[Pan-users] Re: Feature request: CLI operation

[Duncan]

Ron Johnson posted on Sun, 27 Mar 2011 15:49:53 -0500 as excerpted:

> The GUI interface is great, but not so optimized for downloading the
> headers of big groups.
> 
> So, I request a separate program pan-get-headers (which is part of the
> pan2 source tree and thus gets built when you run "make") that uses the
> already-existing pan nntp functions.
> 
> It would work something like:
> $ pan-get-headers --group=alt.really.big --all 
> $ pan-get-headers --group=alt.really.big --new
> $ pan-get-headers --group=alt.really.big --days=nn
> $ pan-get-headers --group=alt.really.big --latest=nn

While I haven't looked at it in awhile nor really ever used it myself, pan 
already has a "gui-less" CLI interface, that specifically /does/ have 
header-fetch functionality.  Try running pan --help in a terminal window 
and see if it has what you're looking for. 

A quick-look says it fetches new headers (no options for the others).  
Good for a continuing cronjob or the like, to keep up with a group so when 
running pan normally it's already up2date, but not so much for 
initializing a big group.

Unfortunately, there's presently no mechanism for getting the new headers, 
scoring, then caching full content for messages scoring at/above a 
particular category (probably watched by default), all automatically, 
based on CLI options.  It's a feature that has long been requested...  
Imagine being able to cron-job all that, so you can load up pan and have 
everything already cached, ready to be examined and saved or deleted. =:^)

Of course, that would require setting the cache size rather larger than 
the default 10 MB, something that's only possible by direct-editing the 
config file at this point, as Charles thought a cache-size setting was too 
complex for normal users (resulting in the gnomification/dumbification of 
pan with the rewrite, losing that option from the GUI), but setting a huge 
cache (12 gig dedicated partition for my binary pan instance), downloading 
headers, doing an initial delete of what I /knew/ I didn't want, then 
downloading to cache and doing something else while it did so, then coming 
back to sort thru what's downloaded when I didn't have to wait for it, is 
how I've /always/ worked with binaries.  If that could be automated via CLI 
options and cronjob so all I had to do was sort thru the pre-cached 
messages, I'd be a happy camper indeed. =:^)  Unfortunately... =:^(

-- 
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