Conference hydra::amiga_v1

    Try : DIR/OUTPUT=filename note-range
    I access AMIGA,AMIGA_TECH and AMIGA_USENET from Germany and had
    simlar problems. But once you extract the hole directory into a
    file, you'll be able to edit this file and serch for whatever
    you want.
    
    Volker      

	I wouldn't claim to be an expert on Notes protocols, but surely
	the length of time taken to do a search is a function of the
	filesize alone. Once  your DIR command has winged its way over 
	to the US then the searching is done locally. The time taken
	to send the command line is probably insignificant judging by
	the good response times for other operations.

	Splitting the file has some merits but you would have to know
	which file you want search. In the end you could end up searching
	both files anyway!

	Anyone care to re-structure this file as a relational data base? :-)

    
    I agree with Ian (.0), because Volker's idea is sound but
    
    a) extracting the directory would take *hours*
    b) it would be out of date very quickly.
    
    There are arguments against Ians suggestion though... there are
    some *very* old notes which are still active.
    
    					Alan T.

    re .3
    It took me about half an hour to extract the USENET directory, and,
    if this is too long, make a batch job.
    Updating is very easy, once a week I extract new notes and
    APPEND them to the old file.
    I'd prefer this rather than splitting the whole file into one
    (next year perhaps two, in 5 years ...) files.
    
    Volker

    
    This is the 2nd try at replying, the network partner keeps getting
    lost!!
    
    Anyway.....
    When I said everything from before jan 1st should be put into another
    conference I really meant those notes that were `dead'. The ones
    such as 1156.* (Fred Fish Disks) could be used in both conferences
    and both could be updated. When notes have not had any more replies put in
    for a certain amount of time (e.g 4 months) they could migrate into
    the old conference, this would therefore keep the conference up-to-date
    and also reasonalby fast for U.K users and others.
    
    Problems with Volkers idea are:
    	1. It would take a long time, even running it in batch,
    	2. The amount of disc space it would take up is enormous, and
    	3. It would be out of date very quickly indeeed.
    
    The only other solution I could find is to find a node in the U.k
    or Europe and have exactly the same information on it as there is
    on this conference, but even then I doubt that access times would
    improve very much.
    
  			Cheers, Ian.
    
    
    Ian Norris, Dec Park I, England.
   	

I also access this conference from DECpark in England. I find it is one of
the faster conferences I look at! Searching and doing directories to files
really doesn't take that long - especially in batch. Of course, I do try
and access at non-US core times - and even run the batch jobs at both
non-US and non-UK core times.

Separating dead notes from live notes is a good idea - if a) you could find
someone who could spend a large amount of time doing it and b) we could
all decide which were dead and which were alive - just because no-one
replies doesn't mean that no-one reads them. 

Experiences from other conferences that have split (eg Rainbow) show that
the same old questions get asked again in the new conference - rather
faster than they might have done if the conference had not be split.

Nick

    I also use the conference from the UK. I do not find any serious
    problems, although directories do, sometimes, take a while. However,
    I've installed AVN (Automatic VAXnotes) on our system since then,
    and that helps tremendously. Can I suggest you talk to your system
    manager?
    
    I would be against splitting the conference, as I have occasionally
    had to go back to old notes. Splitting will make looking for info
    much more tedious.
    
    Martin

    AMIGA_TECH?? Where is it?

< AITG::WISNER "Paul Wisner, ...I have a totally traditional haircut..." >


>    AMIGA_TECH?? Where is it?

    
    So someone thinks access from the UK is slow? You should try it from
    Australia.
    
    FRSOLD:: USENET stuff is fine, but BOMBE::AMIGA is a real dog. We have good
    high speed links to the US through the Pacific, so something seems
    to be dying beyond MA. We dont get this on most of the other
    conferences. 
    
    Contention is unlikely to be the problem, as my day access is in the
    US midnight to dawn timezone - or is that when all you hackers are out
    on the town??!!!!!
    
    The only manageable ways to use the conference at the moment
    are Batch DIR/OUTPUT and interactive SEARCH.
    
    What impoverished group is donating the BOMBE host? What type of node
    is it and what type of disk? Does anyone have an idea what bottleneck
    is? 
    
    If someone can fix the speed, I will throw in a free beer (proof of
    fix needed!). I mean, this is the most important conference at DEC,
    it needs at least a 8900 or a Cray!
    
    Regards.

    Re .10:
    
>   FRSOLD:: USENET stuff is fine, but BOMBE::AMIGA is a real dog. We have good
>   high speed links to the US through the Pacific, so something seems
>   to be dying beyond MA. We dont get this on most of the other
>   conferences. 
    
    Unfortunately, the high speed links don't make it to Boxborough, MA
    where the BOMBE cluster lives.  It's not for lack of horsepower HERE
    that the response seen elsewhere is less than wonderful.  (FRSOLD::
    access isn't real spiffy from THIS vantage point for the same reason.)
    
    There's a continuing battle to improve our network service for our OWN
    needs.  As we make progress, access to BOMBE::AMIGA should also improve.