyright notice.

[chris57]

It may be but it flashes up as an update in the CSI installer so many will automatically update. I was just wondering if anyone else had updated and if things were ok still for them as ive not had a download since i updated.

I've updated but forgot to save my conf file so I'm having to renter details and have not had anything push over and download yet. I have a few of my shows on tonight so I'll see how thing go and if nothing is there I'll try a standard download etc. and check my settings first.

[Slevin]

I've been using the latest nzbget (r407) for the the last 6-7 days..
Just remember to backup your settings as they get reset after the csi install.

Everything seems to be working fine with r407.

[jhmiller]

At the moment the conf file is removed for maximum compatability, its been like this since before i started making the CSI packages. I will from now no change the line that deletes the file and just rename it (.bak) so people will still be able to get their setting back without having to redo the whole config.

Sorry about your configs guys.

[jasch]

Maybe the new release package should come with a "nzbget.conf.orig", so it will not overwrite the original file.

If there's no "nzbget.conf" in place, "orig" get's renamed. This way .conf file will never get overwritten.

This is how most PHP installs do it, so you can update without fear of overwriting your setup.

The rename part is optional, as people can just make a copy, remove '.orig' and work on that. It makes good practice also, since you allways have an original template.

BTW, Hubbug confirmed to me that 0.8.0 should come out of 'testing' soon, and a 0.9.0-testing will be commited, with a http client. This will bring some inmediate changes that would benefit users of DOGnzb, like add-by-url, including multiple file uploads at the same time (I could push 10 urls to you, with only one simple and quick connection, and then NZBget fetches the files afterwards). This could pave the way for other features that required an http client (like downloading from RSS feeds) which Hugbug seems interested in adding.

[chris57]

Sounds interesting.

Got it up and running with the latest CSI version of nzbget. Think I had forgot to set the server ip to 0.0.0.0 to get it to listen to all interfaces. So that's a couple of us at least that have n o issues with trhe latest CSI release of nzbget and dognzb.

[jasch]

Hugbug has released 0.9.0-testing, and so far it's working great. I've ran several tests myself, and NZB pushing is way faster, since the NZB doesn't have to be uploaded, just the URL (nzbget fetches the nzb file directly from the server).

This has also paved the way for uploading multiple URLs, on one connection, streamlining the whole process.

If anybody is interested in running 0.9.0, and helping us test it, here's the executable

http://dl.dropbox.com/u/626683/nzbget-0....sel.tar.gz (for 200/300 series):
http://dl.dropbox.com/u/626683/nzbget-0....sel.tar.gz (for 100 series)

It's just a drop in replacement over the previous nzbget in your Apps directory (rename the previous one to nzbget.old or whatever, so you can revert easily).

Also, NZBget assumes categories now automatically from HTTP headers when downloading NZB files, so make sure you have AppendCategoryDir set to NO in your config, if you're using oversight also.

[Slevin]

I just used this on my A110.
I can see NZBget as it should and the server is started..

I'll try to report back in a few days and tell how the things got downloaded (dognzb-push)

[Slevin]

Push and downloads are working ok with the provided 0.9
tpo bad it is a bid hard with oversight integration..
i've made a request to change that part on oversight's side here. please vote it

[jasch]

Even after disabling it on nzbget?

[Slevin]

sorry for the late response..

in NZBget (0.9.0-testing-r414) AppendCategoryDir is set to "no".
but it still creates subfolders per category in the /share/Complete folder. (nzbget-bug?? )

i don't think the AppendCategoryDir has any effect on how oversight handles categories..
as it gets te category value passed trough directly from NZBget when it executes the unpacking/catalog-script.

i haven't heard any feedback yet from lordy/oversight on how to solve this.