Published Papers

Below is a list of my (David Lomet's ) published papers. The papers with links are available electronically. Reprints of the remaining papers are available from me. Send mail to (lomet@microsoft.com ). The papers are grouped by major research area, e.g., concurrency control, access methods, etc.

Transactions

In 1975, while on sabbatical at the University of Newcastle-upon-Tyne, I invented the notion of an atomic action. This was shortly seen to be essentially the same notion as a database transaction, which had been invented the same year by researchers at the IBM San Jose lab. The transaction concept is one of the fundamental notions underlying work in databases, operating systems, and distributed computing for the last 20 years. The paper that I wrote on this has been included in two anthologies of collected papers on reliability and distributed systems. I have subsequently worked on distributed transaction commit protocol optimization. This has included optimizations even when transactions might not have quiesced, using timestamps, and management of the protocol ``database" so as to reduce message cost and log cost simultaneously (with Butler Lampson). Most recently, I have worked with Christian Jensen and Rick Snodgrass on transaction timestamping for temporal databases.

  1. Lomet, D., Vagena, Z., and Barga, R. Recovery from "Bad" User Transactions. SIGMOD Conference, Chicago, IL (June 2006) (to appear) pdf, .150MB
  2. Lomet, D., Barga, R., Mokbel, M., Shegalov, G., Wang, R., and Zhu, Y. Transaction Time Support Inside a Database Engine. ICDE Conference, Atlanta, GA (April 2006) PDF, .30MB
  3. Lomet, D., Snodgrass, R., and Jensen, C. Using the Lock Manager to Choose Timestamps. IDEAS Conference, Montreal, Canada (July 2005) 357-368. PDF, .27MB
  4. Lomet, D., Barga, R., Mokbel, M., Shegalov, G., Wang, R., and Zhu, Y. Immortal DB: Transaction Time Support for Sql Server. SIGMOD Conference, Baltimore, MD (June 2005) 939-941. pdf, .283MB
  5. Jensen, C., and Lomet, D. Transaction timestamping in (temporal) databases. VLDB Conference, Rome, Italy (Sept. 2001) 441-450 PDF,.12MB
  6. Lampson, B. and Lomet, D. A New Presumed Commit Optimization for Two Phase Commit. VLDB Conference, Dublin, Ireland (Aug. 1993) 630-640. postscript, .19MB
  7. Lomet, D. Using Timestamping to Optimize Two Phase Commit. PDIS Conference, San Diego, CA (Jan. 1993) 48-55. postscript, .17MB
  8. Lomet, D. Process structuring, synchronization, and recovery using atomic actions. ACM Conf. on Language Design for Reliable Software, Raleigh, NC SIGPLAN Notices 12,3 (Mar 1977) 128-137. in Reliable Distributed System Software, J. A. Stankovic (ed), IEEE Computer Society Press (1985) and in Reliable Computer Systems, S. K. Shrivastava (ed), Springer-Verlag (1985) PDF, .52MB

back to the top

Recovery

My recovery work was originally prompted by the need for Digital's Rdb database system to provide redo logging in a data sharing context. This led to recovery using multiple logs. Combining access methods with recovery, my work (with Betty Salzberg) showed how to use a history database, organized as a TSB-tree, to provide media recovery should the current database be corrupted. How to combine high concurrency with recovery for B-trees was part of this work. Multilevel transactions are one of the truly important recent (1980's) notions in transaction models, being a generalization of nested transactions that permits substantially increased concurrency. Multilevel transactions have been used informally previously. Weikum made them explicit. But a good general recovery mechanism for them was not known. My MLR recovery method solved this in a particularly elegant way. How to replay the redo log against the stable database state that survives a crash has not been well understood. Recent work (with Mark Tuttle) formalizes redo recovery, prescribes what a cache manager must do to keep the database recoverable, and specifies the properties needed of logged operations that permit them to be successfully replayed. This is seminal work in an area that is very important for fault tolerance, not just for transactional systems but for recoverable editing, etc. My recent work has been in applying the redo understanding resulting from formalizing redo recovery to the recovery of applications as well as databases. This requires mapping application execution to log operations, and then providing a cache manager that correctly posts changed data objects to the stable state so as to keep the system state recoverable. It also requires a recovery process that can decide exactly which operations to execute and which to bypass in recovering the state.  This recent work has been in the context of the Phoenix project at Microsoft Research, with principal collaborators being Roger Barga and Gerhard Weikum.  Two prototype Phoenix systems have been built providing different flavors of application resilience to failures.

  1. Lomet, D. Persistent Middle Tier Components without Logging. IDEAS Conference, Montreal, Canada (July 2005) 37-46. PDF, .09MB
  2. Lomet, D. Robust Web Services via Interaction Contracts. TES'04 Workshop (2004) 1-14. pdf, .09MB
  3. Barga, R.,Lomet, D., Shegalov, G., and Weikum, G. Recovery Guarantees for Internet Applications. ACM Trans. on Internet Technology (August 2004) 289-328 pdf, .36MB
  4. Barga, R., Chen, S. and Lomet, D. Improving Logging and Recovery Performance in Phoenix/App. ICDE Conference, Boston, MA (March 2004) 486-497 Word, .23MB
  5. Barga, R., Lomet, D., Paparizos, S., Yu, H., and Chandresekaran, S. Persistent Applications Via Automatic Recovery. IDEAS Conference, Hong Kong (July 2003) 258-267 pdf, .147MB
  6. Lomet, D. and Tuttle, M. A Theory of Redo Recovery. SIGMOD Conference, San Diego, CA (June 2003) 397-406 pdf, .132MB
  7. Shegalov, G., Weikum, G., Barga, R., Lomet, D. EOS: Exactly-Once E-Service Middleware. VLDB Conference, Hong Kong, China (August 2002) 1043-1046. PDF, .2MB
  8. Barga, R., Lomet, D. Phoenix Project: Fault Tolerant Applications. SIGMOD Record 31, 2 (June 2002) 94-100. PDF, .13MB
  9. Barga, R., Lomet, D. and Weikum, G. Recovery Guarantees for Multi-tier Applications. ICDE Conference, San Jose, CA (March 2002) 543-554 Word document, .34MB
  10. Barga, R. and Lomet, D. Measuring and Optimizing a System for Persistent Database Sessions. ICDE Conference, Heidelberg, Germany (April 2001) 21-30 pdf, .17MB
  11. Lomet, D. High Speed On-line Backup When Using Logical Log Operations.  ACM SIGMOD Conference, Dallas, TX (May, 2000) 3-45 PDF, .22MB
  12. Barga, R., Lomet, D., Baby, T., and Agrawal, S. Persistent Client-Server Database Sessions. EDBT Conference, Lake Constance, Germany (Mar. 2000) 462-477. Word document, .331MB
  13. Lomet, D. and Tuttle, M. Logical Logging to Extend Recovery to New Domains. ACM SIGMOD Conference Philadelphia, PA (June, 1999) 73-84. PDF,.168MB
  14. Barga,R. and Lomet, D. Phoenix: Making Applications Robust.(demo paper) ACM SIGMOD Conference, Philadelphia, PA (June, 1999) 562-564 PDF, .06MB
  15. Lomet, D.B. and Weikum, G. Efficient Transparent Application Recovery in Client-Server Information Systems. (Best Paper Award) ACM SIGMOD Conference, Seattle, WA (June 1998) 460-471. PDF, .135MB; Technical Report with appendices Word document, .570MB presentation slides PowerPoint .223MB
  16. Lomet, D.B. Persistent Applications Using Generalized Redo Recovery. IEEE ICDE Conference, Orlando, FL (Feb. 1998) 154-163 PDF,.12MB
  17. Lomet, D.B. Advanced Recovery Techniques in Practice. in Recovery Mechanisms in Database Systems (V. Kumar and M. Hsu, eds.) Prentice Hall PTR 1998 postscript, .15MB
  18. Lomet, D.B. Application Recovery: Advances Toward an Elusive Goal. Workshop on High Performance Transaction Systems (HPTS 97), Asilomar, CA (September, 1997) Word document, .19MB
  19. Lomet, D. and Tuttle, M. Redo Recovery after System Crashes. VLDB Conference, Zurich, Switzerland (Sept. 1995) 457-468. and in Advanced Recovery Techniques in Practice, Prentice Hall PTR, 1998 postscript, .28MB
  20. Lomet, D. and Salzberg, B. Exploiting a History Database for Backup. VLDB Conference., Dublin, Ireland (Aug. 1993) 380-390. postscript, .23MB
  21. Lomet, D. MLR: A Recovery Method for Multi-Level Systems. SIGMOD Conference, San Diego, CA (May 1992) 185-194. and in Advanced Recovery Techniques in Practice, Prentice Hall PTR, 1998 postscript, .25MB
  22. Lomet, D.B. Recovery for shared disk systems using multiple redo logs. Digital Tech. Report CRL90/4 (Oct 1990) Cambridge Research Lab, Cambridge, MA PDF, .13MB

back to the top

Concurrency Control

During the late 1970's, my work explored how database systems might provide deadlock avoidance (making deadlock impossible) rather than permitting deadlocks and recovering from them. This work grew out of earlier operating system work by Habermann and Holt. But with database system logical resources, a simple graph controls the system. Combining this potential delay graph with the wound-wait technique permits it to be applied to distributed systems. An IEEE TSE paper demonstrated how to avoid indefinite delay by partitioning resources into subsytems.  Subsequently, my concurrency control work had focused on improving concurrency via new lock modes. The important notion of combining lockable resources to reduce lock overhead permits concurrency improvements without increasing lock overhead. These ideas have been applied to key range locking for phantom prevention and to private lock management for reducing lock overhead in a data sharing system.

  1. Lomet, D. Private Locking and Distributed Cache Management. PDIS Conference, Austin, TX (Sept., 1994) 151-159 and Int'l Workshop on High Performance Transaction Systems Asilomar, CA (Sept. 1993) postscript, .21MB
  2. Lomet, D. Key Range Locking Strategies for Improved Concurrency. VLDB Conference, Dublin, Ireland (Aug. 1993) 655-664. postscript, .20MB
  3. Lomet, D.B. Subsystems of processes with deadlock avoidance. IEEE Trans. on Software Engineering SE-6, 3 (May 1980) 297-304. PDF, .55MB
  4. Lomet, D. Coping with Deadlock in Distributed Systems. IFIP Working Conference on Database Architecture, Venice, Italy. In Data Base Architecture, Bracchi and Nijseen (eds), North-Holland (June 1979), 95-105. PDF, .40MB
  5. Lomet, D. Multi-level Locking with Deadlock Avoidance. ACM Annual Conference, Washington, DC (December 1978), 862-867. PDF, .40MB
  6. Lomet, D. A Practical Deadlock Avoidance Algorithm for Data Base Systems. SIGMOD Conference, Toronto, Canada (August 1977), 122-127. PDF, .70MB

back to the top

Access Methods

Since the late seventies, I have authored a series of papers on access methods that are among the most widely read and cited papers in this area. These papers have included improved B-trees, improved hashing, and importantly, the bounded disorder access method, which combines B-trees and hashing (with Witold Litwin), yielding a combination that provides many of the advantages of both, i.e. very fast random key access and decent sequential performance. More recently, access methods for multi-attribute data, including temporal data, have become of wide interest. The hB-tree multi-attribute method and the TSB-tree temporal method (both with Betty Salzberg), are among the very small number of practical methods in these areas.  Finally, access methods must work in the context of transactional systems. This means that high concurrency must be provided while retaining the ability to recover the access method across system crashes. This has led (with Betty Salzberg) to one paper that extends the B-link tree method to provide recovery, and another that shows how historical data in a temporal index can be used to provide media recovery. Additional papers extending this work to multi-attribute access method were subsequently published (with Betty Salzberg and Georgios Evangelidis).  Betty and I (with collaborators) are currently exploring access methods for ``tree-versioned" data where the versioning is not strictly linear (as with temporal data).

  1. Salzberg, B., Jiang, L., Lomet, D., Barrena, M., Shan, J., and Kanoulas, E. A Framework for Access Methods for Versioned Data. EDBT Conference, Heraklion, Greece (March 2004) 730-747 )PDF, .237MB
  2. Lomet, D. Simple, Robust and Highly Concurrent B-trees with Node Deletion. ICDE Conference, Boston, MA (March 2004) 18-28 Word, .15MB
  3. Jiang, L., Salzberg, B., Lomet, D., and Barrena, M. The BTR-Tree: Path-Defined Version-Range Splitting in a Branched and Temporal Structure. SSTD Conference. Santorini, Greece (July 2003) 28-45 pdf, .268MB
  4. Lomet, D. The Evolution of Effective B-tree Page Organization and Techniques: A Personal Account SIGMOD Record 30, 3 (Sept. 2001) 64-69. Word, .09MB
  5. Jiang, L., Salzberg, B., Lomet, D., and Barrena, M. The BT-Tree: A Branched and Temporal Access Method. VLDB Conference. Cairo, Egypt (Sept. 2000) 451-460 postscript, .559MB
  6. Lomet, D. B-tree Page Size When Caching is Considered. SIGMOD Record 27,3 (Sept. 1998) 28-32. Word, .334MB
  7. Lomet, D. and Salzberg, B. Concurrency and recovery for index trees. VLDB Journal 6,3 (Aug. 1997) 224-240. postscript, .26MB
  8. Evangelidis, G., Lomet, D., and Salzberg, B. The hBPi-tree: A Concurrent and Recoverable Multi-attribute Index. VLDB Journal 6,1 (Feb. 1997) (1-25) postscript, 1.30MB
  9. Lomet, D. Replicated Indexes for Distributed Data. PDIS Conference. Miami, FL (Dec. 1996) (108-119) postscript, .19MB
  10. Evangelidis, G., Lomet, D. and Salzberg, B. The hB-Pi-tree: A Concurrent and Recoverable Multi-attribute Access Method. VLDB Conference, Zurich, Switzerland(Sept. 1995) 551-561. postscript, .60MB
  11. Lomet, D. and Salzberg, B. Transaction-Time Databases. Chapter in Temporal Databases: Theory, Design, and Implementation A Tansel et al eds., Benjamin/Cummings(1993) PDF, .14MB
  12. Lomet, D. A Review of Recent Work on Multi-attribute Access Methods. SIGMOD Record 21,3 (Sept. 1992) 56-63. postscript, .16MB
  13. Lomet, D. and Salzberg, B. Access Method Concurrency with Recovery. SIGMOD Conference, San Diego, CA (May 1992) 351-360. postscript .15MB
  14. Lomet, D. and Salzberg, B. Versioned Backups and Index Concurrency: Results of Work-in Progress. Int'l Workshop on High Performance Transaction Systems (Sept 1991) Asilomar, CA postscript, .11MB
  15. Lomet, D. and Salzberg, B. Spatial Database Access Methods. SIGMOD Record 20,3 (Sept. 1991) 5-15. PDF, .16MB
  16. Lomet, D. Grow and Post Trees: Role, Techniques, and Future Potential. (invited paper) 2nd Symposium on Spatial Databases (Aug. 1991) Zurich, Switzerland in Advances in Spatial Databases, Lecture Notes in Computer Science 525 Springer-Verlag (1991) 183-206. postscript, .29MB
  17. Lomet, D. and Salzberg, B. The hB-tree: a multiattribute indexing method with good guaranteed performance. ACM Trans. on Database Systems 15,4 (Dec 1990) 625-658. PDF, 1.95MB also in Readings in Database Systems, 2nd edition, M. Stonebraker ed., Morgan Kaufmann (1993)
  18. Lomet, D.B. and Salzberg, B. The Performance of a Multiversion Access Method. SIGMOD Conference, Atlantic City, NJ (May 1990) 353-363. PDF, .87MB
  19. Lomet, D. and Salzberg, B. Access methods for multiversion data. SIGMOD Conference, Portland, OR (May 1989) 315-324. PDF, .85MB
  20. Lomet, D., and Salzberg, B.J. A Robust Multi-attribute Search Structure. IEEE ICDE Conference, Los Angeles, CA (February 1989), 296-304. PDF, .64MB
  21. Lomet, D.B. A Simple Bounded Disorder File Organization with Good Performance. ACM Trans. on Database Systems 13,4, (December 1988), 525-551. PDF, 1.52MB
  22. Litwin, W. and Lomet, D.B. A New Method for Fast Data Search with Keys. IEEE Software 4,2, (March 1987), 16-24. PDF, .59MB
  23. Lomet, D.B. Partial Expansions for File Organizations with an Index. ACM Trans. on Database Systems 12,1 (March 1987), 65-84. PDF, 1.08MB
  24. Litwin, W. and Lomet, D. The Bounded Disorder Access Method. IEEE ICDE Conference, Los Angeles, CA (February 1986), 38-48. PDF, .85MB
  25. Lomet, D. A High Performance, Universal, Key Associative Access Method. SIGMOD Confence, San Jose, CA (May 1983), 120-132. PDF, 1.43MB
  26. Lomet, D.B. Bounded Index Exponential Hashing. ACM Trans. on Database Systems 8,1 (March 1983), 136-165. PDF, 1.56MB
  27. Lomet, D. Digital B-trees. VLDB Conference, Cannes, France (September 1981), 333-344. PDF, 1.2MB
  28. Lomet, D. Multi-table Search for B-tree Files. SIGMOD Confence, Boston, MA (May 1979), 35-42. PDF, .85MB

back to the top

Sorting and Ordering

Order preserving string compression is very important in many searching and sorting techniques. My work (with Gennady Antoshenkov and Jim Murray) on compressing variable length strings in a dictionary approach is the most general such method yet discovered.  AlphaSort was the world record holder on the Datamation sort benchmark when it was published. AlphaSort was an effort of a group led by Jim Gray. 

  1. Antoshenkov, G., Lomet, D., and Murray, J. Order Preserving Compression. IEEE ICDE Conference, New Orleans, LA (Feb. 1996) 655-663. postscript, .16MB
  2. Nyberg, C., Barclay, T., Cvetanovic, Z., Gray, J., and Lomet, D. AlphaSort: A Cache-Sensitive Parallel External Sort. VLDB Journal (Oct. 1995) 603-627 MS Word, .10MB
  3. Nyberg, C., Barclay, T., Cvetanovic, Z., Gray, J., and Lomet, D. AlphaSort: a RISC Machine Sort. (Best Paper Award) SIGMOD Conference, Minneapolis, MN (May 1994) 233-242 PDF, 1.0MB

back to the top

Languages and Compilers

My dissertation from the University of Pennsylvania showed how to generate parsers from deterministic context free grammars. The parsers so generated were transition diagram systems, with multiple exits to permit bottom-up as well as top-down parsing. A paper based on my dissertation was published in the Journal of the ACM. It showed how all deterministic grammars could be put in Greibach normal form, i.e. that left recursion could be eliminated from deterministic grammars. In a series of papers written in the 1970's and early 1980's, I explored both compiler implementation and programming language topics. My paper on cross procedural data flow was the first to exploit the notion of ``maybe aliases" information. My particular language focus was on using typed pointers and control over when storage can be freed to provide highly efficient pointer safety. This involved an early notion of what is now called ``pointer swizzling".

  1. Lomet, D.: Symbol Binding and Resolution in Programming Languages (unpublished) (1985, revised 1994, reformatted 2002) PDF, .08MB
  2. Lomet, D.B. Making Pointers Safe in System Programming Languages. IEEE Trans. on Software Engineering SE-11, 1 (January 1985), 87-96. PDF, .79MB
  3. Lomet, D.B. A Data Definition Facility Based on a Value-Oriented Storage Model. IBM Journal of R&D 24,6 (November 1980), 764-782. PDF, 1.21MB
  4. Lomet, D.B. Data Flow Analysis in the Presence of Procedure Calls. IBM Journal of R&D 21,1 (November 1977), 559-571. PDF, .77MB
  5. Lomet, D.B. Objects and Values: the Basis of a Storage Model for Procedural Languages. IBM Journal of R&D 20,2 (March 1976), 157-167. PDF, .86MB
  6. Lomet, D. Control Structures and the RETURN Statement. IFIP Congress, Stockholm, Sweden (August 1974), 403-407. PDF, .34MB
  7. Lomet, D. Automatic Generation of Multiple Exit Parsing Subroutines. Colloquium on Automata, Languages, and Programming Saarbrucken, Germany, in Lecture Notes in Computer Science 14 Springer-Verlag (July 1974), 214-231
  8. Lomet, D.B. A Formalization of Transition Diagram Systems. Journal of ACM 20,2 (April 1973), 235-257. PDF, 1.5MB
  9. Lomet, D.B. The Construction of Efficient Deterministic Language Processors. IBM Tech. Report RC2768 (Jan 1970) TJ Watson Research Lab, Yorktown, NY also Ph.d Dissertation, University of Pennsylvania.

back to the top

Architecture

In the 1970's, I worked in the area of high level language machines. My focus was mostly about how to efficiently support a high level language via hardware mechanisms to which a language could be compiled. The central issue addressed was to guarantee pointer safety with efficiency. This resulted in a number of IBM Technical Disclosures, and one paper. One outgrowth of this work was a disclosure describing a form of register windows. This was well before this notion was exploited in the RISC machines developed at UC Berkeley or Stanford. These are listed in an order consistent with the evolution of the ideas.
  1. Lomet, D.B. Scheme for Invalidating References to Freed Storage. IBM Journal of R&D 19,1 (January 1975), 26-35. PDF, .53MB
  2. Lomet, D.B. Stack Discipline for Procedure Activation Tombstones. IBM Techical Disclosure Bulletin 16,9 (February 1974), PO872-0453, 2980-2984 PDF, .71MB
  3. Lomet, D.B. Combining General Purpose Registers with a Display. IBM Techical Disclosure Bulletin 17,12 (May 1975), YO874-0279, 3780-3783 PDF, .46MB
  4. Lomet, D.B. Hardware Assistance for Type Checking. IBM Techical Disclosure Bulletin 18,12 (May 1976), YO875-0031, 4194-4197. PDF, .48MB
  5. Lomet, D.B. Separating the Area Management System from the Storage (Paging) System. IBM Techical Disclosure Bulletin 21,5 (Oct. 1978), YO877-0492, 2164-2166. PDF, .26MB
  6. Lomet, D.B. Areas and Spaces. IBM Techical Disclosure Bulletin 21,5 (Oct. 1978), YO877-0494, 2167-2168. PDF, .19MB
  7. Lomet, D.B. Long Areas and Long Spaces. IBM Techical Disclosure Bulletin 21,5 (Oct. 1978), YO877-0493, 2169-2170. PDF, .18MB
  8. Lomet, D.B. Regions for Controlling the Propagation of Addressability in Capability Systems. IBM Techical Disclosure Bulletin 22,3 (Aug. 1979), YO878-0594, 1286-1289. PDF, .45MB

back to the top

General Topics

From time to time I am invited to part in workshops or public discussions or to provide an article on general topics relevant to the research community.  The topics usually are about research directions or how to make research more effective.

  1. Jones, C., Lomet, D., Romanovsky, A., Weikum, G., Fekete, A., Gaudel, M-C., Korth, H., de Lemos, R., Moss, J. E. B., Rajwar, R., Ramamritham, K., Randell, B., Rodrigues, L. The atomic manifesto: a story in four quarks. in J. UCS 11(5) (2005) 636-651; Operating Systems Review 39(2) (2005) 41-46; and SIGMOD Record 34(1) (2005) 63-69. PDF, .09MB
  2. Lomet, D., Barga, R., Chaudhuri, S., Larson, P., and Narasayya, V. The Microsoft Database Research Group. SIGMOD Record 27,3 (Sept. 1998) 81-85. PDF, .33MB
  3. CRA/NSF Workshop on Strategic Directions in Computing (1996): I participated in the Database Management Systems Working Group. Leading database researchers met in a workshop organized by the ACM. The workshop met at MIT and discussed trends in both database research and in how databases have, do, and might fit into the wider trends in computing, e.g., the world wide web, the commoditization of hardware, object component technology, etc. My submission subsequently appeared in Computing Surveys.
  4. Lagunita Workshops, San Jose, CA. (May 1995) and Palo Alto CA (Feb. 1990). Two workshops on the future of database research were held four years apart with the same intent. I also participated in both. The attendees were drawn from the leading database researchers. The intent of the workshops was to make the case for increased funding of database research by pointing to past accomplishments and emphasizing the important problems remaining. I wrote summaries of both of the reports, which appeared in Computing Research News.
  5. Blakeley, J., Fishman, D., Lomet, D., Stonebraker, M., and Barbara, D. Panel: The Impact of Database Research on Industrial Products (Summary) SIGMOD Record 23,3 (Sept. 1994) 35-40. postscript, .13MB

Miscellaneous Papers (Some Unpublished)

I have worked in a number of areas that do not fit into the tidy scheme that is presented above. Some of this work is on programming languages, some on file systems, some on databases in general. Not all of this work has been published. Where it has been published, I give a citation. Where it has not, I mark it as unpublished. Some of the unpublished work is also "unfinished". I include it here simply because I have no current plans to complete it, but think that it might be of interest to someone in technical community. As Julia Child would say, "Bon Appetit!"

This section is itself a work in progress. Stay tuned for more exciting and little known papers.

  1. Bernstein,P., Lomet, D.: CASE Requirements for Extensible Database Systems Data Engineering Bulletin10,2 (June 1987) 2-9. PDF, .48MB
  2. Lomet, D.: How the Rdb/VMS data sharing system became fast. Rdb Expo (1992) postscript, .25MB
  3. Lomet, D.: The Case for Log Structuring in Database Systems HPTS (October, 1995) PDF, .02MB

back to the top