Best way to download nzb files to then have files downloaded

[neurot]

I have a Windows 7 HTPC that has a 6TB RAID-5 system. This system has SABNZB+ on it, and I usually sit at a Mac. From there, I use Chrome and binsearch.info to find things I'm looking for. My download folder is the Mac's default folder (/user/downloads). On the Windows HTPC, I have SAB listening to that download directory for NZB files. However, the problem I have is when I download an actual nzb file. This temp nzb file goes in /user/downloads, is "consumed" by SAB, then disappears from the Mac (which is my indicator that SAB has picked it up), then it reappears in my completed folder on the HTPC (a:\SAB\complete). Since SAB is not monitoring that folder, it just sits there. I have to copy it back over to my Mac for it to be discovered again by SAB for the real download again.

To make matters worse, once I do drag it back over to my Mac, TWO copies of it start downloading. So something causes it to be seen twice. I could disable it downloading if it detects a duplicate file name, but since the file name is generated by the search results, the search results are very often the same. it would be nice if the resulting nzb files would have random or time generated file names to prevent that from tripping up the duplicate checker.

I suppose that instead of using the "create NZB" function, I could  just use the "download" link that is present in Newzleech and NZBindex, but is not present in binsearch (which is what I usually use). But is there another way around this?

[shypike]

You should not make the "watched folder" and the download folder the same.
Use a sub folder or just another share.

[neurot]

bump

let me make this more simple. if the search sites create nzb files, but they also sometimes show nzb files, what happens when you use them to select an nzb file? they create an nzb file, which then compiles the end file (yet another nzb file), but does not further process that nzb file recursively.

any solution to this problem?

[shypike]

SABnzbd already supports NZB chaining.
If the end result of a download is only one or more NZB file(s),
these get automatically send to the queue as new jobs.
(When other files are present, the NZB is not re-queued. The reason is that many NZBs
contain a full payload plus an NZB).

[neurot]

Then how should I have things set up? When I am surfing around on my Mac, I am saving the downloads locally. On the SAB system, it listens to that same folder, over the network. That is how I get my nzb's processed. Seems to me that the problem is that the output folder isn't getting listened to, so the resulting nzb isn't getting processed. I've been using SAB for several years, and have never experienced NZB chaining.

If I save my downloads to my Mac, then drag the nzb's somewhere else to a watch folder, i don't see how that will affect this issue at all. it seems more dependent on the output folder than the input folder. can you provide some more examples of folder structures and flow of where the files are generated as the SAB process happens?

One related question that may help - how do I get SAB to not create folders for every download? Perhaps if there were no subfolders created, then the output folder would be the same as the watched folder, and the listener would see that there is a new nzb file there and process it.

[shypike]

The watched folder supports only the main folder and any first level sub folder named after one of your categories.
It does not scan the complete folder hierarchy!
When a job that downloads only one or more NZBs is finished, these NZBs are sent to the queue and are removed.
At least that is the idea, however there are some bugs associated with this.
These will be solved in the next (minor) release.