@Ron, yep, I should have said that authorisation is done by domain so you don't need uid/password as you do with the Giga service. Glad you found it out anyway.Further Notes - How to use nzb files and avoid downloading headers.
So you've got Newsbin or an equivalent. I haven't tried Newsleecher so I can't advise on its use but it supports nzb files so the following notes are also relevant.
nzb files are indexes of binary posts that Newsbin can recognise with a plugin. If you download a nzb file, which is quite small, run the plugin then the list of binary files from the post is injected into your header list so you can select it for download. Some groups are huge so this can save an awful lot of header downloads and searching.
You can get nzb files in two ways: the first is simple, they get posted in a newsgroup which can be useful but you still have to find and download it. One or two groups have associated nzb groups. alt.binaries.ftn contains millions of headers but there is also alt.binaries.ftn.nzb which contains nzb files that point to posts in a.b.ftn. The second is to visit a newsgroup listing website. This is a good one (thanks Isis! ) http://www.binsearch.info/ With the latter you can enter a search string for a particular post, it will display any hits it finds from which you can select a post, then generate a nzb file for use in Newsbin.
OK, how about an example? Assuming you have set up Newsbin and tried doing a few downloads. Go here and check out new releases: www.nforce.com Assume you're interested in new movie release "Be Cool".
Visit www.binsearch.info.
Enter the search string "cool". There will be several hits. One is for an SVCD copy and another for a full DVD-R. Look at the DVD-R post. It has a note "Collection". This means that binsearch has collated each individual post and lumped them together for you. You can simply click on the tickbox, select "create NZB" and the NZB file will be built for you.
At the time of writing this there is also a "parts available" number in red which means that the post is currently incomplete. You can click on the "collection" word and get a list of posts then select individual post components beore creating the NZB file. then you can go on downloading the files posted so far then go back later to Binsearch and get the rest once they're posted.
Anyway, you now have the nzb file, the rest is easy. Start Newsbin. Click on 'File' Select 'Load plugin data' Select your newly downloaded nzb file. This will punch the items into Newsbin's post list. Select the items for download.
But supposing the post is incomplete? You have broken headers or even missing components altogether? All is not lost as if the poster has been kind, he/she will have posted par or par2 files. More on this later.
(eccles @ 13 May 2005,12:35pm)
because of the way it [text server] works, it's possible to direct it to the binary servers and get full bandwidth
Interesting comment Curls. How's it done?
Understand that I only know how to do this with Newsleecher, so you have to consider whether it's worth learning a new news client. I believe it is also possible to do the same with Newsbin Pro and BNR2, but I had no success with either, but that could be down to me not knowing the programs so well. Also, I use version 2.1 of Newsleecher and it may well not have the same setting in later versions.
Anyways, it really is simple to do, except there's no help files for Newsleecher, but I can help out there if anyone has any problems.
Fill in the server details for both Giganews and UUnet in the Usenet Manager tab. The UUnet one, in case you don't know, is news-text.dsl.pipex.com. Edit the servers so that you have two connections to both and set the priority of the UUNet server to 1 and the Giganews one to 2. If you don't think you'll use up your 10 gig allocation on Giganews, you can set both priorities to 1. Tick 'Server Requires Login' and fill in username and password for Giganews, leaving the UUnet one blank. In the 'Server Options' tab, tick the top 3 and the bottom one, leaving the middle one blank. Grab a list of newsgroups from Giganews and your finished. Happy downloading.
Oh, one other thing, in the binary groups on Giganews, there can be millions of headers. Consider that it just took me 45 minutes to get 150,000 headers and you begin to realise the problems. You can overcome this in Headers setting in Options by setting a more realistic figure for first time update. I have mine set to 20,000. Once you have got headers in all the ng's you subscribe to, you can then tick Get All Headers.
BTW, that 150,000 headers is only about 7 meg, which shows how slow Giganews really is. Actually, what it really shows is how little bandwidth Pipex bought from them.Thanks for the update Curls, but I really don't understand how that can work. I assumed you would be using some form of socks tunnelling to direct the uunet binary servers through the uunet text servers. But all your setup seems to do is put the text server as first priority and the giga server as second priority. But are you saying that by giving a binary header list to the text servers they can pick up binary content? Incidently, with Newsbin you can tick all servers but one to be slave servers. This means that they are used to get content only and the header list from the unticked server is used by all slaves. This is why I like Astraweb as it can be used to get headers without impinging on your d/l allowance and is quite quick in doing so, much quicker than Giga.Eccles, I don't even begin to understand it, I just know it works. I still have the post from August last year that put me on to it. It's too long to post here, but if you want to PM me an email address, I'll send you a copy. The priorities have nothing to do with it. It's the 'Send group Command' that does the trick. This command isn't available in most news clients as far as I know.
Edit: I just tried it with Newsbin but it won't get binaries from the text server. I've cut and pasted the relevant parts below, from the post I got the info from. There's more, but it's mostly waffle...
A typical news server will create an overview database for each newsgroup that allows XOVER responses to be generated very quickly. The overview information may be indexed by ARTNUM, although modern news server software often uses its own internal IDs for articles. But the whole point of the overview database is to facilitate fast retrieval of overview information (certainly for XOVER commands and more often than not for XHDR and XPAT as well, at least for fields in the overview database) and retrieval by ARTNUM.
Retrieval by MID is a different matter. If a GROUP command has been issued before a retrieval by MID then some news servers may use the overview database to try and locate the article, but this is not usually that fast since the overview database isn't typically indexed by MID. The more usual method of finding the article from its MID is to use the history database (which the news server needs in order to check whether it already has articles offered by peers - this is indexed by MID and needs to be fast in order to handle the incoming feeds). Article lookups via the history database aren't dependent on whether a GROUP command has been issued or not (incoming feeds from peers aren't handled on a per group basis, so the history database isn.t structured that way).
So in the context of Pipex's setup, what is easy to deduce is that
a) the front-end readers (what you see as news.dsl.pipex.com) don.t have a means of retrieving articles by <message-id> - if they did (e.g. if they used the overview database) then the Xref header would be there.
b) the back-end servers don.t know about article numbers (and hence about groups) - if they did then the Xref header would be in posts retrieved by MID.
So what you may be asking. Well what do you think happens if you request a message that was posted to a binary group by MID but from news-text.dsl.pipex.com rather than news.dsl.pipex.com?
The answer, is perhaps, obvious by now. news-text.dsl.pipex.com runs the same software as news.dsl.pipex.com, so it doesn't have the ability to locate a message by MID and so it passes it to the same back-end(s) - and hey presto you get the post.
news-text.dsl.pipex.com is, of course, not overloaded (or wasn't until I posted this) so there are no delays and download speeds on a single connection to the news server will easily max out ADSL 500, no bother.
But there is more -
What do you really think the difference between news.dsl.pipex.com and news-text.dsl.pipex.com is? Well, since I've probably now for all practical purposes destroyed news-text.dsl.pipex.com I may as well tell you.
There is only one difference and that is that the alt.binaries.* groups have been removed from the active file on news-text.dsl.pipex.com (I'm not going to even and try and explain everything about how a news server works, but the active file is effectively just a list of the news groups with high and low watermarks, plus flags relating to whether the group is read-only, moderated, etc). No doubt the significance of that escapes you as well, but what it means is that although you can't switch to an alt.binaries.* group on news-text.dsl.pipex.com using a GROUP command (because a GROUP command needs to consult the active file to get the high and low watermarks), other NNTP commands (depending on the server software) may still work. As it happens Pipex's news server software (which IMO is pretty crap) does allow you to change to a binary group on news-text.dsl.pipex.com using the LISTGROUP command - and having done that you can then use XOVER to get the 'headers' (at lightening speed), retrieve articles by ARTNUM or do whatever else you could do on news.dsl.pipex.com. The LISTGROUP command is, of course, absurdly inefficient and should have been removed from the NNTP specification years ago - but what the hell, news-text.dsl.pipex.com is (or rather was) lightly loaded, and if Pipex choose to support the LISTGROUP command, they can't complain when someone uses it.Well, that told me didn't it? Thanks for that Curls - I must give Newsleecher a try after all that.Here is the concluding part of my little mini-tutorial on how to download from newsgroups.
PAR and PAR2
OK so you've downloaded a game or a movie or a music cd and you have a load of files suffixed zip or rar or ace. First you have to unpack them. If you don't have WinRar then do a google and find it and download it. It's free but you have to pay for it if you want to get rid of the nag when you start it. WinRar will unpack all the popular compression formats so it's the only decompression utility that you need.
So you've kicked off WinRar only to find that one or more of your files is missing or corrupted. All is not lost. Did the poster also post par or par2 files? If they did, you're in luck. Go here http://www.quickpar.org.uk/ and get QuickPar and install it. When asked if you want to associate QuickPar with par and par2 files reply yes.
PAR and PAR2 makes use of similar coding techniques that are used with striping in RAID systems. Think of a PAR2 file as a stripe sector on a RAID disk. If the disk goes down the stripes can be used to recover the broken disk.
OK, so now look back at the newsgroup headers and hunt for a file called xxxx.par or xxxx.par2. Download it then double click on it. Quickpar should automatically kick off and scan your files highlighting which ones are bad or missing. If the file was called xxxx.par then you will have to download a par file suffix pxxx (where xxx are variable numbers) for every file that is missing or broken. Par2 is more sophisticated - even a broken file can have useable blocks and Par2 can use them when doing a repair. Clicking on the Par2 file will again kick off QuickPar which will scan your files and tell you how many blocks are needed. You merely have to download enough par2 component files to fill the blocks. The files are labeled volume xx + nn where xx is the volume number, and more important nn is the number of repair blocks the file contains. There will usually be some small ones with one or two blocks increasing to several large ones comprising many blocks. just download enough to add up to the number of blocks that QuickPar has reported that you need.
Once you have your repair files (stored with the main download), click on the par/par2 file again to kick off QuickPar and it'll then ask you if you want to repair your download. Click yes and it'll fix it. It may take a few minutes.
You can then go back to WinRar and unpack the whole thing.
Pretty good huh?
As always, you should be careful of viruses with your downloaded files, but at this point in time most newsgroups are surprisingly virus-free, far more so than P2P systems such as Kazaa.
And that concludes the little tutorial. It's by no means conclusive but it should give you a flavour for what to do when downloading from binary groups.