Missing files from Giganews and accurate Newsbin status

[markeh]

I downloaded a set of files from my usenet service, Giganews.
Newsbin showed all of them complete. But some of the downloaded files were incomplete, a few were missing.

One of the log entries shows:
[13:59:29] HIGH - Failed to retrieve Article from Server:Giganews-europe ARTICLE <4bd14d00$0$5660$c3e8da3@news.astraweb.com>
| 430 no such article

I don't doubt the article was missing - I guess that's a problem with Giganews. The post date (2010) should be well within the advertised retention time, so I don't know why its not there. I may need to try a different Usenet service to see if they can do better...

The problem is that Newsbin shows everything complete. If I had understood that there were missing parts, I might have skipped the download.

Is there a way for Newsbin to accurately reflect what's actually on the server??

Windows 7, x64
Version 6.42: 00 64 39 F9 E2 60

Both display and Storage age set to 9999

thnx,

[Quade]

Where did you get the information about how how to download?

Here's the problem, the header data becomes stale over time. Meaning if you download headers a year ago or 2 and those are the headers you're looking at they're just not accurate any more. When they were downloaded, they were accurate. Now, years later, they're not accurate any more. The problem is the same with NZB files and our own search. Over time, bit-rot sets in and the data you have that shows how to download isn't what the server really has on it. The problem isn't the data. It's that the server loses parts of the download. If you have enough PAR files this doesn't matter. The files can still be fixed. If you don't, the downloads are trash or you need another server that has the missing pieces.

In my experience lately, Astra actually has better long term retention than giganews. Anything over 1000 days on Giga is a crapshoot.

Newsbin has a "test download" function in the right click menu. It'll show you how much of the file can be downloaded without actually downloading it.

[markeh]

the header data becomes stale over time. Meaning if you download headers a year ago or 2 and those are the headers you're looking at they're just not accurate any more. When they were downloaded, they were accurate. Now, years later, they're not accurate any more.

I get your answer. How can I "refresh" the headers from Giganews?? If I "download all headers", does that do it??

thanks,

[Quade]

You can try that. You're setting you self up for gigas of downloads. It also assumes that Giga's header database isn't itself stale.

I might pick a group and try it and see if it's more accurate. Testing the download will let you be able to find out without having to download everything. If it has pars and at least 90% of the file is there it should download and repair. Might use 95% to be sure.

I downloaded some files from 1200 days ago the other day and even with 2 servers a couple of them were unrepairable.

[markeh]

I downloaded some files from 1200 days ago the other day and even with 2 servers a couple of them were unrepairable.

I guess meaning that the advertised retention time (1879 days binary retention stated on their web site) is mostly b.s..

I guess the most reliable strategy is to pay for two independent providers.


thanks,