[Date Prev] [Date Next] [Thread Prev] [Thread Next] Indexes: Main | Date | Thread | Author

[ba-ohs-talk] (offtopic) regex, backlink database data


> It won't match. The ^ means start of line.
>
> Eugene, I think what you might want is /^--\s?$/ because otherwise
> aren't you matching too much?    (01)

Ack! Yes. I'm getting rusty.
Sorry, Eugene. And thanks, Chris.    (02)

Although my mail client (Outlook Express) does have an alarming tendency to
reformat mails after I've sent them.
So '-- ' in the middle of a sentence might well drop to
'-- ' on the odd rare occasion (with Eugene's regex, not Chris').    (03)

What does 'foofie' mean?    (04)

Cheers,
Peter    (05)

----- Original Message -----
From: <cdent@burningchrome.com>
To: <ba-ohs-talk@bootstrap.org>
Sent: Sunday, December 09, 2001 6:52 PM
Subject: Re: [ba-ohs-talk] DKR Progress, backlink database data    (06)


> On Sun, 9 Dec 2001, Peter  Jones wrote:
>
> > > Question: when do links need context?
> > Um, not sure. But a little descriptive metadata can go a long way in
> > interaction -- crucially, to aid lookahead. If it's in a standard format
it
> > goes further in enabling more interesting forms of interaction to be
built
> > in. But again, lots of limiting choices have to be made.
> > 'Lookahead' is the key word for me.
> >
> > Cheers,
> > Peter
> >
> > P.S. Eugene, what happens to /^--\s*$/ when it hits my last opinion
above?
> > Methinks it gets messy. ;-)
>
> de-lurking for a moment...
>
> It won't match. The ^ means start of line.
>
> Eugene, I think what you might want is /^--\s?$/ because otherwise
> aren't you matching too much?
>
> Important to remember in this discussion that the matching is not to
> edit the message but to constrain the searching for URLs. As such
> Eugene's plan seems pretty good. Especially since what we are talking
> about here is gathering some URLs where none were gathered before.
> Some better than none, right?
>
> This list has a tendency to get a bit distracted by the details,
> delaying actual experimentation, don't you think? What Eugene is
> talking about here isn't particularly complicated, could gather some
> interesting data and could expose some patterns of understanding that
> lead somewhere interesting. Even if fails to create a product of some
> kind the process of experimenting with something semi-tangible has a
> much different impact than just thinking about it.
>
> Or if you want me to be a little more foofie about it I can come back
> with a bunch of noodling about Barsalou, Landauer, Winograd and Flores,
> Suchman, Schon, etc. etc. and of course Engelbart but that would belabor
> things.
>
> --
> Chris Dent  <cdent@burningchrome.com>
http://www.burningchrome.com/~cdent/
>
>    (07)