Date: Wed, 03 Apr 2002 00:56:08 +0000 From: "Matthew Hanson" <[EMAIL PROTECTED]> Subject: Re: [LIB] List archive search problem
>Date: Tue, 2 Apr 2002 16:35:50 -0700 (MST) >From: Michael Kopplin <[EMAIL PROTECTED]> > > >On Tue, 2 Apr 2002, Matthew Hanson wrote: > > Another thing I've noticed in the searchable archives is that some posts >do not include EOL characters. Thus you have to scroll back and forth to >read incredibly long sentences that are all on single lines. Again, >Raymond's posts show examples of this.< < > >Matt, > >I made a few changes, and rebuilt the archive. I did a quick >check and this seems to be fixed now. I had noticed them in the >past, but it was only an occasional message, so I didn't take >the time. It has been increasing though, so time well spent. Terrific Mike... Sounds like it may have been a relatively easy tweak if you got it ironed out this fast. I'm guessing this will be greatly appreciated by the silent majority. > > >But Mike... I wonder why when you're doing a regular search for some >term, frequently the results don't return hits for 2002, or frequently only >a few and not all.< < < > >The problem with the "All years" is that due to a hardware >failure, I switched to a backup server. The backup doesn't have >quite enough ram to reindex the entire archive, so the All years >index is not up to date. I should probably remove that choice >for now. Oh hecks... it's nice to have that multiple year option when you're searching for something relatively arcane. Maybe for the moment you could explain the limitation in your help file until you get it all worked out. >I will soon be switching back to the main server, which >can update all the indexes daily. Right now, only the yearly >indexes are updated, and those not every day. Not quite sure if I'm following you. Once a past year's worth of messages has been indexed, why would they have to be re-indexed every day? >Another reason you might not get any hits is if your search term >is too common. If a word appears in too many messages, e.g. any >word that appears in the footer, it is not included in the >index, and will be ignored in your query. That's a useful bit of info. Maybe it'd be nice to add that to your help file too. Great job at setting this all up though Mike! I forget how long back I discovered your archives, but they've been extremely useful. It seems you've chosen not to list the URL on the major web engines, which I think is great. It gives the list members a kind of a firewall from the web email address spam worms. Ciao, Matt _________________________________________________________________ MSN Photos is the easiest way to share and print your photos: http://photos.msn.com/support/worldwide.aspx ************************************************************** http://libretto.basiclink.com - Libretto mailing list http://www.silverace.com/libretto/ - Archives -------TO UNSUBSCRIBE------- Reply to any of the list messages. The reply mail should be addressed to: [EMAIL PROTECTED] - Then replace any text on the message's subject line: cmd:unsubscribe --------TO UNSUBSCRIBE DIGEST------ Do above but with this on subject line: cmd:unsubscribe digest **************************************************************
