Being a backseat reader(troller if you will) I basically will D/L more
than U/L but am in the process of learning that monster soon(!)...
My problem is every time I open said group I cannot retrieve before a
certain date. I would like to retrieve articles(cds) posted prior to
when the slider will let me go(?)
for instance,when Iopen the group and go to the very "start" position
it always goes to the same date. I know this is basic for most of you so
be kind,for some reason I just dont see how to reset the starting point
so I can retrieve something from an earlier date.
Thankyou for the guidance,knowledge and criticism,just cant get over
this hump?...
--
Posted via NewsDemon.com - Premium Uncensored Newsgroup Service
------->>>>>>http://www.NewsDemon.com<<<<<<------
Unlimited Access, Anonymous Accounts, Uncensored Broadband Access
XNews does have a limit on the number of headers it can retrieve from a
newsgroup. I have seen occasion on busy newsgroups (such as the one you cite)
where the NSP does have the posts on the server but Xnews is unable to
retrieve them.
I know of no way around this limitation.
Wow...Talk about quick response!!!
Thanks for the information and advice,I'll go back to the requests group
and hope some kind soul will repost....
Not for nothing but why would a company state they have a retention of
XXX days....is it the newsgroup that cannot retrieve to that date or is
it a gimmick to lure you in?...Thanks again for the quick response(help)
The problem is with Xnews, unfortunately. I'd bet other newsreaders can
retrieve them without a problem.
I have yet to find a newsreader I like even a quarter as much as I like
Xnews... so I've started using Grabit and NZB files for downloading from
the busy newsgroups.
I certainly hope the OP is not taking what you wrote in this
thread as actual advice. No newsreader has any influence on the
retention time set up on a server. The OP described the
situation unambiguously and it is a matter of retention.
> I have yet to find a newsreader I like even a quarter as much as I like
> Xnews... so I've started using Grabit and NZB files for downloading from
> the busy newsgroups.
>
Mike "Xnews was not meant as a binary d/l tool" Yetto
--
In theory, theory and practice are the same.
In practice they are not.
> Bada bing JimH <n...@nunya.biz> bada bang:
>> The problem is with Xnews, unfortunately. I'd bet other newsreaders can
>> retrieve them without a problem.
>>
>
> I certainly hope the OP is not taking what you wrote in this
> thread as actual advice. No newsreader has any influence on the
> retention time set up on a server. The OP described the
> situation unambiguously and it is a matter of retention.
>
>> I have yet to find a newsreader I like even a quarter as much as I like
>> Xnews... so I've started using Grabit and NZB files for downloading from
>> the busy newsgroups.
>>
>
> Mike "Xnews was not meant as a binary d/l tool" Yetto
Well... yes and no.
If the post isn't on the server, the news reader won't be able to retrieve
it.
However... Xnews has a maximum amount that it can subtract from the most
current message number. If the desired message number precedes this, Xnews
will not be able to retrieve it on ANY server.
I had to double check the exact amount... but Xnews is limited to retriving
only the 50,000,000 most recent posts in a newsgroup regardless of the
server retention.
If server's retention is such that a newsgroup has less than 50 million
posts in on it, you can retireve them all. If it has over 50 million, you
cannot get them all.
> For uploading, you want PowerPost. It's simply the best tool you could want
> for the task. http://powerpost.cjb.net/
I've never looked at Powerposter but I'm guessing it also makes a pretty
good bulk uploader for spammers. My filter logs contain thousands of
(non-binary) sex spam messages every day with the X-Newsposter header
that seems to be a characteristic of Powerposter.
PowerPost is not for uploading single spam posts to multiple groups. It's
designed for uploading multipart binaries to one or several groups at a
time. Those ubiquitous sex spam binaries disguised as .avi files are
generally single files, not multipart posts.
--
Martin S.
It would take forever to scan 50 million headers with Xnews, a quarter
million or so at a time. (teranews shows close to 319M headers on
boneless, but you're right, Open Special with Xnews shows only 50M.)
If server retention is the problem, Giganews has the longest. Most
servers have a longer retention on text posts than on binaries.
--
Martin S.
The trick is to pay attention to the Article Number.
When Xnews opens a server, it queries the server as to the status of
your subscribed groups. The Group window shows the results of this
query. The "First" column lists the first (earliest) article number
that the server reports as being presest. The "Last" column lists the
last (latest) article number that the server has. When you open the
group (with sliders if necessary) and you find that you need an article
that is earlier than the "First" article number, that is an indication
that your news server has purged that article and it is beyond the
retension of that news server. Your only recourse is to find another
news server with better retension. This is not an Xnews problem.
Xnews does have a limit on how many headers you can download at once,
but even groups with huge numbers of articles can be comfortably viewed
by downloading small subsets of headers at any one time (incremental
retrieval) and purging before downloading the next subset.
HTH,
John
> When you open the group (with sliders if necessary) and you find that you
> need an article that is earlier than the "First" article number, that is an
> indication that your news server has purged that article and it is beyond
> the retension of that news server. Your only recourse is to find another
> news server with better retension. This is not an Xnews problem.
>
This is false when the number of articles exceeds 50,000,000. I have proved
this to myself time and time again with NZB files. There have been dozens of
occasions where I wanted to download a file, but Xnews would not retrieve it.
I went to binsearch and created an NZB file and downloaded it just fine with
Grabit.
Same server, same retention. Xnews wouldn't access it, Grabit had no problem.
On a very basic level, your statements do not make sense -- or at least
detail is missing such that I don't understand what you are trying to
say.
At the lowest RFC-3977 level, articles can be retrieved by either
Article number or Message-ID. Usually with Xnews this is done by
Article number which requires the header to be retrieved first in order
to identify the article number.
The total number of articles should make no difference to Xnews except
that
1) there is a limit as to how many headers you can download at any one
time and
2) If the article number itself exceeds 2^31 (2,147,483,648) you have
to use the 64-bit version of Xnews.
I don't think that 50,000,000 headers can be downloaded in one chunk
into Xnews. Is that the problem you're having? I never personally
download more than 300,000 at a time using incremental retrieval.
I'd be curious to know the details about how you tried to download an
article w/ Xnews and "couldn't". I'm not a NZB user so I just assume
that it is a way to correlate a list of articles with Message-id for
downloading. Xnews should have no trouble using incremental retrieval
to download small chunks of headers at a time until the desired
articles show up.
-- John
Ok, here's an example:
alt.binaries.erotica
Oldest article number on server: 288,254,267 (commas added for clarity)
Newest article number on server: 472,524,824
Xnews will refuse to acknowledge the existence of any article prior to:
422,524,825
Articles 288,254,267 through 422,524,824 are on the server but unreadable
by Xnews.
I've never had a problem with the article limit on Xnews.
I wouldn't use Xnews for binaries because it can't download multiple files
at the same time.
You can download as many binaries at the same time as there are Queue
folders. I have 8 Queue folders and would often download 8 binaries at the
same time - without invoking multiple instances of Xnews. (The server has
multiple ports, and I've entered each of these into their own server entry
in Setup)
To put it on a very basic level: when you Open Special (Ctrl+Enter on
the Groups page), it shows the Start header as 422524825 and the End
header as 472524824, with the total number of headers available to Get
as 50000000. There is no way to select a header prior to 422524825, even
though the Groups page shows a header range of 288254267 to 472524824.
Just try it on ab.erotica or ab.boneless.
If I really wanted to scan all 50 million headers, I'd do it in blocks
of no more 300,000. But why would anyone want to scan 50M headers?
--
Martin S.
I guess this is your way of getting round the Xnews default of no more
than 4 connections to one sever. With up to 4, you only need one Setup
entry. That's usually enough to max out my 14.4Mbps (1.8MB/s) bandwidth.
--
Martin S.
If I used 4 channels, I'd get a total of about 500 kBps. Adding 4 more
channels would get me an extra 100-200 kBps.
I used to think that 700 kBps was the max my ISP/NSP combo could provide.
I've discovered after using Grabit that I can get a reasonably consistant 830
kBps with it.
If I try to use more than 8 channels with Xnews, the computer gets so bogged
down that I can't do a darn thing with it... not even play Solitaire.
Boneless suffers from its own problem. The article number is so high that
its not being reported accurately with Xnews' 32 bit variable.
> If I really wanted to scan all 50 million headers, I'd do it in blocks
> of no more 300,000. But why would anyone want to scan 50M headers?
>
I don't scan... I search for specific keywords. I can use either my NSP's
search engine or a web site such as binsearch. When I use my NSP's search
engine, it's a simple matter to find out what the exact article number is.
I just finished a newsgroup scan of my NSP (Newsguy). The following are all
the newsgroups that currently contain more than 50 millions articles:
alt.binaries.ath
alt.binaries.bloaf
alt.binaries.boneless
alt.binaries.cores
alt.binaries.dvd
alt.binaries.dvd.german
alt.binaries.dvdrs.pw
alt.binaries.erotica
alt.binaries.games
alt.binaries.ghosts
alt.binaries.hdtv
alt.binaries.hdtv.german
alt.binaries.hdtv.x264
alt.binaries.hou
alt.binaries.kleverig
alt.binaries.misc
alt.binaries.mom
alt.binaries.multimedia
alt.binaries.nl
alt.binaries.pwp
alt.binaries.test
alt.binaries.town
alt.binaries.tv
alt.binaries.tv.deutsch
alt.binaries.u-4all
alt.binaries.usenet2day
alt.binaries.vcdz
alt.binaries.x
>> I'd be curious to know the details about how you tried to
>> download an article w/ Xnews and "couldn't". I'm not a NZB user
>> so I just assume that it is a way to correlate a list of articles
>> with Message-id for downloading. Xnews should have no trouble
>> using incremental retrieval to download small chunks of headers
>> at a time until the desired articles show up.
>>
>> -- John
>>
>>
>
> Ok, here's an example:
>
> alt.binaries.erotica
>
> Oldest article number on server: 288,254,267 (commas added for
> clarity) Newest article number on server: 472,524,824
>
> Xnews will refuse to acknowledge the existence of any article
> prior to: 422,524,825
>
> Articles 288,254,267 through 422,524,824 are on the server but
> unreadable by Xnews.
>
>
Unbelievable -- but true!
When I first query the group, it shows:
First = 233589000
Last = 474607552
Unread = Total = 241018553
But if I do a Ctrl-Enter only the last 50,000,000 are selectable and it
has marked all but the last 50,000,000 as being read. I had never
known that Xnews has this limitation. I guess I've been hanging around
the text groups too long...
Bummer. Thanks for opening my eyes. ;-)
-- John
Each of your 8 queues download separately. You can't have a master queue
feeding articles to whichever of the 8 connections is free.
True. You have to juggle headers between queues if you want to finish as
quickly as possible. Simultaneous connections don't always download at
the same speed, for some reason. Often a second or third connection will
go faster than the first one.
I used to use a binary downloader, whose name I forget, from the same
stable as PowerPost IIRC, that would do this. However, the reading and
marking of headers in large groups was cumbersome.
--
Martin S.
All 8 are constantly busy - none of them are "free." What more could you
ask for?
One other, perhaps minor, item. Compacting such a large number of
headers in an ng. The perceived percentage of fragmentation is so
low that, without altering the defaults, the manual storage button
isn't of use.
Question on nzb files.
I understand that they can be filtered / configured to down load
specifics?
Xnews doesn't really support nzb fully, so...
I go somehwere on web (binsearch or such), I search for subject, poster
or whatever I am interested in.
does it create just that, or will other posts even off topic, get
included too?
> My name wrote:
>
>> I go somehwere on web (binsearch or such), I search for subject,
>> poster or whatever I am interested in.
>> does it create just that, or will other posts even off topic, get
>> included too?
>
> The NZB created will contain whatever articles you select at the site
> that's generating it for you.
>
> You can open an nzb file with a text editor.
>
Thanks, I was wondering that!
<snip>
> Unbelievable -- but true!
>
> When I first query the group, it shows:
> First = 233589000
> Last = 474607552
> Unread = Total = 241018553
>
> But if I do a Ctrl-Enter only the last 50,000,000 are selectable
> and it has marked all but the last 50,000,000 as being read. I
> had never known that Xnews has this limitation. I guess I've been
> hanging around the text groups too long...
<snip>
Last week I stumbled over another limitation:
The message ID controls on the Refresh Headers Special dialog for
'start' and 'end' (and possibly 'get' too, but there is the 50M limit)
are only able to contain 9 digit numbers.
They display anything, even numbers beyond this 9 digit limitation, but
if you try to edit them, you can only enter up to said 9 digits.
Give it a try, e.g. take a.b.dvd currently starting at 9 something M,
adding 50M, ending at 1.0 something G.
Try to fetch the last 200K.
I simply copied the end value to the start field, deleted a digit and
... et voila ... I wasn't able to enter any digit.
This happend with version 2006.08.24, but with the new 2009.something
TEST-64 version too.
First I thouhgt Wine is to blame, but it happens even on Windows
(Windows 2000 that is).
I think the Delphi controls are limited, either the
implementation/configuration or the datatype used is causing this.
My guess, if more and more groups Luu Tran is interested in hit the 1G
barrier, he will change that.
Would be nice if he could remove the 50M limit as well.
Cheers
QvL