[Pan-users] Re: Pan derived product

[Duncan]

-=[Miko]=- posted <address@hidden>, excerpted below,  on Sun, 14
Aug 2005 10:59:03 +0200:

> Hello world, I want to make a little project, but doing it only will be 
> very fastidious...
> I explain.
> 
> Spenting lot of time of grabbing articles with pan is very borring. So I 
> want to make an automated pan, wich grab all the articles of a group 
> every 5 minutes, and eventually doing an par2verify/par2repair when 
> neccessary.
> 
> So I have to remove all gui present in pan, and simplify it for command 
> line uses.
> I there are anybody interested by this little project, contact me at
> axel <dot> lc <at> free <dot> fr
> 
> Thanks a lot for reading me, and I hope you to contact me.
> 
> Best regards, Miko
> 
> (As you can see my english is not perfect, french users are bienvenue)

Your English gets the point across quite well, which is what matters. =8^)

This has been on the todo list for PAN for quite some time.  The idea has
always been a --background or --nogui or --daemon or --fetchonly mode of
some sort, that would either sit in the background, fetching periodically,
as you outlined, or be designed as a single-shot deal, that could be run
as a cron job or similar, if periodic checking and fetching was desired.

However, there were always other things to do first.  

Presently, PAN is undergoing a fairly major backend rework, ultimately, to
use the sqlite database library for its backend storage.  If you want the
feature now, go ahead and work on it.  However, be aware of the fact that
because the PAN backend is changing, future PAN GUI versions will quite
likely be incompatible with any fork at the present time. 

In any case, I'd recommend that you take a look at the pan-devel list.
It's not very heavy traffic, but there are several folks there working on
various PAN projects.  K. Haley is the one that's gotten the farthest on
an sqlite implementation, tho as of the last update, there were still
several portions of the project to go, and how close that implementation
will look to what Charles ultimately implements for mainline PAN is of
course not yet known at this point.  However, they are coordinating at
least to some extent, and some others do smaller stuff once in awhile.

I should also mention that others have voiced the opinion that a periodic
fetch is /exactly/ what a news server does.  If you desire that
functionality, why don't you just run a proper news server, then have PAN
connect to it locally, either over your LAN, or thru 127.0.0.1/localhost? 
There are a number of people on this list that do exactly that.  You then
run PAN itself with a small cache, since it's all stored locally on the
news server anyway, and access is essentially instantaneous.  No need to
go to all that work of rewriting PAN to remove the GUI and become what
amounts to a news server of sorts, when there are already apps available
that do the job quite well!

-- 
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 in
http://www.linuxdevcenter.com/pub/a/linux/2004/12/22/rms_interview.html