[Pan-users] Re: a little something for the pan heads

[walt]

On Tue, 23 Jun 2009 17:14:53 -0600, K. Haley wrote:

> walt wrote:
>> WFM too, thanks.  If you're still bored, there is an obvious bug in how
>> pan marks articles as read/unread after fetching new headers, but no
>> one has so far been able to identify the buggy code.
>>   
>>   
> Well, I looked through the code and couldn't find anything wrong.  I did
> a little cleanup on the newsrc code but I don't think that will fix your
> problem.  Since pan uses the newsrc files to track read articles and I'm
> not seeing the problem, the only thing I can think of is that the
> article numbers are changing on the server or perhaps out of sync round
> robin servers.

Good work with your gmime updates, thanks.

I'm replying to this post to explain that my "obvious" read/unread remark
was too glib, sorry.  I don't see the problem in text groups like this one.
I see it definitely in multipart binaries groups but I can't define how or
when it occurs.

I visit binaries groups rarely, so the bug may be triggered by articles
that have expired from the server but still exist in my local cache.

Or maybe my NSP's European servers are not actually using the same article
numbers as their U.S. servers as they claim.

Dunno, just tossing out ideas, but I've learned over the months to ignore
pan's read/unread flags as unreliable.

Very much appreciate your work on pan, thanks!