newsbin.com

Had a situation today which would have been a lot easier to handle in previous versions...

Loaded a .rar set (no pars) for a video file. Downloaded fine, but a couple of files were corrupt, so it wouldn't unrar. Found a separate post of the same .rar set, but with pars added in. Queued that up, and it failed the whole thing because it had already downloaded those .rars, instead of downloading the pars which were brand new.

I had to pause downloading, re-queue the file, bypass filters, expand the set, pause the rars and un-pause the pars, and then un-pause downloading. That's a lot of manipulation to get a few par files.

In the old version, the pars would have automatically downloaded, as they were new files even if the rars weren't. Is there a way to have that behaviour in the new version?

Thanks,
thwartme

You did not have to queue the whole set. You could have opened up the set in the Post list, and selected just the files you wanted, and queued just them.

It wasn't a post list, it was an nzb file. I would have had to open the nzb file in a text editor, and manually edit out the bits I didn't want. Again, a far more cumbersome solution than having Newsbin simply recognize files it has already downloaded vs files it hasn't, like it used to.

Naw, load the NZB using the "Load NZB" button. It's the first one on the toolbar. Then you can manipulate it just like it WAS headers.

If Newsbin's got the PAR file, it shouldn't re-download the already downloaded files anyway. It'll check them by downloading the first chunk and then making sure it's the file the PAR file mentions. You might want to be using B6 though. It works best for this. I use that for sampling. Add a couple files to download, unrar and sample, then add the rest of the set.

You might want to disable the duplicate checker. It's in the advanced options. It's not very useful for rar files and gets in the way sometimes.

With a NZB file (which I use all the time) there are basically two choices:

• load the NZB via the Load NZB toolbar or menu items. You can then select individual items to queue (opening up compact items if necessary to select specific items). This is the most efficient in terms of download volume.
  
• load the whole NZB into the download list making sure the same folder will be used for the files as the previous time. If you have signature checking active use the "bypass filters" option. NB will download the first chunk of each file, and if it is already on disk Mark the file as complete and move onto the next. This is the least effort but does mean that the first chunk of each file gets redownloaded which may be an issue for some people.

This second option was not possible in NB5 as it would cause all files to be downloaded again in their entirety.

Okay, I understand that there are ways to make it work under the new version. I detailed one myself, in the OP. My issue is that in the old version, it was much easier. I dropped an nzb into newsbin's download list, and the program did all the work. Now, it doesn't. This is a step backwards, not an improvement, in my opinion.

Thank you,
thwartme

If 6 is better 95% of the time and you have to occasionally jump through hoops for messed up posts. The question becomes how much effort should I put into these rare cases. I haven't said I won't let you delete individual ones in the future. The problem is that it'll likely break autopar which means it'll probably be worse then jumping through hoops in the first place. It's a trade-off.

Something I'll have to think about. Newsbin is designed so, that you can just dump the entire new set in and it'll figure out what needs to be downloaded for you. As long as you use the same download folder.

At the moment the behavior seems to be that as soon as any file in the set gets caught by the signature match, the whole set gets rejected as a duplicate (unless you have used bypass filters) so you do not get to the point at which there may be new files in the set if it is at the end. I suspect the problem raised by the OP would not happen if each file in the set was tested as being a duplicate so that you could get past any duplicates and download any new files.

itimpi wrote:At the moment the behavior seems to be that as soon as any file in the set gets caught by the signature match, the whole set gets rejected as a duplicate (unless you have used bypass filters) so you do not get to the point at which there may be new files in the set if it is at the end. I suspect the problem raised by the OP would not happen if each file in the set was tested as being a duplicate so that you could get past any duplicates and download any new files.

Right, that'd be great.

You might want to disable the duplicate checker. It's in the advanced options. It's not very useful for rar files and gets in the way sometimes.

Like I said...

Quade wrote:

You might want to disable the duplicate checker. It's in the advanced options. It's not very useful for rar files and gets in the way sometimes.

Like I said...

I DO want Newsbin to avoid downloading duplicate files. I DON'T want Newsbin to flag as duplicate files which it hasn't yet downloaded.

You're right though, this is a relatively rare situation, and if you don't think it's worth the time and effort of amending the code, that is of course your call.

Between this and not being able to do the partial sorts on a download list like I used to do, I may have to revert to an earlier version.

Thanks,
thwartme

I simply re-queue in these (as you say rare) conditions the whole post with by-pass filters and then get the behaviour you want. That works as if the file already exists on disk NB does not redownloaded the whole file again even when by-pass filters is used.