Version 6.70 Beta 5 Available...

[Quade]

Found that if i deleted all posts from one poster and then clicked on 'clear the search entry' the deleted posts would pop back up,when i shut down newsbin pro and started it up gain those posts were gone.

I think I fixed this. They really are getting deleted from storage but they stay in memory till you reload the group.

[Teddy]

Most recently, NewsBin automatically makes backup, Einstellung.nbi.bak, and Einstellung.nbi.old. Can hang it so together.

Yeah when it saves, it first write the new NBI file to the "nbi.old" then when it finishes the write, it copies the old one to ".bak" then copy's the new one to the real name.

I doubt that has anything to do with this issue. The change means that in a crash, if Newsbin is in the middle of saving the NBI you still have it original NBI in place when you restart.

Ok I figured out what's going on. "Save As" wasn't actually saving the NBI, it was just setting the new path so, parts of the program were unhappy that I was setting a path for a file that didn't exist yet.

Thanks for the report. Kinda surprised people are actually using "Save As". I've fixed it. It'll be in the next beta. You can manually copy the NBI file for now using explorer.

Thanks, It works again in the 6.70 beta 6

[majestyx]

Look at the size of the import folder, multiply it by 10 and that's how much data was downloaded from the news server. Depending on the groups, it's a bunch of data. I have about 200 groups and even though my machine is pretty powerful it takes 10-15 minutes to import them all.

I could import them faster by say doing 2 GZ's at a time but I think many PC's would groan under the load. One thing you can do it set Newsbin to download headers every 30 minutes to an hour. Then there's never a huge lump of headers to process in one go.

As a follow-up to my previous query, I've noticed in groups with huge post counts that while downloading headers, the Import folder creates what I'm guessing is the queue of headers it is downloading, usually storing them in 25-35 MB gz files. However, every so often it creates a 1 KB file and once the program gets to this file, it really grinds to a halt with a ton of disc activity, spending at least 15 minutes and sometimes more than 30 minutes doing something in the background once it gets to that 1 KB file. Is this by design? This morning I finished downloading headers from a group that had almost 900 million of them. It then took the program more than 6 hours to place them all into the group's spool .DB file, each time stalling as I've described each time it encounters a 1 KB file. Does this sound correct, and if not, any idea what's causing this? Does anyone else have this issue?

I know there's a new beta, but just wanted to post here since I haven't installed Beta 6 yet. I do know it does the same thing on the current release version, which is why I decided to install Beta 5, hoping this would clear it up.

[Quade]

This morning I finished downloading headers from a group that had almost 900 million of them.

900 million * 350 bytes = 315,000,000,000

It's about 315 gigs of data Newsbin has to chunk through. Even though it's compressed, Newsbin still needs to decompress it, process it, then recompress parts of it to place in the DB3 files. I don't consider 6 hours to be that bad to be honest.

I've never seen a 1KB file stall. Not sure how I'd reproduce it.

[majestyx]

This morning I finished downloading headers from a group that had almost 900 million of them.

900 million * 350 bytes = 315,000,000,000

It's about 315 gigs of data Newsbin has to chunk through. Even though it's compressed, Newsbin still needs to decompress it, process it, then recompress parts of it to place in the DB3 files. I don't consider 6 hours to be that bad to be honest.

I've never seen a 1KB file stall. Not sure how I'd reproduce it.

Okay, thanks for explaining the math. Now that I see how much space that is uncompressed it makes a lot more sense as far as the time is concerned. As for the 1KB files... it seems pretty random as to how many or how often it creates that size of file. Just downloaded 150 million more and it went through them much more quickly, but I didn't see ANY 1KB files. Not sure if it creates a file that small due to a "hiccup" in my internet service, which isn't the best where I live, or something else.

Anyway, thanks again for your response and for continuing to work on Newsbin after all these years. I'm a very happy customer!

[Quade]

What security software are you using? Kaspersky is known to interfere with compressed header downloads.