home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: alt.gopher
- Path: sparky!uunet!usc!snorkelwacker.mit.edu!thunder.mcrcim.mcgill.edu!sifon!news
- From: peterd@bunyip.com (Peter Deutsch)
- Subject: Re: index for all of gopherspace
- Message-ID: <1992Jul28.023815.9882@sifon.cc.mcgill.ca>
- Sender: news@sifon.cc.mcgill.ca
- Nntp-Posting-Host: expresso.cc.mcgill.ca
- Organization: Bunyip Information Systems (the archie people)
- References: <1992Jul28.012727.22716@ra.msstate.edu>
- Distribution: na
- Date: Tue, 28 Jul 1992 02:38:15 GMT
- Lines: 67
-
- In article <1992Jul28.012727.22716@ra.msstate.edu> fwp@CC.MsState.Edu
- (Frank Peters) writes:
- > In article <1992Jul28.002902.8570@sifon.cc.mcgill.ca>
- peterd@bunyip.com (Peter Deutsch) says:
- > :
- . .
- > Indeed. I didn't use the reference to archie lightly. It seems to
- > provide a very nice framework for indexing an uncoordinated and
- > distributed information pool like the internet.
- >
- > The problem is that the information necessary to build a useful
- > information index isn't really there in a lot of protocols including
- > FTP and gopher. I used archie to find a software package recently.
- I
- > found files at a site that had a recently created version of the
- > package...yet that version turned out to be older than other versions
- > made available previously.
- >
- > It might be a useful exercise (one that, no doubt, has been done
- > before) to identify what information about a given gopher entry or
- > package available for FTP would be useful for indexing and see how
- much
- > of it can be practically made available. Things like (this isn't a
- > complete list...just off the top of my head) size, version ID, a
- brief
- > (one line or so) description, a keyword list, an archive date.
-
-
- You should pick up a copy of the DRAFT IAFA document that is
- being created by the Internet Anonymous FTP Archives Working
- Group of the IETF, it addresses exactly this kind of question by
- defining a set of recommended informational templates that we
- will be asking archive site operators to make available. There
- are templates to store additional information about software
- packages, Abstracts for papers, lists of available services, etc.
- We didn't list "Gopher sites", but there are several that either
- could be used or could serve as a model for defining a new template.
- The idea with such templates is that they are inherently easier to
- fill out, so people are more inclined to fill them out, and they
- are fairly easy to pick up, index and serve back to the net, so
- hopefully a variety of such indexing services will come into existance.
- Given that we think the Yellow Pages problem (ie the problem
- of locating specific services and service providers in a sea
- of a million hosts) is about to go away, we think such an indexing
- approach has a lot of potential.
-
- BTW, we are not trying to define "all possible information", merely get
- some of the most obviously useful stuff standardized and get people
- into the habit of providing this kind of "meta-data" about the
- information they are serving. I have described the last IETF
- as the "year of the template" since the idea seems to be simultaneously
- being discovered in multiple user services working groups I attended.
- I smell a trend in the air....
-
- You can get the latest draft of the IAFA doc from
- "archives.cc.mcgill.ca" in the files "pub/iafa/draft.part.I" and
- "pub/iafa/draft.part.II". the email list for this work is
- "iafa@cc.mcgill.ca" and if you'd like to join in the discussion send
- email to "iafa-request@cc.mcgill.ca". There is also an archive of the
- list on archives.cc.mcgill.ca in the directory "pub/mailing-lists".
-
- We hope to freeze this doc by November, so feedback from anyone
- who is interested in eagerly sought ASAP. Please send it to the iafa
- list so everyone sees it.
-
-
- - peterd
-