The ROADS Frequently Asked Questions List (FAQ)

Version: $Id: roadsfaq.html,v 1.14 1997/03/11 17:17:12 lisap Exp $

Overview

This is the Frequently Asked Questions document about ROADS, Resource Organisation and Discovery in Subject-based Services.

Table of Contents

1) Introduction

1.1) Conventions
1.2) Where can I get the ROADS FAQ?

2) What is ROADS?
2.1) How is the ROADS software used?
2.2) Where can I get the ROADS software?
2.3) How is ROADS funded?
2.4) Glossary

3) Administrator Questions

4) User Questions

5) Miscellaneous Questions

6) Other Resources

7) Authorship
7.1) Acknowledgements
7.2) Permissions


1) Introduction

1.1) Conventions

1.2) Where can I get the ROADS FAQ?

2) What is ROADS?

2.1) How is the ROADS software used?

ROADS has four main objectives:

2.2) Where can I get the ROADS software?

If you think you might be interested in using the ROADS software for your own project / research / needs contact the ROADS liaison officer on: { Notes on availability, licencing, etc. to be supplied }

2.3) How is ROADS funded?

ROADS is a development project funded by the UK Higher Education Funding Councils via the Information Services Sub-Committee (ISSC) of the Joint Information Systems Committee. It is part of the Electronic Libraries Programme (eLib), set up as a result of the Follett Report, which is steered by the Follett Implementation Group on IT (FIGIT).

ROADS fits into the ANR (Access to Network Resources) section of eLib.

2.4) Glossary

Every subculture needs its list of buzzwords and acronyms. Here's a collection for ROADS.

I18N Internationalisation ('I', followed by 18 letters, followed by 'N').
IAFA templates
MIME Multipurpose Internet Mail Extensions - RFC 1521
RFC request for comments; proposed or standard Internet protocols
URL WWW uniform resource locator; access-method://host/path
WHOIS++ a distributed directory protocol
WWW the worldwide web

3) Administrator Questions

3.1) What technical documentation exists for ROADS?

The ROADS technical report archive is at

<URL:http://www.roads.lut.ac.uk/Reports/>

3.2) What ROADS releases are available? Where can I get them? Which version do I use for a production service?

You can find information on ROADS availability and release status on the software distribution pages

<URL:http://www.roads.lut.ac.uk/>

ROADS v 0.25 is the current production version. However ROADS v 1 offers much more functionality that v 0.25 so in general we'd recommend that sites wanting to use ROADS use that version instead.

3.3) Is there a ROADS bug list?

This information is also available from the software distribution pages.

You can go directly to the ROADS v1 bug tracking page for information on the status of ROADS v1 bugs

<URL:http://www.roads.lut.ac.uk/bugtrack1.0.html>

3.4) How does relevancy ranking work?

The ranking all happens in lib/rank.pl. This piece of Perl takes the query and the templates handles that have been matched by the search mechanism. It first removes all booleans, WHOIS++ constraints, punctuation and other guff from the query string. It then cycles over all the templates handles and looks for occurances of each term in the munged query in each line of the template. A counter is incremented for each match and at the end of the template the handle is placed in a hashed array with a key equal to this count (there maybe handles with this key already; if so the handle is appended to the existing list).

Once all the templates have been checked, a normal array is built by doing a reverse sort on the keys of this hashed array (so that the largest counts, i.e. most hits, come first) and pushing the handles from the hash array's values on the normal array. This array is then returned for use by the rendering routines.

Note that if you wish you can put your own ranking algorithm in place by replacing lib/rank.pl with your own Perl routine. The subroutine must be called "rank" and must take arguments of the query string and a list of unranked result handles. It must return a list of rank result handles (the lists are normal Perl arrays).

Answer supplied by: Jon Knight <jon@net.lut.ac.uk>, February 1997

3.5) How do I set up multiple databases under ROADS version 1?

First it is important to remember that there is no support for real multiple databases under ROADS version 1. However you can enable multiple logical databases by making use of the 'Destination' attribute in ROADS templates. The following steps should get things going.

Essentially the value of the 'Destination' attribute is the 'name' of the logical database. For the purposes of this example we'll assume that all records will go into one of two logical databases, 'Biology' or 'Biochemistry'.

  1. Make sure that there is a 'Destination' attribute listed for every template type in the 'config/outlines' directory.
  2. Update any existing records in your database, giving them each a 'Destination' attribute value of either 'Biology' or 'Biochemistry'.
  3. It is important to remember that there is no way of explicitly searching for records with an empty 'Destination' attribute (though you can search for 'ALL' records) so normally you will want to add a 'Destination' to every record that you create. To enforce this you may want to make the 'Destination' attribute mandatory and offer a selection of values for it at template creation time. To do this update each of the files in the 'config/outlines' directory making the 'Destination' lines look something like this:
    Destination:::Biology|Biochemistry:m:
    
  4. Configure the two logical databases by updating the 'config/databases' file. Typically this file will contain a single entry. Something like
    ROADS TEST SERVER:roads.ukoln.ac.uk:6663::roadsukolnacuk01
    
    This single entry should be replaced with two (or more) entries:
    Biology:roads.ukoln.ac.uk:6663:Biology:roadsukolnacuk01
    Biochemistry:roads.ukoln.ac.uk:6663:Biochemistry:roadsukolnacuk01
    
  5. Make sure that the 'Destination' attribute is listed as one of the searchable attributes for each of the template types listed in 'config/search-restrict'.
  6. Re-index your database:
    bin/mkinv.pl -a
    
    Finally, use your Web browser to bring up the 'cgi-bin/search.pl' search screen. You should now be presented with, and be able to search, multiple databases.

Answer supplied by: Andy Powell <a.powell@ukoln.ac.uk>, September 1996

3.6) What template types are there?

Michael Day at UKOLN is in the process of creating a registry of ROADS template types at:

<URL:http://www.ukoln.ac.uk/metadata/roads/templates/>

Answer supplied by: Andy Powell <a.powell@ukoln.ac.uk>, February 1997

3.7) Is it safe to have several people updating a ROADS database at the same time?

3.8) I seem to be having some problems with standard UNIX commands because my ROADS 'source' directory is so big?

3.8)

4) User Questions

5) Miscellaneous Questions

5.1) Why IAFA templates?

Based on IAFA but now referred to as ROADS templates or WHOIS++ templates. Simple, flexible, easy to use metadata format - same as the 'on the wire' format delivered by WHOIS++.

5.2) Why WHOIS++?

Seemed like a good choice at the time! Now looks like LDAP might possibly have been a better choice (everyone, including Microsoft and Netscape seem to be going down the LDAP route) - but still not absolutely clear that is the case. WHOIS++ and CIP (Common Indexing Protocol - which provides query routing) will provide good solution to distributed searching. LDAP and CIP will happen but probably later than with WHOIS++?

The ROADS software takes a toolkit approach. This allows, for example, reasonably easy conversion between different formats of metadata. If necessary it would be possible to rip out WHOIS++ server and replace with LDAP - leaving other ROADS tools in place.

5.3) How big can a ROADS database be?

This is not clear yet but is being investigated. The answer will obviously depend on complexity and size of records as well as simply on the number of records. If necessary it is possible to rip out the ROADS database and slot in, say, an Oracle database behind the ROADS WHOIS++ server using the ROADS API. Again, this is a feature of the toolkit approach.

5.4) How configurable is the ROADS user interface?

Very flexible. Most of the HTML generated by the ROADS software is based on configurable skeleton files - including the rendering of search results. Compare results from the eLib project database, SOSIG and ADAM for example. There is no reason why skeleton files couldn't contain JavaScript and make use of cookies to configure the output further - though I don't think anyone is doing this yet.

ROADS also contains support for I18N based on client supplied HTTP headers.

6) Other Resources

Web pages, mailing lists, DESIRE, NWI, SOSIG, ADAM, EEVL, etc, eLib.

Services using ROADS:

7) Authorship

The layout of this FAQ was based on the MMDF FAQ.

Answers have been taken from existing ROADS Web pages and from various questions and answers sent to the roads-liaison, open-roads and cousns mailing lists.

The ROADS FAQ's current maintainer is Andy Powell <a.powell@ukoln.ac.uk>.

7.1) Acknowledgements

Contributions to this FAQ have come from:

John Kirrimuir, Jon Knight, Andy Powell

{ Your name and, optionally, e-mail address or URL, could be here! }

7.2) Permissions

Permission is granted for unlimited redistribution of the unedited ROADS FAQ.