RE: [ba-unrev-talk] RE: Immutable Granular Addressability, was ...
Jack - thanks for the pointer to the Lucerne search stuff. (01)
I've been wondering why there are not more search tools like
AltaVista Discovery. Anyone have any insights? (02)
(Discovery is apparently no longer supported. It was a version
of AltaVista's search engine that ran on your PC. It understood
about 150 file types, including hmtl, pdf, word, email, ppt, etc
etc etc. Essentially, you could do an AltaVista search on the
contents of your hard disk. I found it most useful. 20 GB is
a lot of information: I really don't enjoy the task of filing
stuff so I can find it later. :-( ). (03)
I must admit, though, I stopped using it after about a year
because my machine effectively ground to a halt when the indexer
was running. (04)
Both AltaVista and Google are trying to sell services like this
at the enterprise level. I don't know how much success they're
[mailto:firstname.lastname@example.org]On Behalf Of Jack Park
Sent: Monday, October 07, 2002 10:35 AM
Subject: [ba-unrev-talk] RE: Immutable Granular Addressability: Was: Re:
[ba-unrev-talk] Re: International Purple Numbering Standard? (07)
Terrific ideas, Barry. (08)
The open source email client ZOE http://sourceforge.net/projects/zoe
includes code that allows you to use the google engine. To use that code,
you must sign up with google for your own account. That email client also
includes the Lucene search engine http://jakarta.apache.org/lucene which
indexes your entire inbox, meaning all your emails are ready for keyword
At 10:19 AM 10/7/2002 -0700, you wrote:
>Google presents the user with the option to look at the current version
>of a page, or at the version that the indexing engine used. Sometimes
>you find some very interesting things by looking at the older version :-)
>Google is great for finding things on the web (at least compared to what
>came before it :-). But it's *useless* for keeping track of or summarizing
>what you've found.
>I'd like to see something like the Google toolbar that helps with the
>problem of recording useful (re-usable! shareable!?) references to things
>found on the web. I described the following extension to the folks at
>Google during an interview some time ago: they thought it was technically
>straightforward and 'interesting'.
> Once I've found a useful piece of information on a web page, I'd
> like to be able to select the text in question, then e.g. do a
> mouse click and select 'create reference' from the pop-up menu.
> (Google and Kontiki have both added 'search' commands to the
> right mouse button pop-up menu.)
> The 'create reference' verb would:
> 1) Save the selection, along with
> 2) the URL
> 3) time of capture
> 4) any notes the user might want to add
> 5) a snapshot of the page
> At a later time, the user would be able to examine both the
> original page from which the referenced item was extracted, and
> the current version of the page. Without the cooperation of
> the page's owner.
>If I may attempt to put this in Doug's vocabulary: I think Google is a
>point tool for the "intelligence collection" portion of the knowledge
>activity, but it does not connect well with the "recorded dialog" or
>product" related activities.
>By the way, Google has recently provided a web services interface to their
>engine. I've not looked at it closely: I wonder if their linkbase is
>as part of the web service?
>[mailto:email@example.com]On Behalf Of Mei Lin Fung
>Sent: Monday, October 07, 2002 1:29 AM
>Subject: RE: Immutable Granular Addressability: Was: Re: [ba-unrev-talk]
>Re: International Purple Numbering Standard?
>Why not build on top of Google's link database (they don't call it that,
>but it effectively is a db of links)?
>They already cache all the pages and archives.
>We need someway to have links to between the time series of snapshots.
>Google's back link feature might play a role in this.
>If you download the Google Tool, you'll see a command called "Page Info"
>with a little " i " info sign. Pull down that menu and you'll see back
>links as an option. Click on that and you can see all the pages that
>link to that page.
>[mailto:firstname.lastname@example.org] On Behalf Of SAsites
>Sent: Monday, October 07, 2002 8:38 AM
>Subject: Immutable Granular Addressability: Was: Re: [ba-unrev-talk] Re:
>International Purple Numbering Standard?
> > What happens when the document being referenced grows or shrinks by
> > one or more array items? People's links elsewhere are now broken.
> > Which is basically what you are saying with:
> > > By the way, another pressing issue intimately related to this matter
> > > remains unresolved is that of link integrity. i.e., how do we keep
> > > to our sources from breaking if we don't control the source?
> > >
> > > I believe I have solved this problem and will disclose details
> > I'm very curious to hear this.
>(laughing) Chris, you sound as if you are a little skeptical. If anyone
>I had made that statement I would be too ;) Note however that I said "I
>believe I have solved..." (emphasis on I). I never said anyone else
>believe me. ;))
>I had a snafu with my pc this weekend. I had hoped to have this
>described. I hope I don't disappoint (part of the reason for the delay),
>want to formalize/clarify my thinking. But the solution is surprisingly
>simple. As a matter of fact why delay...
>---- IMMUTABLE GRANULARLY ADDRESSABILITY: SOLVED ----
>All one need do is archive the page at the time that the link is made.
>a snap shot of the page as it exists at that moment and link to the
>snap shot. (Think about it for a few seconds).
>---- END SOLUTION ----
>http://web.archive.org is doing this now sorta, but their archive is,
>according to them, about 6 months behind (I am not sure why there is
>lag). While I can't use their archive to link to today's NYTimes, I can
>(using PurpleSlurple) link back to an archived version of Fleabyte and
>quite confident that link will remain intact. <== think about that for a
>(NOTE: In the above example PS is not required because Henry had the
>sense to us Eugene's Purple. But I wanted to relate something that this
>group is familiar with, and besides Henry was directly responsible for
>showing me the solution).
>So how do we get immutable granular addressability to today's news? If
>haven't figured it out already - stay tuned :)
>Matthew A. Schneider (011)
XML Topic Maps: Creating and Using Topic Maps for the Web.
Addison-Wesley, ISBN 0-201-74960-2. (012)