RSS feeds dont start

[Usenet]

Im probably blind or plain stupid, but I cant get the rss feeds to start downloading.
I have two feeds from nzbs.org . One is a personal feed where you can ad nzbs as you like, the other is based on search criteria.
In general settings, Ive set rss checking interval to 15 minutes.
In rss settings, I can preview both feeds and get an answer on matched items. I have not set scheduled.
I have after the set up of feeds added several nzbs to my feed but they dont get downloaded.

As I have understood it: scheduled will run one time a day according to the schedule. If not set the feed will be read every x minutes set in the general settings. The match of the feed in the first set up will filter out items in the feed so only new ones gets downloaded (after set up). right?

[shypike]

Automatic handling requires the "schedule" tickbox of the feed to be enabled (Config->RSS).
The interval can be set in Config->General (minimum is 15 minutes).
There  is just one interval setting, no "once a day".

The very first batch of a newly defined feed is NOT downloaded by the first scheduled scan.
This is to prevent a huge number of jobs to be downloaded the first time.
All new (and matching) entries in subsequent scans will be downloaded.

If you use the "Download" button, the "skip first batch" rule will be igonored.

[Usenet]

Aha, so "scheduled" means activated and has nothing to do with schedules ...

Some how I've managed to get all top model s11 episodes.. well at least the wife gets happy

Thanks for the quick answer

[shypike]

Aha, so "scheduled" means activated and has nothing to do with schedules ...

It has to do with schedules, because the scans are scheduled.
You just don't see them in Config->Scheduler.
OK, maybe I think too much like a software developer instead of a user 

[hal]

Hi Shypike
I have to admit that I was confused by this too.
Is there any way that this could be made more intuitive for the user? 

For example, on the Scheduling page, maybe you could add a block called "Auto-Scheduled RSS Feeds" that lists the Feeds ticked on the RSS page, and the time in minutes until the next auto-scheduled poll and download.

On the RSS page, you could change the "Scheduled" tick box to "Auto-scheduled", and put a note on the page that the user must set the "RSS checking interval" on the Config/General/Tuning page.


I am curious  - can you tell me where sabnzbd+ keeps the record of which RSS items it has downloaded (so that it doesn't download them multiple times), and how and when this record is reset?


Thanks, not least for a fantastic piece of software!
Hal

[hal]

Any responses to my questions would be very much appreciated. 

Many thanks
Hal

[switch]

Hi Shypike
I have to admit that I was confused by this too.
Is there any way that this could be made more intuitive for the user? 

For example, on the Scheduling page, maybe you could add a block called "Auto-Scheduled RSS Feeds" that lists the Feeds ticked on the RSS page, and the time in minutes until the next auto-scheduled poll and download.

On the RSS page, you could change the "Scheduled" tick box to "Auto-scheduled", and put a note on the page that the user must set the "RSS checking interval" on the Config/General/Tuning page.


I am curious  - can you tell me where sabnzbd+ keeps the record of which RSS items it has downloaded (so that it doesn't download them multiple times), and how and when this record is reset?


Thanks, not least for a fantastic piece of software!
Hal

I'll probably be changing the RSS layout for smpl to be more clear in 0.4.6/0.5, may extend that to the other skins too.

Will be either like this:


or move the checkbox and add a label like: 'Enable RSS checking" or "Enable Automatic Checking".

As for your other suggestions, they are currently more difficult and problematic to implement until we move to a database structure in storing certain information. The current RSS information is simply stored in a python dictionary that gets dumped to a file in the cache folder at program exit. It is a pythond ata structure and is not easily editable.

[hal]

Hi Switch, thank you for you last reply, and apologies for my late one! ;-)

If information cannot be moved around, then I think the best option is to make the wording of the labels more intuitive, like you have suggested, or like I suggested in my previous post...

"On the RSS page, you could change the "Scheduled" tick box to "Auto-scheduled", and put a note on the page that the user must set the "RSS checking interval" on the Config/General/Tuning page."

Pass on my gratitude to the rest of your great team.

Best wishes
Hal

[hal]

The current RSS information is simply stored in a python dictionary that gets dumped to a file in the cache folder at program exit. It is a pythond ata structure and is not easily editable.

Regarding the above response, would you mind clarifying something, please?

When do the records expire of all the downloads retrieved via RSS?
Alternatively, are they lost only if the cache is manually deleted?

Many thanks
Hal

[switch]

The current RSS information is simply stored in a python dictionary that gets dumped to a file in the cache folder at program exit. It is a pythond ata structure and is not easily editable.

Regarding the above response, would you mind clarifying something, please?

When do the records expire of all the downloads retrieved via RSS?
Alternatively, are they lost only if the cache is manually deleted?

Many thanks
Hal

I recently looked further into this as I didn't write the code in question and was getting a lot of duplicates myself.

The only thing recorded are the current items in the queue and if they have been matched or not, once the item falls off the RSS feed it is deleted from the records and could be added to the queue if it enters the RSS feed again; so there is no proper duplicate detection in sabnzbd yet.

0.4.6 will included something that could be useful, enable the nzb backup folder (will create an nzb for every item in the queue) and you can disallow adding items to the queue that the nzb already exists in the nzb backup folder.

However for 0.5 I am looking into adding proper duplicate detection that will hopefully be easily editable.

[hal]

Switch
I have only recently started using the RSS facility, and am finding it quite a revelation - in other words, very useful indeed! ;-)
I hope you understand, the reason for my questions was simply to understand the way SABnzbd RSS feeds work, and certainly not to complain.

I understand there are currently some slight oversights with the way RSS feeds work, but on the whole it works quite successfully.
Having said that, the changes you propose would probably make it more robust.

Thank you for explaining the feed matching process, as this is much clearer now.  Can you tell me, how long do the RSS feed items and their respective matched/unmatched flags last?  Will they persist across sabnzbd restarts?

Thanks very much
Hal

[switch]

They will remain for as long as they are in the RSS feed that is fetched, once not listed in the feed they are removed from the list.
And yes, it persists throughout restarts.

[hal]

That's great, Switch - thank you very much once again!