WHAT IS URLib?

by Gerald Jean Francis Banon
<urlibservice@gmail.com>

First version: August 1995.
Last change: June 2024.
Portuguese version.

CONTENT

1. URLib and URLibService
2. URLibService features
3. Repository name formation rule
4. Persistent access
5. Original document and copyrights
6. URLibService releases, license and download
7. Examples of sites using the URLibService
8. Quotation
9. References


1. URLib AND URLibService

URLib is an acronym for Uniform Repository for a Library.
 
The so-called URLib platform is a computational environment for hosting
hyperdocuments in order to ensure fully persistence of links.

Think of the URLib platform as a universal digital library platform with
a kind of smart bookshelf in the sense that there is always a single space where
any new document can fit, regardless of who deposits it.

The concept of Uniform Repository was first introduced in 1995
to solve the problem of integrating electronic documents produced
by people using independent file systems.

Each uniform repository (or simply repository) contains a unique electronic
resource called document.

An electronic document can be any collection of files and directories under
a same directory called document folder.

The collection of an Archive is the set of all reachable documents deposited
in uniform repositories. It can be distributed over the net along a predefined
set of other Archives, each of which exposing their so-called local collection.

URLibService is the name of the software which manages the URLib platform.

At the present stage of URLibService development [2] each Archive use
Apache as a Web server.

By using URLibService, one can create new repositories to
deposit new documents in a local collection and install new copies of existing
repositories/documents from other local collections.


2. URLibService FEATURES

You can use URLibService to:

- identify universally the documents by means of IBI (Internet Based Identifier);
- organize the storage of the documents in the file system of the computers;
- deposite and update the documents remotely using browsers;
- store the documents in any file format (HTML, PDF, DOC, ...);
- post the documents on the Web;
- retrieve the documents on the Web through keywords;
- search physically distributed URLib sites with a single query;
- generate search reports, tables, summaries and author indexes;
- make easy the download of documents with their metadata;
- restrict the Web access of each document through ip or password;
- share a given local collection with other ones;
- install a copy of a document from another local collection;
- control the document versions;
- localize on the Web the original version of a document;
- authenticate the copies of a document;
- transfer a document to a new local collection;
- register a document as being original;
- display the appropriate copyright warning for each document;
- display the document publisher stamp;
- store and display the access statistics of each document;
- allow relative links between documents, turning the access persistent;
- resolve universally links defined in terms of IBI (persistent URL); 
- guarantee the link robustness in cases of local collection fusion or division;
- allow user language preference sensitive links;
- create persistent link to the last edition of a document;
- warn automatically the documents under construction;
- display the document metadata in BibTeX, Refer (EndNote) and BibINPE formats;
- post the document metadata through the Open Archives Initiative (OAI)
  Protocol for Metadata Harvesting;
- assign documents to reviewers and select them based on reviews.



3. REPOSITORY NAME FORMATION RULE

The repositories are called uniform because they are created by using
a same construction rule all over the local collections. In this way,
any of them can be installed within any local collection without
name conflict.

A uniform repository is a sequence of four successive directories
whose names are respectively:
- a subdomain name,
- a domain word, and eventually a port number, separated by "." or "@",
- a year and
- a month, day, hour, minute and eventually second separated by ".".

More precisely, the repository name syntax is:

<subdomain-name>/
<domain-word>[.|@<port-number>]/
<year>/
<month>.<day>.<hour>.<minute>[.<second>]

Here are two examples of repository names:

<dpi.inpe.br/banon/1995/09.01.10.50>
<dpi.inpe.br/banon-pc2@1905/2005/12.07.19.19>.

In order to guarantee that the above path is unique among all paths
created up to now and all paths that will be created in the future,
the following rules have to be observed.

- The string <domain-word>.<subdomain-name> must refer to the domain name
  of the computer hosting the local collection, and the <port-number> must
  refer to the http port given access to this local collection.
  
  In the above examples, the repositories were created within local
  collections whose access URLs were (at the repository creation time),
  respectively, banon.dpi.inpe.br and banon-pc2.dpi.inpe.br:1905
  
- The <year> and <month>.<day>.<hour>.<minute>[.<second>],
  expressed numerically, must correspond to the GMT date/time of the repository
  creation, in order to solve the problem of local collection migration
  between time zones.

More details about repository name formation can be found in [9]


4. PERSISTENT ACCESS

URLibService allows the construction of stable hypertexts [4]
whose parts can be copied or moved from one URLib site to
another without the necessity of linking change or calling a
global name service up.

Additionally, URLib Service offers a global name service that resolves
persistent URL written in terms of IBI (Internet Based Identifier) [9].

Hence, one of the highlights of URLibService is to offer a nice
solution to the CITATION LINKING problem. Wherever is physically a
document in the URLib collection, its access is persistent.


5. ORIGINAL DOCUMENT AND COPYRIGHTS

Depositing a document in a repository is a publication act.

In order to respect the copyrights and not duplicate entries, ONLY
copyright holders (or authorized persons) should make up the
repositories. In other words, the possible local administrators are:

- the first author of pieces of work (if the copyrights have not been
  transferred to a publisher, otherwise with the publisher permission);

- the publisher of pieces of work (once the copyrights have been
  transferred by their author(s));

- someone with the copyright holder permission.

A document is considered original if it is in the copyright holder
local collection.


6. URLibService RELEASES, LICENSE AND DOWNLOAD

The first release of URLibService [1] was launched in
August 1995. This release has been used until April 2001.

The second release of URLibService [2] was launched in
August 1998.

The use of URLibService is free for public institutions and
nonprofit organizations [5].

URLibService is multiplatform, running under the following
operational systems: Windows, SunOS and Linux [3].

This program is protected against unauthorized use by
Brazilian Laws n. 9.609, of february 19th, 1998,
ruled by Decree n. 2.556, of April 20th, 1998,
combined with Law n. 9.610, of february 19th, 1998,
being registered at "Instituto Nacional da Propriedade Industrial – INPI"
(Decree nº 2.556/98, art. 1º) under n. 00072253.


7. EXAMPLES OF SITES USING THE URLibService

At the present time, more than twenty URLib sites are installed.
Some are mirrored from others. The most interesting ones are:

- The "INPE Library" site
<http://www.inpe.br/biblioteca>
which gives access to references and/or full documents of pieces of
work (like thesis and conference papers) published by the
"Instituto Nacional de Pesquisas Espaciais" (INPE) [6][8].

- The "Biblioteca Digital - SBSR" site
<http://urlib.net/ibi/83LX3pFwXQZ5Jpy/CxGU3>
which gives access to the proceedings of all conferences held [7].


8. QUOTATION

"Toute oeuvre est l'oeuvre de bien d'autres choses qu'un "auteur""
Paul Valéry (1871-1945)
(Le Monde, Selection Hebdomadaire Edition Internationale, Jeudi 27/7/95, p.10)

"Any work is the work of many other things than an "author""
(translation by G. J. F. Banon)


9. REFERENCES


[1] Banon, Gerald Jean Francis :1998:
    Uniform repositories for a digital library - URLib
    URL: <http://urlib.net/dpi.inpe.br/banon/1997/08.14.12.32>


[2] Banon, Gerald Jean Francis ::
    URLibService User's Guide
    URL: <http://urlib.net/iconet.com.br/banon/2000/12.31.20.38>


[3] Banon, Gerald Jean Francis ::
    Downloading the URLib Environment
    URL: <http://urlib.net/iconet.com.br/banon/2000/12.30.22.40>


[4] Banon, Gerald Jean Francis ::
    Hiperdocumentos versus URLib
    URL: <http://urlib.net/dpi.inpe.br/banon/2002/10.10.08.39>


[5] Banon, Gerald Jean Francis ::
    URLibService Copyright
    URL: <http://urlib.net/dpi.inpe.br/banon/2000/03.04.00.05>


[6] Banon, Gerald Jean Francis; Ribeiro, Marciana Leite & Banon, Lise Christine :2004:
    Preservação digital da memória técnico-científica do INPE
    URL: <http://urlib.net/dpi.inpe.br/lise/2004/03.02.15.20>


[7] Banon, Lise Christine & Banon, Gerald Jean Francis ::
    "Ferramentas on-line" associadas a uma biblioteca digital para publicação em eventos:
    Projeto XII SBSR
    URL: <http://urlib.net/sid.inpe.br/ePrint@80/2005/10.05.21.15>


[8] Banon, Gerald Jean Francis ::
    Bibiloteca Digital da Memória Tecnico-Científica do INPE
    URL: <http://urlib.net/dpi.inpe.br/banon-pc2@1905/2005/12.07.19.19>


[9] Banon, Gerald Jean Francis :2011:
    Identificador com base na Internet (IBI): Sistema de identificação
    URL: <http://urlib.net/dpi.inpe.br/banon-pc2@1905/2005/12.07.19.19>