Page 1 of 1
Check files before downloading doesn't work
Posted: March 11th, 2013, 7:58 am
by harhar
I have these two options checked, and restarted the latest version of SabNZBd and it isn't working.
Abort jobs that cannot be completed
When during download it becomes clear that too much data is missing, abort the job
Check before download
Try to predict successful completion before actual download (slower!)
Repair failed, not enough repair blocks (230 short)
Repair failed, not enough repair blocks (126 short)
Am I doing something wrong? If the NZB was already in the que before I made the changes, do I have to re-add the NZB after I made the changes?
Re: Check files before downloading doesn't work
Posted: March 11th, 2013, 1:07 pm
by shypike
The first option is only useful when most of the content is missing.
"Check before download" does a more extensive check, but unfortunately it won't work
with all servers due to the way those servers deal with DMCA takedowns (i.e. replace with crap articles).
We're looking for a more reliable method for a future release.
Keep in mind that there will never be a magic method that will prevent all fringe cases.
Re: Check files before downloading doesn't work
Posted: March 11th, 2013, 4:46 pm
by harhar
shypike wrote:The first option is only useful when most of the content is missing.
"Check before download" does a more extensive check, but unfortunately it won't work
with all servers due to the way those servers deal with DMCA takedowns (i.e. replace with crap articles).
We're looking for a more reliable method for a future release.
Keep in mind that there will never be a magic method that will prevent all fringe cases.
Thanks for the reply... I heard SabNZBd had the capability of detecting the crap DMCA articles so it knows to use a backup server. I guess it can't use the same way to detect when it won't repair?
Re: Check files before downloading doesn't work
Posted: March 12th, 2013, 12:55 pm
by shypike
Currently we don't read article headers, because that would cost you bandwidth and possibly quota.
However, several servers (especially astraweb) do not remove articles but replace them with junk
and a few telling headers.
Using the "abort" option is the fastest way to detect removed posts, which will cost you relatively little bandwidth and quota.
Re: Check files before downloading doesn't work
Posted: March 13th, 2013, 6:05 am
by harhar
shypike wrote:Currently we don't read article headers, because that would cost you bandwidth and possibly quota.
However, several servers (especially astraweb) do not remove articles but replace them with junk
and a few telling headers.
Using the "abort" option is the fastest way to detect removed posts, which will cost you relatively little bandwidth and quota.
Will do, thanks!