In article <55a0oehjnl36hhb7ekh3d46qoecekkkqah@4ax.com>,
SL <SerpentLord@Evil.Incarn8> wrote:
> My question from the other day...how do "I" make an nzb file?
> I see Gollum sending them with his posts. He's not getting them
> after-the-fact from one of the nzb search engines; he's making them,
> but how?
That last part is for Gollum to answer. :->
An NZB file consists merely of re-statements of all of the message-IDs
of the parts of a binary post or series of posts, along with
re-statements of the "Newsgroups" and "From" and "Date" headers from
those posts. It's all formatted as a series of XML elements and put
into a text file (you can edit that file with any text editor).
Some of the more modern newsreaders have options for creating NZB files
from posts that you select when you have a group open, similar to
downloading except that you just end up with the .nzb rather than the
files themselves. That's for the situation where you want to give the
NZB to other folks so they can grab those same files. But if you know
about a Usenet post somewhere and you want an NZB for it so you can
download it, then obviously starting with your newsreader isn't any
help, because if you can find the post there, you can download it in
the newsreader, and there's no need for the NZB. That's why there are
the Usenet-indexing web sites, which are merely (in addition to the web
side, the user-facing side) passive NNTP servers that never store any
of the bodies of the Usenet articles that they receive; they only store
the headers. So only a tiny fraction of the disk space is required on
those servers, but they do need to be Usenet peers with good
connectivity, which is difficult in these days of corporate news
servers. So not just anybody can set up a Usenet-indexing server;
you need to know someone who knows someone, etc.
A long time ago I wrote a script (will work on any Unix/Linux/BSD or
Mac OS system but not on Winblows) to do that massaging of the data
into XML and create an .nzb file, given the headers of the binary posts
in a text file as input. However, there are two things wrong with it:
it's pretty closely tied to the Mac newsreader that I use (it reads
text files created by that program in a certain format), and I never
got around to developing it to the point where it could handle
multi-part posts! So it's useless for your purposes, sorry. However,
that might be irrelevant, because it's merely another example of what
I mentioned where the newsreader sees the posts complete on your server
and writes an NZB file for you to distribute. And maybe Agent can do
that for you already; I dunno. If Agent 7 can't, then maybe 8 can?
:->
|
|
){kind=link}