Subject:
|
Re: Weird thing I just noticed
|
Newsgroups:
|
lugnet.admin.general
|
Date:
|
Mon, 22 Jan 2001 00:09:31 GMT
|
Viewed:
|
435 times
|
| |
|
|
In lugnet.admin.general, Eric Joslin writes:
> > This appears to happen to me too - all the older stuff shows up 1st. I'd
> > imagine a good default would be newest 1st????
>
> I wasn't so concerned about the order it appeared in (which I beleive is
> ranked by what best matches your search criteria, not time-based), but the
> fact that it displayed partial content for an article you can't actually
> view.
>
> I don't mind if Todd wants to keep the articles for Lugnet historical
> purposes and make them unreadable, and I don't mind if he chooses to make
> them completely visible to the public in an archive that everyone can read
> but not post or respond to. But I'd really rather that if someone can't
> read *all* of an article I wrote, that they not have a way to see *any* of
> an article I wrote.
This has been a problem ever since the first search function back in
1998, and I'm sorry for all the confusion it has caused when articles
labeled "Canceled" show up in search results.
In a feverish moment of clarity the other night, I thought of a simple, clean
solution to this longstanding problem: If I have the system maintain a giant
linear bit-table of cancelled articles and feed this table into the search
query engine as an "exclude these" list, then it could ignore cancelled
articles in an extremely efficient way -- similar but opposite in polarity
to the way it has "include these" lists for the subgroup queries. In effect,
the cancelled articles would be ignored early on and never show up in the
search results, rather than ignored later on during the final display page.
This direct approach was never feasible with the old query engine, but is
straightforward with the new one.
--Todd
|
|
Message is in Reply To:
 | | Re: Weird thing I just noticed
|
| (...) I wasn't so concerned about the order it appeared in (which I beleive is ranked by what best matches your search criteria, not time-based), but the fact that it displayed partial content for an article you can't actually view. I don't mind if (...) (24 years ago, 18-Jan-01, to lugnet.admin.general)
|
6 Messages in This Thread:
  
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
