Dr. Joanna Richardson, University Library, Bond University, Gold Coast, Queensland Australia. Joanna_Richardson@bond.edu.au
Antony (Tony) Barry, Australian National University Library,
Canberra, Australia. tonyb@dynamite.com.au
URL managers, information storage and retrieval, digital libraries
Information previously stored in different places, and in different ways, when based on paper may have been held in local paper files, as collections of books or magazines, or as bibliographical references to material to be consulted in a library. With the advent of the web all of these can potentially be accessed in the one way - via a URL reference. Bookmark files to control these get out of hand as the collection becomes large. This paper reviews software which has been developed to control access to URLs and discusses integration with paper based information.
This paper addresses the problem individuals have in keeping track of large bodies of Internet resources. It is in part a summary of work on a database to describe software to do this and a discussion of its content. Therefore the methodology of establishing the database will be discussed as well as conclusions drawn in establishing it and conclusions with regard to the content. This however is a work in progress and the content of the database will continue to evolve. To ensure that the paper remains relevant to the database an updated version [HREF 1]will be maintained.
The Internet is generating new problems in documentation for the general public and for scholars. Material which would have been held in books and magazines in relation to, for example -
can all be held as URL links in browsers such as Bookmark files
(Netscape) or Favorites (Internet Explorer).
New problems of organisation arise, as formerly disparate material which may have been stored in different physical ways, now ends up stored together. As the volume of this material rises, the ability of simple hierarchical arrangements and text searching for a single string to locate material becomes overwhelmed. This may not be a problem for many people, but for those with an in-depth interest in a topic something better is needed. For scholars who collect a large amount of material this is particularly acute.
In the world of paper, scholars collected bibliographical
references to material which interested them. Prior to computers
these were often collected as card files arranged in various
idiosyncratic ways. The most sophisticated used marginal punched
cards to aid in sorting and retrieval. When computers entered the
equation, personal bibliographical software such as -
provided much more sophisticated capabilities. As well as providing searching features it was possible scan drafts of research papers so that footnotes could be generated automatically.
With the advent of the web scholars are increasingly collecting URL references -- a process which raises new problems in document management.
A bookmark or favorite is a memory for an Internet URL. Netscape calls them "bookmarks" and Internet Explorer (IE) calls them "favorites", but they are basically the same.
Internet Explorer (on Windows) stores every single favorite in a separate file with the extension 'url'. It uses normal directories on the hard disk as folders in the favorite structure. Because of this, you have to specify a directory as the location of the favorites, not a single file.
Netscape, on the other hand, stores the complete bookmark structure in one single file, most commonly named Bookmark.htm (Windows) and bookmarks.html (MacOS and Unix). Because of this, you have to specify one single file as the location of the bookmarks.
A challenge arises when people use one or more browsers, e.g. IE in their office but Netscape at home, and want to manage their URL links. The differing data structures do not allow you to simply merge these links into one single file that your browser can automatically read. In addition, the inherent constraints of the browser determine your "view" of your URL links, which may not necessarily suit your modus operandi. Hence the need for management tools. In addition GVA's annual survey [HREF 5] finds that people have difficulties with bookmarks/favorites
The problems with hierarchies are:
A variety of solutions has arisen but so far none is dominant. These include -
These vary in their capability to-
Few solutions provide integrated support to paper-based information (other than traditional personalised bibliographical software which is moving in this direction). There are disjointed developments which separately may help future integration of paper and web based information.
One way is to convert paper to electronic form but this is quite expensive. Paper based information delivered in machine readable form via document delivery services such as Uncover [HREF 8] can be received in fax or email form. There are fax to email gateways, such as the services provided by JFAX [HREF 9] Some document delivery suppliers and certainly the library community are providing scanned images via email via the Research Library Group's Ariel software [HREF 10] as upgraded by the JEDDS project [HREF 11]. We can expect to see a rapid expansion of this as the LIDDAS [HREF 12] software is deployed,
For those library systems which provide Web enabled OPACs with stable URLs to entries in their catalogue, direct URL links to items in their collection are possible A number of online bookshops do this, Amazon [HREF 13] for instance. They maintain a viable business model for virtual bookshops this way. These techniques allow you to keep a URL pointer to a printed item and integrate print references with the handling of URL references.
Personalised tools are also appearing that enhance the ability of individuals to find and select material of interest to them on the web. There are -
Tools which support individuals are more developed than those which support groups although -
Some of these tools are discussed below concentrating on software which can maintain and manipulate databases of URL.
As we all recall, step one in eating an elephant is to find an elephant! Fortunately well-known sites such as TUCOWS and FileDudes offer reviews of what are generically referred to as "Bookmark Utilities", which is much easier than doing a keyword search on one of your favourite search engines. However, the authors of this paper found that reviews are often not as helpful as they might first appear.
The fact that a utility is awarded three cows or four stars or a partridge in a pear tree does not necessarily tell you, the user, whether you will be "happy" with this product. We have found software which, although not allocated a high score, proved to be very useful in terms of the narrow functionality for which it was designed. In other cases, the "review" is simply a cut-down version of what the author/vendor has to say about their u-beaut product. Or authors / vendors draw your attention to wonderful reviews, which on closer inspection can be as much as three years out of date.
The word "review" inherently implies a "critical" analysis. And it is this aspect which seems to be lacking at the major archive/review sites. In most cases, it has been simpler for the site/ reviewer(s) to list the major functions of the product under discussion, allocate a number (usually between 3 and 5) and then leave you, the user, to determine whether it's worth downloading a trial copy. Very seldom does one find truly personal comments such as "cumbersome to do x, y, z" or "merging files is very intuitive".
As the authors of this paper found when reviewing search engines for AusWeb96, one of the most important criteria for choosing a preferred utility is quite subjective and has to do with design features rather than actual functionality. The way in which people organise their "world"-- be it office, desk, computer desktop-- will have a significant influence on the bookmark utility(s) which they choose. If a person really likes the Windows Explorer hierarchical/tree "view", then they will be comfortable with a bookmark program which adopts the same approach (QuikLink Explorer). On the other hand, if they tend to organise themselves in terms of graphics/folders, then a program such as AcqURL may be more suitable.
A search was done on several of the major archives/reviewing sites for Windows/NT and MacOS. In nearly all instances, an evaluation copy of the bookmark utilities listed at those sites was downloaded and tested. In the case of PC software, the reviewer used Netscape Navigator 4.08 as the principal browser and Microsoft Internet Explorer 3.0 as the secondary browser. In the case of MacOS, the reviewer used Netscape Communicator 4.06 as the principal browser and Microsoft Internet Explorer 3.0 as the secondary browser.
The reviewers have attempted to personalise the entries in the web-enabled database by sharing their own experiences and impressions of actually using these tools, rather than merely focusing on features. Every attempt has been made to identify exactly how a bookmark utility might be useful -- assuming that the reviewers found it to be so!
The reviewing process and subsequent creation of a review database is really a "meta-project": it involves not only the actual process of reviewing and the writing reviews but also determining how to present that information to the reader in a meaningful way as the database will be publicly available.
Tools discussed will be reviewed for -
The paper is supported by a web enabled database [HREF 16] with links to reviews and copies of the software or demonstrations where available. The two authors were located in Queensland and the ACT and coordinated their work through the database as well as email and a draft of this paper on the web. As they each used different HTML editors some formatting problems arose.
The database software used is FileMakerPro 4 [HREF 17]. This is a relational database which has a web server built in that can access almost all functions of the database. A web interface is therefore available to the database without the need for cgi scripting or a separate web server. The database has in fact been utilised as a general purpose web server using the relational capabilities to assemble prepared headers and footers to maintain a large site [MORGAN [HREF 18]].
Those features that are not supported could be accessed via cgi scripts on the server (a Macintosh) as FileMaker Pro is fully scriptable via AppleScript. So far it has not been necessary to do this, although support to edit multi-valued fields via the web may require it. FileMaker Pro also has a Windows version.
The database web server responds to embedded propriety commands in HTML to initiate its functions. These are in a proprietary markup language called CDML (Claris Dynamic Markup Language). A companion product (although Apple's recent divestment of FileMaker Pro confuses this), Claris Home Page, supports the use of CDML tags and integrates it with editing HTML and JavaScript. Simple tools and collections of styles are provided to enable the automatic generation of HTML pages to access a database and provide the ability to edit with authentication.
Do we have anything to learn from the Library community? After all they are heavily into the business of organising access to information objects. There is a considerable amount of experimentation taking place. The subject tree approach is being actively developed by the UK ROADS [HREF 19]project which provides one of the best examples. At OCLC the NetFirst [HREF 20] database gives an example of treating internet resources like print material and cataloguing them. Within the MARC specification [HREF 21] which libraries use to hold and transfer cataloguing information there is a field, #856 [HREF 22], which is designed to hold URL information. But few libraries now can afford to engage in cataloguing material from scratch, instead they buy records from other institutions. Tennant [HREF 23] thinks that we should be looking at online bibliography and that expensive cataloguing is a wasted effort. The Dublin Core community are providing an alternative approach which may come into its own once the Resource Description Format (RDF [HREF 24]) gets widely deployed in the form of independent servers providing indexing information for remote objects. Barry [HREF 25] has speculated on some of these developments and their future path.
Another approach to organising URL's is not to! Instead you can rely on personal indexing robots to do the task for you. Provide them with the URLs you want organised and they will trawl the sites indexing every word on the page (or pages) in a common directory which then provides you with your own search engine index to your chosen sites. There are ethical problems concerning this. Frequent updates to your index will increase traffic to the sites adding an unwanted load and distorting their usage statistics. If such a practice became common it might add to traffic globally. Two tools were examined by the authors -
but these are not discussed further in the context of this paper.
The main archives for PC tools are software archival collections.
Main tools reviewed are-
[Note: please consult the (1) authors' database [HREF 33] for the actual reviews and (2) updated version [HREF 34] of this paper for ongoing additions to this list.]
In looking at the major utilities reviewed to date, there are several potentially useful ways to categorise, i.e. compare, them:
|
Internet-Based |
Your Own Machine |
|
Bookmarks2Go |
All the others! |
|
MURL |
|
|
MyLinx |
|
Cost
|
Free |
Cost (<US$20 ) |
Cost ( >US$20) |
|
Bookmark Wizard |
Bookmark Converter $8 |
AcqURL $25 |
|
Collectibles |
BookMarx $15 |
Alert $25 |
|
QuikLink Explorer |
URL Address Book $19.95 |
Address Organizer $43 |
|
URLMenu 98 |
URL Organizer 2 $15 |
Powermarks $24.95 |
|
|
|
WebQuick $39.95 |
|
|
|
WebTabs $34.95 |
|
|
|
|
|
Basic |
Medium |
High |
|
Bookmark Converter |
Alert |
AcqURL |
|
Bookmark Wizard |
URL Address Book |
Address Organizer |
|
BookMarx |
URLMenu 98 |
Powermarks |
|
Collectibles |
WebQuick |
QuikLink Explorer |
|
URL Organizer 2 |
|
WebTabs |
|
|
|
|
|
Windows Explorer hierarchy |
Category / Folder |
Other |
|
Bookmark Converter |
Alert rotateable index |
BookMarx |
|
QuikLink Explorer |
AcqURL folders (buttons) in column array |
Powermarks keywords indexed A-Z |
|
WebTabs |
Address Organizer alphabetical tabs |
|
|
|
Bookmark Wizard html document in same order as Favorites folder(s) |
|
|
|
Collectibles categories and subcategories |
|
|
|
URL Address Book categories |
|
|
|
URLMenu 98 folders |
|
|
|
URL Organizer 2 categories and subcategories |
|
|
|
WebQuick similar to Windows Programs display |
|
|
|
|
|
The authors of this paper/project would be interested in feedback from readers as to what method(s) of comparison would be most useful to them.
Once can conclude from the above tables that price generally but not always indicates the degree of functionality. The two notable exceptions are (1) QuikLink Explorer which has high functionality and has a "freeware edition" and (2) WebQuick which is one of the most expensive and yet has "basic to medium" functionality. WebQuick appears to this reviewer to have been developed before browsers such as Netscape offered "History" (for viewing recent pages) as well as more sophisticated handling of bookmarks. In addition, the "rave reviews" at the WebQuick home site date from 1996 which rings some alarm bells! Unfortunately Web "first-timers" may happily purchase this overpriced product unaware of these aspects.
In examining design features, it would seem that bookmark tools tend to replicate (in terms of display) what the user is already familiar with -- either a Windows Explorer hierarchical / tree structure or folders/categories containing nested folders, which is how browsers such as Netscape and Internet Explorer handle storing URL links. Of the tools examined to date, only Powermarks has moved outside the square; it uses a flat, hierarchical list and replaces folders with keywords. [Note: BookMarx serves a special function --reporting on validity of links; display aspects have little bearing and so it has been grouped with Powermarks under "Other".]
The kind of functionality/features which one can find in current market offerings (but not necessarily in a single product) for PCs include:
The main archives for MacOS tools are reviewing sites and software archival collections.
Main tools reviewed are -
A similar analysis as that used for Windows tools follows.
|
Free |
Cost (<US$20 ) |
Cost ( >US$20) |
|
BookmarksExport |
Bookmark Organiser $US10 |
URL Manager Pro $US25 |
|
Dragnet |
|
WebArranger $A149 Discontinued? |
|
URL Clerk |
|
|
|
URL Collector stack |
|
|
|
URL Forward |
|
|
|
URL Grabber |
|
|
|
What URL?! |
|
|
|
Basic |
Medium |
High |
|
BookmarksExport |
URL Collector stack |
Dragnet |
|
Bookmark Organiser |
URL Database |
URL Manager Pro |
|
Netbook |
URL Grabber |
WebArranger |
|
URL Clerk |
What URL?! |
|
|
URL Forward |
|
|
|
|
|
|
|
Folder hierarchy format |
Database format |
|
Others |
URL Collector stack |
|
|
URL Database |
|
|
URL Grabber |
URL Manager Pro was the outstanding tool for the Mac although for those with a yen for programming the Hypercard and FileMaker Pro based tools could be developed further.
[Note: please consult the (1) authors' database [HREF 39] for the actual reviews and (2) updated version [HREF 40] of this paper for ongoing additions to this list.]
This paper was originally conceived as a review of URL managers but it became clear that it should also look at other sources if information and how these are controlled. Our conclusions are -
The database of software reviewed will be available for public suggestions and searching at http://ningaui.anu.edu.au:591/helpers/index.html [HREF 41]. Anybody wishing to draw our attention to new software or search for a particular program can access it. An expanded bibliography will be available as well as an "updated version" of this paper which will be modified as time permits.
Joanna Richardson and Tony Barry, 1999. The authors assign to Southern Cross University and other educational and non-profit institutions a non-exclusive licence to use this document for personal use and in courses of instruction provided that the article is used in full and this copyright statement is reproduced. The authors also grant a non-exclusive licence to Southern Cross University to publish this document in full on the World Wide Web and on CD-ROM and in printed form with the conference papers and for the document to be published on mirrors on the World Wide Web.