[Pan-users] Re: Rule to block crossposts

[Duncan]

Daniel Mahoney <address@hidden> posted
address@hidden, excerpted below, on  Fri, 22 Aug 2008
12:32:48 -0500:

> 0.14.2.91. I didn't even know there was a Pan2. But I'm now on the web
> page and I see that, and I'm getting it via svn now. Thanks.

You're talking about checking the source code and you haven't even 
checked the site, very clearly available directly from pan itself, to see 
if you're using the latest version?  I take it you /like/ beating your 
head against walls other people have already punched thru, duplicating 
code, etc?

More seriously, I suppose you know to check, /now/, but anyway, do 
realize that new-pan is a rewrite, and does some things differently.  
It'll take a bit of getting used to if you're used to old-pan, but it's 
VASTLY better scaling in large groups (try old-pan on more than a couple 
hundred thousand headers in a group and see how long it takes to digest 
them... new-pan handles millions), and is fully automated multi-server 
now.

The multi-server rework in particular changed the GUI layout and methods 
somewhat.  But keep in mind that's what it's designed for now, and many 
of the changes make sense.  Meanwhile, there's a number of settings that 
aren't available (or are but at significantly less detail) in the GUI as 
Charles wanted to simplify it some, but a power user will find a lot of 
tweaking can be done by directly editing the config files or similar, if 
desired.

The biggest thing missing in new-pan is the rules set.  There's no way to 
automate downloading watched posts, or deleting ignored posts, for 
instance as there was in the old version.  Other than that, most features 
in old-pan are there in some form, if hidden a bit.  Ask if you get stuck.

One nice new feature is that pan now honors the PAN_HOME environmental 
variable, which can point it at a default location other than ~/.pan2.  I 
combine that with starter scripts that set it, to run multiple pan 
instances, each with its own settings.  Here, I have a binaries instance, 
a text instance, and a test instance, but of course you can set them up 
as desired.  Cache size is now set by editing preferences.xml only, and 
the number of connections per server can be set beyond the four allowed 
by the GUI (due to GNKSA) by editing servers.xml.  That's a sampling of a 
few of the "advanced" options now available to those willing to edit the 
config files directly.

Conversion... There's very little of your old-pan settings and data that 
can be used with new-pan, due to the changes in format that went along 
with the scaling improvements.  There are, however, two exceptions.

The scorefile is almost the same -- except it's stricter toward SLRN 
style.  Old-pan allowed the newsgroup/section entries to be in regex, 
while slrn and new-pan only take * wildcard format newsgroup/section 
entries.  That may mean a bit of hand tuning on your score file if you 
used newsgroup regexes before.  The major differences now are that pan's 
scorefile is still case insensitive by default, and that pan doesn't yet 
implement some of the more advanced stuff (includes, etc) that slrn does.

And, the actual article cache should be transferable, if you like me run 
a multi-gig cache.  The overviews/headers aren't -- they must be 
redownloaded after setting up the servers again -- but if you use the 
same cache, the articles themselves shouldn't need redownloaded, only the 
overviews/headers.

That should get you started.

-- 
Duncan - List replies preferred.   No HTML msgs.
"Every nonfree program has a lord, a master --
and if you use the program, he is your master."  Richard Stallman