Header Fetching currently not supported

[Frasier]

But I can't find "a.b.u-4all" in the binsearch tab of AltBinz, because it only lists the first 1000 groups. Therefore I can not browse this group. Binsearch does seem to index it, but it doesn't do this right - many postings do not show up in Binsearch, but they exist when you browse the headers with NewsLeecher.

And it wouldn't be necessary to maintain a complete list of all headers between sessions - just some way of temproarily view the newest x headers in a group would do fine! I'm sure this is not difficult to implement.

Regards!

[Hecks]

To get all the Binsearch 'Other' groups, you need to replace the bsglist2.abz file following the instructions here:

https://www.altbinz.net/wiki/Using_the_Search_Engines#Binsearch

But as I said, the group you're looking for is indexed by NZBIndex as well.  I'll regard the "many posts don't show up" statement as an exaggeration until I see meaningful proof.

I think you underestimate the complexities of header handling and how this could be implemented in Alt.Binz.  If it's access to unindexed groups that's a priority for you, and your server doesn't support XPAT searching, to have any kind of search functionality you'll want to download more than just latest X headers for each group.  In any case, you'd be gradually building up a header cache for all of your groups to match your server retention.  That's a lot of local hard disk space and bandwidth to sacrifice for not much benefit in the long run.  And you'll know if you've used Newsleecher how frequently such caches become corrupted and unusable, meaning you have to start the whole process again, wasting even more bandwidth.  By design, Alt.Binz lets the indexing server do all of that grunt work for you. 

Who knows if Rdl will implement header downloading eventually.  Personally, I think it's not a priority feature on a pure effort vs reward basis.

[grogert]

Free search engines like binsearch are kinda shitty because files like nfos go missing alot.  I still have hope that header fetching will be implemented in alt.binz

[Hecks]

Free search engines like binsearch are kinda shitty because files like nfos go missing alot.

Proof?

[grogert]

I just checked some search engines and the only one that isn't missing files is nzbindex, but headers should still be implemented because there isn't any reason not to do it, especially after you requested posting features (and cr4zyfr4g supporting the idea) in the program but don't want header support?  I'd say thats quite idiotic. Furthermore it shouldn't take rdl more than a year to add this to the program as this thread is more than a year old already 

From the last 10 days of headers I used some random req id's from alt.binaries.cd.image and binsearch, newzleech, aeton and beta.binaries.nl are all missing nfo's for these (except for nzbindex):
16580 16660 16656 16657 16654 16644 16636 16633 16630 16624 16625 16626 16623

In other words, all except for 1 of them are crap and totally useless for people who collect releases.

[Hecks]

because there isn't any reason not to do it

Of course there is, this thread is full of them.  Finding non-indexed .nfos is a rather trivial reason to implement something like this, IMHO, especially if NZBIndex carries them (and of course NLSS does too).  There are plenty of sites that index those separately, if you're missing any (http://www.nfoogle.com).  Posting is an enirely separate function and has nothing to do with managing headers locally.

Please try to keep the discussion courteous.  We're only sharing opinions here, it's up to Rdl to decide whether to move this forward or not.  It's not that I don't want header support, it's just that I don't think the rewards are worth the (user) effort in this case.  I have my views, plenty of other people think differently to me (including Rdl, as I recall ).  That's life.

[grogert]

That reason is good enough because it shows how low the overall quality of those search engines are. Headers have no problems at all. I really wonder if you don't mind about header support being added to the program as you are trying to find tons of reasons not to do it. As if there is anything else worthwhile for him to work on other than fixing bugs and wasting time on weird requests like SSL binsearch instead of something extremely basic like header fetching, something that every good client should have. 

BTW, I was happily using Newsleecher SS until my subscription expired not too long ago and now I'm stuck with these free search engines as I am not going to pay for a program that I won't be using (Newsleecher). Guess he doesn't know/gives a shit about people who just want to pay only for SuperSearch. 

[Hecks]

Header fetching is only "extremely basic" in the way that a steam engine is "extremely basic", i.e. a redundant technology.   In this case, superseded by quicker methods that offload bandwidth and local disk usage to an indexing server.  Sometimes the indexing algorithms get it wrong and miss part of a post here or there, especially with posts like the ones you cited that have obfuscated subjects (and why is that?).  I'd be more worried if the free engines regularly missed whole posts rather than trivial .nfos that can easily be found elsewhere.  I simply don't see the point of clogging up my drive and wasting my bandwidth for such a meagre benefit.  And I'm glad that Alt.Binz development over the last year hasn't been held back by a need to incorporate and bug-fix a backward approach to using binary usenet.

This doesn't mean it won't be in future.  As I said, Rdl disagrees with me, if I recall our last discussion about this properly.

Anyway, this thread is going round in circles, and I meant what I said about keeping the discussion civil.

[davidq666]

why r u guys not using the build-in xpat search? by puting in a space u get (nearly?) all posts of a specific newsgroup. ok you can only browse one group at the time and when u changed groups you have to do the search again, but if you limit ur search to the last 1000-10000 posts it's quick enough. by the way hitnews is one of the suposed rare providers supporting it.

[Mystic2010]

why r u guys not using the build-in xpat search? by puting in a space u get (nearly?) all posts of a specific newsgroup. ok you can only browse one group at the time and when u changed groups you have to do the search again, but if you limit ur search to the last 1000-10000 posts it's quick enough. by the way hitnews is one of the suposed rare providers supporting it.

That might just be why not many (don't know whether that's true or not) use it, the provider you use needs to support it for it to work

[Rdl]

why r u guys not using the build-in xpat search? by puting in a space u get (nearly?) all posts of a specific newsgroup. ok you can only browse one group at the time and when u changed groups you have to do the search again, but if you limit ur search to the last 1000-10000 posts it's quick enough. by the way hitnews is one of the suposed rare providers supporting it.

Correct way would be to use wildcard and search for * . That way you'll get all posts. All providers I've used/am using have xpat... xlned, hitnews, easynews, newshosting. The only one I'm sure doesn't support xpat is giganews.

[davidq666]

why r u guys not using the build-in xpat search? by puting in a space u get (nearly?) all posts of a specific newsgroup. ok you can only browse one group at the time and when u changed groups you have to do the search again, but if you limit ur search to the last 1000-10000 posts it's quick enough. by the way hitnews is one of the suposed rare providers supporting it.

Correct way would be to use wildcard and search for * . That way you'll get all posts. All providers I've used/am using have xpat... xlned, hitnews, easynews, newshosting. The only one I'm sure doesn't support xpat is giganews.

tanks for that info :-)

[Sparkss]

This is about headers, but not necessarily about Dling all headers for a group. Is there, or can there be, a way to download and read a single header ? For example if there is a collection and someone has posted a reply to the collection, is there a way to just grab and read that single response (hearder contents) ?

Our previous reader had that ability, and to be honest, I don't miss much else about the old reader at all, except occassionally for that one feature 

[Hecks]

If you use NLSS, you can right-click + Open (to view article contents) and/or right-click + View Header (just article header info).  Extending this to other search engines has been requested before - although it isn't quite as straightforward, given how articles are indexed & grouped by these.

[Sparkss]

Ahh, ok, Thanks for the explanation. And NLSS is a paid subscription index/search engine ? (if I read other comments correctly).

Thanks again !!!