Purge History problem

[Ramrodian]

I'm using version 0.4.12 with Plush.

I've have 6 downloads that seems to stuck into the history while that flowerthingy is going round and round and round and not finish in a star. (hope you understand).
So now all other downloads that get finished is stuck into that loop too. And are remaining in the Incomplete folder and not moved, for example to the TV folder.

Can I fix it somehow?
I've tried to purge the history, all before those 6 downloads disappeared, but these remained.
I don't want to start the software with -clean because I have a queue that's 120GB now.

Edit: It can be because SAB is trying to make a folder that has too long name. But shouldn't it just stop and move on?
Because it gives a warning about that SAB can't make those folders.

Also when I shutdown SAB it says an error occurred and it created a log file in the install folder named SABnzbd.exe.log.
The Log says:
Exception in thread Thread-1:
Traceback (most recent call last):
 File "threading.pyo", line 488, in __bootstrap_inner
 File "sabnzbd\postproc.pyo", line 217, in run
AttributeError: 'NoneType' object has no attribute 'replace'

[Ramrodian]

Try shutting down sabnzbd and starting it up again.

I have done that several times, that's when I get that error message and that log file entry.
The only thing I haven't tried is to shut it down, restart the computer.
I forgot to say that I'm on Windows XP SP3.

[shypike]

Somehow the queue administration got corrupted.
Probably the only way to solve it, is to clear out the queue.
This will mean that you loose History and all that is now in the download queue
and the post-proc queue.

You may be able to salvage the jobs pending post-processing
by looking in the "incomplete" folder and manually do the Par2 check
and unpack.

Start up SABnzbd with the --clean parameter once to clean the queue.

[Ramrodian]

Somehow the queue administration got corrupted.
Probably the only way to solve it, is to clear out the queue.
This will mean that you loose History and all that is now in the download queue
and the post-proc queue.

You may be able to salvage the jobs pending post-processing
by looking in the "incomplete" folder and manually do the Par2 check
and unpack.

Start up SABnzbd with the --clean parameter once to clean the queue.

Okay, as I was suspecting. I already manually do par2 check and unpack, because it will not load my "download" computer as much, because everything getting moved to my main computer anyway.

I'll finish the queue and do a -clean run after that.

[shypike]

The release after 0.5.0 will have data integrity as highest priority.
There's a fundamental problem in the way the queue is organized
that makes queue corruption possible in some scenarios.

The least we will do is make "stuck" jobs removable.