http:^^www.cs.cornell.edu^info^people^lagoze^papers^www.html
来自「This data set contains WWW-pages collect」· HTML 代码 · 共 776 行 · 第 1/3 页
HTML
776 行
two page images, which made correct page mapping impossible in thedocument server. While it was easy to find files with this problem(by reading the height and width from the TIFF header with apublically available TIFF package <!WA11><!WA11><!WA11><!WA11><!WA11><!WA11><!WA11><!WA11><!WA11><!WA11><!WA11><!WA11><!WA11><!WA11><!WA11><!WA11><!WA11><!WA11><!WA11><!WA11><a HREF="#Leffler">[Leffler]</a>),reasonably quick correction required handcrafting c-code tosplit the files. Even with the handcrafted code, the location andcorrection process took over a week of compute time on a powerful workstation.<p>In addition to manual scanning of documents, we also had to manuallyenter the RFC-1357 bibliographic files. While it would have been easy to writetranslators between RFC-1357 and other common bibliographic formatssuch as BibTex, refer, etc, a consistent electronic bibliographicformat was not available for all the TR's.<h2> The Web is an imperfect document viewing technology</h2>Basing our system on the World Wide has had both benefits andshortcomings. The obvious benefit is wide availability overpublically available browsers. The shortcoming is that HTML, HTTP,and Web browsers lack a number of features important for digitaldocument display and navigation. In this section we enumerate thesefeatures with the goal of inspiring discussion and enhancement of thetechnologies by the Web community.<h3>Facilities for display of compound documents</h3>The Web has insufficient mechanisms for displaying documents thatconsist of multiple textual and non-textual parts. In the electronicmail world, this issue is addressed by MIME (Multipurpose InternetMail Extensions) [MIME]. Although HTTP uses MIME typing to allowbrowsers to map to the proper viewer for a document, documents areallowed to have only a single MIME type - multipart MIME is not one ofthem. The only facility for multi-format documents is the ability toembed images (either GIF or X Bitmap) in an HTML document. Yet thereare gross inefficiency problems with image embedding since the HTTPbrowser must initiate an HTTP GET message for each embedded image.For a document with many embedded images, this can lead tounacceptable document download times. Furthermore, there are othertypes that one might like to embed in documents; for example, <code>MPEG</code> clips. <h3>Ability to display in-line TIFF images</h3>Among the many digital image formats (GIF, JPEG, PBM, etc.), TIFF isthe most flexible and extensible. The TIFF specification isconstantly evolving with the latest being Revision 6.0, finalized in1992 <!WA12><!WA12><!WA12><!WA12><!WA12><!WA12><!WA12><!WA12><!WA12><!WA12><!WA12><!WA12><!WA12><!WA12><!WA12><!WA12><!WA12><!WA12><!WA12><!WA12><a HREF="#TIFF"> [TIFF] </a>. The most significant evolution,from the standpoint of reducing network bandwidth in image transfers,is the growing number of compression schemes for TIFF images. Theability to display in-line TIFF images in HTML documents would takefull advantage of this rich de-facto image standard and permitimmediate display of images produced by scanners, fax machines, andmost paint and photo-retouching programs without computationallyexpensive conversion to GIF format.<h3>Arbitrary rectangle section from the client</h3>Viewing of document images on the Web would be greatly enhanced if theuser of a client were able to select across an arbitrary rectangle inthe image and transmit the selected coordinates back to the server.The server could then retransmit a "zoomed" image of the selectedimage, if the higher resolution were available (which it often is in ahigh resolution TIFF image). Image zooming is an important featurewhen the image being viewed is a document page that contains figuresor tables with small fonts.<h3>Client feedback on display capabilities</h3>A main contribution of Dienst is that it supports the notion ofmultiple formats for the same document. The user can select among theavailable formats and use the view appropriate for that format. Wewould prefer to, at the server end, chose the "best" format to displayon the respective client. This would be possible if the client HTTPrequest contained information on the display capabilities of theclient system, especially display depth and size.<h3>Authentication</h3>The ability to restrict who is able to access a document is anessential feature of a production digital library. While our systemis intended for non-commercial publishing, limiting access is requiredeven in this domain; say, for example, documents that should only beread by members of a campus community or employees of a corporation.To do this, we require that server be able to guarantee the identityof those making protocol requests.<h2>Summary</h2>We have described a system, Dienst, that simplifies documentpublishing on the Internet. This system makes two importantcontributions.<p>First, Dienst provides a uniform protocol for search, retrieval, anddisplay of documents. This protocol addresses a flexible documentmodel where each document has a unique name, can be in multipleformats, and consists of a set of named parts. These parts can bephysical, such as pages, or logical, such as chapters and tables. Inaddition, the protocol allows full interoperability betweendistributed digital library servers. The result is that the user seesa single virtual document collection.<p>Second, Dienst provides a set of tools that permit easy management ofa digital library. These tools automate document submission, permit adocument librarian to manage the collection, and facilitate theproduction of archival hardcopy.<p>We plan over the next year to build on this technology in a number ofways. Installation of the digital library server is too difficult. Weintend to implement tools that will "auto-configure" the server.The search engine in the current implementation is primitive.We intend to include more advanced search engines, for examplefull-text search, to make documentdiscovery in a collection more powerful and easier. The currentstrategy of conducting a parallel search over all servers does notscale over a very large number of servers. We intend to usemeta-information about individual document servers to improve thesearch strategy. With this facility, one could, for example, choose to search only thoselibraries that have a high probability of containing computer sciencedocuments. We plan to examine and possibility incorporate current work oncopyright servers, so Dienst might be used for commercial documents.Finally, we hope to use some of the current work inlocation-independent identifiers to refine the method by whichdocuments on the net are addressed in Dienst.<H2>Acknowledgements</H2>This work was supported in part by the Advanced Research ProjectsAgency under Grant No. MDA972-92-J-1029 with the Corporation forNational Research Initiatives (CNRI). Its content does notnecessarily reflect the position or the policy of the Government orCNRI, and no official endorsement should be inferred. This work wasdone at the Design Research Institute, a collaboration of XeroxCorporation and Cornell University, and at the Computer ScienceDepartment at Cornell University.<h2>References</h2><a name="Cohen">[Cohen]</a>Danny Cohen. A Format for E-mailing Bibliographic Records <!WA13><!WA13><!WA13><!WA13><!WA13><!WA13><!WA13><!WA13><!WA13><!WA13><!WA13><!WA13><!WA13><!WA13><!WA13><!WA13><!WA13><!WA13><!WA13><!WA13><aHREF="file://nic.merit.edu/documents/rfc/rfc1357.txt">RFC-1357 </A><p><a name="Dienst">[Dienst]</a>James R. Davis, Carl Lagoze. A protocol and server for a distributeddigital technical report library.Cornell University Computer Science Department Technical Report <!WA14><!WA14><!WA14><!WA14><!WA14><!WA14><!WA14><!WA14><!WA14><!WA14><!WA14><!WA14><!WA14><!WA14><!WA14><!WA14><!WA14><!WA14><!WA14><!WA14><aHREF="http://cs-tr.cs.cornell.edu/TR/CORNELLCS:TR94-1418">94-1418</A>,June 1994.<p><a name="DIENSTPROT">[DIENSTPROT]</a>James R. Davis, Carl Lagoze. Dienst, A Protocol for a Distributed Digital DocumentLibrary. Internet Draft.<p><a name="EZPUB">[EZPUB]</a>Cornell Information Technologies. How to Use EZ-PUBLISH and the Docutech Printer atCornell Information Technologies. November 24, 1993.<p> <a name="Leffler">[Leffler]</a>Sam Leffler. Public TIFF package. Available via ftp from <!WA15><!WA15><!WA15><!WA15><!WA15><!WA15><!WA15><!WA15><!WA15><!WA15><!WA15><!WA15><!WA15><!WA15><!WA15><!WA15><!WA15><!WA15><!WA15><!WA15><a HREF="ftp://sgi.com/graphics/tiff/v3.2beta.tar.Z">sgi.com/graphics/tiff/v3.2beta.tar.Z </a>.<p> <a name="MIME">[MIME]</a>Nathaniel S. Borenstein, Ned Freed.MIME (Multipurpose Internet Mail Extensions). <!WA16><!WA16><!WA16><!WA16><!WA16><!WA16><!WA16><!WA16><!WA16><!WA16><!WA16><!WA16><!WA16><!WA16><!WA16><!WA16><!WA16><!WA16><!WA16><!WA16><a HREF="file://nic.merit.edu/documents/rfc/rfc1521.txt"> RFC-1521 </a>.<p> <a name="PBMPLUS">[PBMPLUS]</a>Jef Poskanzer.Extended Portable Bitmap Toolkit. Available from many anonymous FTP sites including<!WA17><!WA17><!WA17><!WA17><!WA17><!WA17><!WA17><!WA17><!WA17><!WA17><!WA17><!WA17><!WA17><!WA17><!WA17><!WA17><!WA17><!WA17><!WA17><!WA17><a HREF="ftp://ftp.ee.utah.edu/pbmplus">ftp.ee.utah.edu</a>.<p><A NAME="Putz">[Putz]</A>Steve PutzDesign and Implementation of the System 33 Document ServiceXerox PARC<!WA18><!WA18><!WA18><!WA18><!WA18><!WA18><!WA18><!WA18><!WA18><!WA18><!WA18><!WA18><!WA18><!WA18><!WA18><!WA18><!WA18><!WA18><!WA18><!WA18><A HREF="http://www.xerox.com/PARC/dlbx/other-papers/system33.ps">P93-00112</A>, 1993<p><A NAME="SIFT">[SIFT]</A>Online serviceat <!WA19><!WA19><!WA19><!WA19><!WA19><!WA19><!WA19><!WA19><!WA19><!WA19><!WA19><!WA19><!WA19><!WA19><!WA19><!WA19><!WA19><!WA19><!WA19><!WA19><A HREF="http://sift.stanford.edu/">http://sift.stanford.edu.</A><p><A NAME="TIFF">[TIFF]</A>Aldus Corporation. TIFF Revision 6.0 Specification.<p><a name="URC"> [URC] </a>Michael Mealling. Encoding and Use of Uniform ResourceCharacteristics. <!WA20><!WA20><!WA20><!WA20><!WA20><!WA20><!WA20><!WA20><!WA20><!WA20><!WA20><!WA20><!WA20><!WA20><!WA20><!WA20><!WA20><!WA20><!WA20><!WA20><a HREF="http://www.gatech.edu/iiir/urc2.paper.html">Internet Draft.</A><p><A NAME="WATERS">[WATERS]</A>Kurt J. Maly, Edward A. Fox, James C. French and Alan L. Selman. Wide Area Technical Report Server.Published online in <!WA21><!WA21><!WA21><!WA21><!WA21><!WA21><!WA21><!WA21><!WA21><!WA21><!WA21><!WA21><!WA21><!WA21><!WA21><!WA21><!WA21><!WA21><!WA21><!WA21><A HREF="http://www.cs.odu.edu/WATERS/WATERS-paper.ps">http://www.cs.odu.edu/WATERS/WATERS-paper.ps</A><h2>Biographies</h2><b>Jim Davis</b> works for Xerox at the <!WA22><!WA22><!WA22><!WA22><!WA22><!WA22><!WA22><!WA22><!WA22><!WA22><!WA22><!WA22><!WA22><!WA22><!WA22><!WA22><!WA22><!WA22><!WA22><!WA22><AHREF="http://dri.cornell.edu">Design Research Institute</A>, anon-proprietary consortium at Cornell University which which seeksways to improve the engineering design process. He received a PhD in1989 from MIT's Media Technology Laboratory. His thesis, the <i>BackSeat Driver</i> was a computer program which provided spokendriving instructions to the operator of a car in real-time. Prior tothat, he worked in research and development at a number of placesincluding Atari's Cambridge Research Laboratory. At the DRI he workson developing electronic corporate memory. His most recent projectis a system for <!WA23><!WA23><!WA23><!WA23><!WA23><!WA23><!WA23><!WA23><!WA23><!WA23><!WA23><!WA23><!WA23><!WA23><!WA23><!WA23><!WA23><!WA23><!WA23><!WA23><AHREF="http://dri.cornell.edu/pub/davis/annotation.html">shared group annotation</A> using the World Wide Web. He also playselectric bass and is learning Dutch.<p><b>Carl Lagoze</b> works for the <!WA24><!WA24><!WA24><!WA24><!WA24><!WA24><!WA24><!WA24><!WA24><!WA24><!WA24><!WA24><!WA24><!WA24><!WA24><!WA24><!WA24><!WA24><!WA24><!WA24><AHREF="http://www.cs.cornell.edu"> Computer Science Department </A> at<!WA25><!WA25><!WA25><!WA25><!WA25><!WA25><!WA25><!WA25><!WA25><!WA25><!WA25><!WA25><!WA25><!WA25><!WA25><!WA25><!WA25><!WA25><!WA25><!WA25><A HREF="http://www.cornell.edu"> Cornell University </A> as a SeniorSoftware Engineer in the <!WA26><!WA26><!WA26><!WA26><!WA26><!WA26><!WA26><!WA26><!WA26><!WA26><!WA26><!WA26><!WA26><!WA26><!WA26><!WA26><!WA26><!WA26><!WA26><!WA26><AHREF="http://cs-tr.cs.cornell.edu/Info/cstr.html"> CSTR project </A>.He received a Master of Software Engineering from the Wang Instituteof Graduate Studies in 1987. After receiving his degree he worked inboth academia and the commercial world developing tools for thegeneration of language-specific editors. Over the past two years hehas discovered the joys of digital libraries and the fascinating worldof information capture and access. From the view of his non-technicalfriends, he is doing "something on that information superhighway."Mr. Lagoze is also the proud parent of the cutest baby ever and anavid cyclist and canoeist.<p><b>contact author:</b> davis@dri.cornell.edu 607-255-1134</BODY></HTML>
⌨️ 快捷键说明
复制代码Ctrl + C
搜索代码Ctrl + F
全屏模式F11
增大字号Ctrl + =
减小字号Ctrl + -
显示快捷键?