[Pan-users] Re: Pan on windows?

[Duncan]

Sapient Fridge posted on Fri, 08 Apr 2011 08:46:14 +0100 as excerpted:

> 2) Is there any way of automatically downloading all the message bodies
> rather than just the headers?  My ISP's news service is a bit
> intermittent and my old newsreader (Turnpike) would download and cache
> everything (including the bodies) it could as soon as available so I
> rarely had to wait for anything  to arrive.  I realise I can select
> everything and hit "Cache Article" but is there any way of
> *automatically* downloading and caching new message bodies as soon as I
> open Pan?

There's no way of automatically downloading all bodies, unfortunately, A 
feature has been discussed that would work with the scoring categories, 
configurable to the score category for each, but with the idea that 
watched articles would be downloaded automatically, ignored would be 
deleted automatically, and negative-scored would be marked-read 
automatically.

But pan has just for practical purposes returned from the dead, as Charles 
Kerr, the former maintainer, lost interest, and few enough people do news 
any more that it just seemed nobody with the requisite developer skills 
was interested in pan.

But khaley came along and started at least gathering all the patches that 
had been proposed over the years together, plus a bit of extra stuff 
around the edges, and that git repo was what many of us regulars were 
running for some time, building from source.

khaley wasn't however interested in being an official gnome developer, I'm 
not sure the reason, so that still left pan without an official way 
forward within the gtk/gnome infrastructure at least.  Until petr kovar, 
not really a dev but already a gnome translator and pan user, 
volunteered.  So now it's khaley doing most of the development work and 
pkovar interfacing with gnome, and that's what got us the 0.134 release 
after years of little activity.

But it remains to be seen if khaley (or anyone else for that matter) is 
interested enough to code up any serious new features such as that, or 
whether it'll be mostly maintaining pan as is, mainly just keeping it 
building with current gtk, gcc, etc, plus opportunistic integration of any 
useful patches that do appear.

Of course I and others are hoping for the former and there's certainly 
features that could be added that would further improve pan, but in the 
FLOSS world it's the coders that decide where an app goes, and for those 
of us that aren't coders...

As for caching, be aware that for binary use at least, if you're going to 
be making much use of download to cache, you'll need to increase pan's 
default 10 MB cache size.  The GUI setting for this was removed as too 
complex and unnecessary for users that direct download anyway, but it 
remains an option exposed in the text file configuration itself.  So 
editing preferences.xml using a text editor is necessary to change the 
cache size.  It's not difficult to search for "cache" and find it, then 
carefully edit the setting so as not to screw up the xml formatting, and I 
can confirm that cache settings of into the double-digit gigs at least, 
work fine, as that's what I use here.  (For binary mode, which I actually 
don't do very often any more, I always download to cache first, then sort 
out what's there when it's all local and I don't have to wait for it to 
download.  I have a 12-gig dedicated partition that I use for that, and 
yes, it works fine, at least on Linux.)

> 3) How do I see the raw source text of an article?   I sometimes send in
> spam complaints but to do that you have to send in the full headers and
> I can't work out how to get them from Pan.

There's three ways.  If you're just wanting a quick look at the headers, 
use the show-all-headers toggle under view, body pane.  Here I have it 
mapped to the "h" key as the keyboard accel.  That used to be the default 
and I think it still is, but I'm not sure as I long ago customized my 
accels using the accels.txt facility.  But with that set, a quick "h" 
toggles view headers on and off, as necessary.

You can also use the save article function, and choose to save text, not 
binary (both is also an option).  This is at least /almost/ raw, and does 
indeed save the headers as well.  I believe the result here is identical 
to the file from the third option below, save for the name and location, 
but I'm not positive of that.

Finally, you can grab the message directly as it appears as a file in 
cache.  First, use the view-headers toggle to get the message-id.  That in 
turn is the name pan uses to save the file in the cache, naturally with 
various filesystem illegal characters fudged away into something 
filesystem legal.  The .msg extension is appended.  Once you have the 
message-id, then, it's possible to pull the message from cache directly, 
as long as it's still there, of course.  (With the 10 MB default cache it 
won't be for long when pan's actively downloading binaries.)

Caution!  At least if you decide to increase the cache size as mentioned 
above, you probably do NOT want to try loading the entire cache in a 
standard file manager!  You could be waiting awhile!  Consider that it's 
very possible you'll have ~100 K of cached files to deal with, EACH of 
which the file manager will try to scan enough of to assign an icon...  
Even trying to pick it out with the file open dialog isn't a good idea.  
Rather, open the file open dialog to the pan data dir if desired, then 
type in the article_cache directory and message-id.msg, and hit enter 
without ever attempting to load/show the whole cache dir in the filepicker!

-- 
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