It sounds like 1.1 is going to be really great and I'm sure your working hard on it Alex. Any idea when (roughly) you will be releasing the final version? I don't need a specific date or anything, just a time estimaet. Thanks!

*** Please note this message is not meant to be any type of pressure on Alex or demand for an update ***


------------------
Ted Sindzinski
www.infinityinternet.com

i don't think "soon" or "next week" would be your answer.. i'm not sure if Editors is gonna be part of the final of 1.1.. but it looks to me as if he hasn't really finished it yet..

although he does have the feature to add them.. he doesn't have the feature to login to the admin as the editor..

when i made the admin mod for links2.. it took me 4 days.. plus a ton of bugs.. that kept popping up in the following week.. making huge changes like this take much time..

jerry

[This message has been edited by widgetz (edited October 07, 1999).]

[This message has been edited by widgetz (edited October 07, 1999).]

Thanks a lot!

Well, we keep adding features. =) For instance, the next version will have support Altavista style syntax in the searches:

+"Gossamer Threads" -sewing

as well as the original mode type searches. It will still be indexed to avoid full table scans as well. Should be a great new addition.

As for editors:


This is there. If you go to:

http://gossamer-threads.com/perl/links/user.cgi

and log in with any editor account, you will be able to add/remove/modify links. We still have some work to do in cleaning it up a lot.

Our ETA is to have a final released version within 14 days.

Cheers,

Alex

Alex...

Cool stuff!

Let me ask an interesting question (I know I can see it in operation in 14 days but..) if you are using the quoted "Gossamer Threads" to find things next to each other, yet still index it, wouldn't that require an exponential or logarithmic expansion of the keywords file?

Or are you going to do an "in order" where "Gossamer Threads" would find a link that had both words, in that ORDER (not necessarily next to each other) -sewing and drop links that had sewing in them?

IMWTK

Also, if you want a beta testor, please feel free to send me the code I'm not running it 'live' yet, so I have no problem installing intermediate versions and looking for bugs/problems/quirks, and even when I do go live, I'll have to have a pre-release site for integrating the upgrades, changes and whatnot before deploying it - so I'll always accept new code

Until I figure out the 3-level graphic display thing I can't make the SQL version live, but I'm also trying to integrate my postcards script with SQL and roll out the two changes together (which is why I'm not bugging you to death on things <G> ). There's a steep learning curve on this, and I'd rather get near the summit before I put my business into it's hands -- I can hack flat-files and basic perl to death, but SQL/MySQL/Database-driven/OOP/OOPerl etc is a whole new beastie.....

Yes, we are using an "in order" way of looking for the records. It shouldn't be too hard for anyone to hack the code to post-humously check the result for a perfect match. (the search is, well, O(nK^(m-1)) where n is number of records in the database, and m is number of search terms entered. K is > 1 but < 2, I don't know any offhand estimates of K but, 1.5 could be a good guess)

The current alpha api tries to remove itself from the target table as much as possible. This may change.

Incidentally, does anyone here use keyword*suffix type searches?

ended up posting twice. oops

[This message has been edited by kitsune (edited October 08, 1999).]

I'm not sure exactly what you mean, but if you mean something like "a*phobia" to find all phobias starting with "a" I'm not sure that is a very common way of searching, since most 'normal' people don't think like that.

Computer, Computer Resource or programming sites would have a higher incidence of people doing 'regex' type searches, but I can't see my partners or my mother even thinking of the possibility of an "xx*xx" type search.