Peter Shipley, 2003. personal communication.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....Veritas reported a 50 improvement in LADDIS peak throughput with the new virtual memory allocator [Taylor99] # Web Service. On a large Starfire system running 2700 Netscape servers under Softway s Share II scheduler, vmem reduced system time from 60 to 10 , roughly doubling system capacity [Swain98]. # I O Bandwidth. An internal I O benchmark on a 64 CPU Starfire generated such heavy contention on the old rmalloc( lock that the system was essentially useless. Contention was exacerbated by very long hold times due to rmalloc( s linear search of the increasingly fragmented kernel heap. ....

Peter Swain, Softway. Personal communication.

....determine a predication involving A, x and t. Therefore, no observation of system states and environment states is going to help you with analysing the chance of this accident. Ergo, hybrid predications are essential information. 58 An Example: Playing Golf It has been suggested by colleagues [Mel00, Lev00] that safety concepts can be applied to a simple system such as a golf game. Recall that all that is needed to have a system is objects and behavior. Intuitive Interpretation The idea is that an accident is some event such as playing too many strokes (and thereby losing the game) Landing ....

Peter Mellor. Personal communication. February 2000.

....only insertions and searches. While the term implicit originated in [20] it has also been the subject of papers taking a somewhat di erent point of view, including a long lists of results in perfect hashing [14, 8] boundeduniverse dictionaries [7, 21] and cache oblivious data structures [6, 24]. These results were obtained for less stringent models di erent from the model adopted by Munro and Suwanda and in following papers. Yao [27] examined the special case of keys belonging to a bounded universe U , so that each key can be interpreted as an integer ranging in 0 : jU j 1 and ....

Peter Sanders. Personal communication (ALGO'02), 2002.

....of them unit (see [Bau96] for more details) The major limitation of linearizing completion is that it often diverges on the input set of clauses. In some cases, however, it is possible to partition into two sets 2 such that linearizing completion on each of them separately succeeds [Bau01] Until now this fact was not extremely useful because, clearly, neither of the two reasoners produced this way would be a background reasoner for when taken separately. The two reasoners, however, can now be combined with our approach to reason over in cooperation. Although this idea looks ....

....useful because, clearly, neither of the two reasoners produced this way would be a background reasoner for when taken separately. The two reasoners, however, can now be combined with our approach to reason over in cooperation. Although this idea looks promising, the cases mentioned in [Bau01] involve very simple theories. More work needs to be done to find more interesting cases of theories that can be partitioned into subtheories on which linearizing completion converges. Acknowledgments. I would like to thank Melvin Fitting for pointing out Malitz s interpolation theorem to me. ....

Peter Baumgartner. Personal communication, 2001.

....as so far described: First, the diagram implies that Bob is obtaining access to Carol, but computation as so far described gives Bob no means for holding on to this access. Second, we understand the diagram to say that Alice is giving Bob access to Carol herself, not a copy of Carol [8]. However, in computation as has been described so far, Carol is indistinguishable from a copy of Carol. We cannot distinguish between pass by reference sharing and pass by copy, but the Granovetter Diagram clearly intends to show specifically pass by referencesharing. Were computation adequately ....

Peter Deutsch, personal communication.

.... over all possible values of v, and using r n 1 #, we get that the probability of situation 1 is at most: l # v=3 v 2 r v 1 n v 1 c 2 r 2v # c 2 rn # # v=3 v 2 (n r) v 13 c 2 # n 2 = O(1 n 2 ) The above derivation follows a suggestion of Sanders and Vocking [29], and improves the O(1 n) bound in the conference version of this paper [25] In the second situation there is a sequence of distinct nestless keys b 1 , b v , v # (2t 1) 3, such that b 1 is the key to be inserted, and CUCKOO HASHING 11 x j 2 x j 1 x i 2 x i 1 x 1 = x i j 1 x ....

Peter Sanders and Berthold Vocking, 2001. Personal communication.

....differential equations which currently cannot be formulated in Modelica. Furthermore, these equations require very high efficiency in computation, as well as very detailed control of solution method choice, which is missing in current implementations. However there is ongoing research in this area[32, 58]. The MBS library is a collection of Modelica classes. In order to do modeling of a mechanical system certain classes need to be instantiated in the user s model. These should be properly connected, according to certain rules. Mechanical system models can be used for several purposes, since there ....

Peter Fritzson, personal communication, April 2000.

....them unit (see [Bau96] for more details) The major limitation of linearizing completion is that it often diverges on the input set T of clauses. In some cases, however, it is possible to partition T into two sets T 1 and T 2 such that linearizing completion on each of them separately succeeds [Bau01] Until now this fact was not extremely useful because, clearly, neither of the two reasoners produced this way would be a background reasoner for T when taken separately. The two reasoners, however, can now be combined with our approach to reason over T in cooperation. Although this idea looks ....

....useful because, clearly, neither of the two reasoners produced this way would be a background reasoner for T when taken separately. The two reasoners, however, can now be combined with our approach to reason over T in cooperation. Although this idea looks promising, the cases mentioned in [Bau01] involve very simple theories. More work needs to be done to nd more interesting cases of theories that can be partitioned into subtheories on which linearizing completion converges. Acknowledgments. I would like to thank Melving Fitting for pointing out Malitz s interpolation theorem to me, ....

Peter Baumgartner. Personal communication, 2001.

....of which paragraph was hit. Apart from this interface, control and presentation are entirely separate. In other words, Ensemble follows the model view controller (MVC) architecture espoused by the architects of Smalltalk [17] In practice, most Smalltalk programs merge the view and the controller [9]. In Ensemble, though, there is a special design constraint which motivates the separation: the need for multiple presentations of each document. When these presentations are closely related for instance, one and two column versions of the same document style much of the control code will be ....

L. Peter Deutsch, April 1994. Personal communication.

....query sequence. The small probability of making too many mistakes is determined by the coin flips of the learning algorithm and not by the query sequence selected by the adversary. We contrast this result with an n Gamma 1 mistake bound when the learner selects the query sequence [28], and an n Gamma 1 mistake bound when a teacher selects the query sequence. The remainder of this paper is organized as follows. In the next section we formally introduce the basic problem, the learning scenario and the extended mistake bound model. In Section 3 we present our results for ....

....an n Gamma 1 mistake bound. The teacher can uniquely specify the target total order by giving the n Gamma 1 instances that correspond to consecutive elements in the target total order. Since n Gamma 1 instances are needed to uniquely specify a total order, we get a matching lower bound. Winkler [28] has shown that under the learner selected query sequence, one can also obtain an n Gamma 1 mistake bound. To achieve this bound the learner uses an insertion sort, as described for instance by Cormen, Leiserson, and Rivest [8] where for each new element the learner guesses it is smaller than ....

Peter Winkler. Personal communication. 42

....energy E generalizes the energy E for knots K ae IR 3 studied in [Oha] and [FHW] E(K) ZZ K ThetaK 1 jx Gamma yj 2 Gamma 1 dK (x; y) 2 ds K (x) ds K (y) where dK (x; y) is the shorter arclength distance within K from x to y. In fact, Peter Doyle and Oded Schramm [DS] introduced, in the one dimensional case, the idea of a regularization by a multiplicative factor depending on angle, and observed that E(K) E(K) Gamma 4 when the factor used is (1 Gamma cos ff) One way they explained this equivalence of E and E is through a picture which interprets the ....

Peter Doyle and Oded Schramm. Personal communications.

....circuit breakers which contain electronics to detect arcing (so called arc fault detection technology) and which trip on the occurrence of arcing. One such has been developed by ETA Technologies in Germany (which amongst other things provides circuit breakers for European military aircraft) [14, 11]. Other detectors have been developed by the company Square D for home use [15] and Eaton Corp has patented arc fault detection technology suitable for aviation [10] The issues of faulty and deteriorated wiring have attracted the attention of US Congressional committees. 3.2.2 Summary of the ....

....or insulation damage, although it has not been determined which damage, if any, was causative and which was a consequence of the re. In this area there runs avionics wiring, as well as electrical system wiring, and wiring for the in ight entertainment system, which was installed post manufacture [11]. Damage to wiring may occur through faulty wiring system design, faulty installation, mechanical deterioration, or chemical deterioration of the insulation, through use or age or both. If wiring is routed through a bend, then a factor of two in the radius of the bend can increase the ....

Peter Meckler. Personal communication, September 29, 2000.

....imposter which includes the sun and the knowledge of the window frame, we can screen an object to avoid sending shadow rays if the object cannot see the sun through the window. A similar example is a luminaire imposter for a lamp with built in knowledge of its surrounding lamp shade as in [59]. This extra information in an environment with multiple luminaires can help to derive a p.d.f. whose samples converge faster since a luminaire, which is invisible to the object of interest, will not be selected. 3.3.3 Imposters on a virtual domain In lighting computation, a new geometric shape in ....

Peter Shirley. Personal Communication.

....subtyping should be based on the behavior of objects. If instances of class x meet the external interface of class y, then x should be a subtype of y. The example above demonstrates that the implementation hierarchy need not be the same as the type hierarchy (as defined by object behavior) [Canning85]. In that example, stack inherits from deque but is not a subtype of deque, and deque is a subtype of stack but does not inherit from stack. Behavioral subtyping cannot be deduced without formal semantic specifications of behavior. Lacking such specifications, one can deduce subtyping based ....

Peter Canning. Personal communication.

....keys. Key stretching only reduces a short key s resistance to certain attacks. If at all possible, system designers should use long keys. Several open questions remain. It would be nice to see an analysis of various key schedules economies of scale for keysearch attacks. For example, Peter Trei [Tre97] has recently demonstrated that a software DES keysearch can step through the DES key schedule in Gray code order, making each new key much cheaper to schedule for the keysearcher than a key is to schedule for an ordinary user. What other ciphers have this or worse properties, which make keysearch ....

Peter Trei, personal communication, 1997.

....1) and let f(i) x i1 with probability 1 2; x i2 with probability 1 2. Then, IP[fg(x) x] 1 2, so IP[fg(x) x] #, for any epsilon less than 1 2. But notice that (1 2) X = S . Curiously, once # exceeds 1 2 we must have X # S , under the assumption of independence. Examples [52] show that the assumption of independence is necessary. Independence is a reasonable assumption if we 6 try to apply this result for evolutionary tree reconstruction and holds automatically if the tree reconstruction method is deterministic. This lower bound applies to an arbitrary algorithm, ....

Peter Winkler, personal communication.

....every bit must flip during a Gray code, any graph induced by a code must be connected. Thus trees, the edge minimal connected graphs, are of particular interest. Furthermore, reflection, the key step in the standard construction of Gray codes, adds a leaf to the graph of a code. Both Slater [12] and Bultena and Ruskey [1] note that the standard reflected n bit Gray code induces a simple tree the star K 1;n Gamma1 (one central vertex connected to n Gamma 1 leaves) Another family of simple trees, the paths Pn , behaves differently. Paths with two or three vertices are stars and thus ....

....can be constructed (see Figure 5) However, Slater [11] Bultena and Ruskey [1] and the current authors have all verified computationally that no P 7 code exists, and we all conjecture that no longer path has a Gray coding. Some results are known for graphs outside these simple families. Slater [12] gives two families of diameter 4 trees with Gray codings. Bultena and Ruskey [1] construct ingenious cyclic Gray codings of all trees of diameter 4. They also catalogue the Gray codability of all trees on 6 or fewer vertices and show that no tree of diameter 3 has a cyclic Gray code. The current ....

[Article contains additional citation context not shown here]

Slater, Peter J. Personal communication, 1989.

....pf (controls A F) pf (A F) pf F = P1] P2] imp e (def2 e P1) P2. Figure 5: Representation of controls and controls e. cause an invalid proof to be accepted by the checker. Necula [11] has also implemented an LF proof checker; its current size is about 2000 lines of commented C code [7]. Clients who want to generate proofs will need application specific definitions and lemmas, which might amount to several hundred lines of Twelf, but this will be outside the trusted computing base the lemmas can be independently checked by the recipients of proofs. Proofs are really just ....

Peter Lee. personal communication, 1999.

....verification time proportional to the number of loop free paths as the second approach does. The latter conclusion is important for certified code frameworks (such as proof carrying code) that construct verification conditions at verification time via a form of weakest pre condition computation [11]. If done naively, the constructed condition can have exponential size by creating a different logical clause for every loop free path. For a compiler to exploit the weakness of such a pre condition, it would need to have optimized based on an exponential amount of path sensitive information. We ....

Peter Lee. Personal communication.

....much explicit argument in scientific papers. Although Landauer et al. claim that scientific reports go from problem to observation to inference [LER 93, p. 74] that does not seem true of computer science and information science. Foltz found that psychology articles have a very rigid format [Fol96] however this observation does not seem to apply to articles in the fields of computer science and information studies. 19 3.1.3.2 Structural Links that make explicit connections that are often presented in ordinary documents by headings, cross references and citations are termed structural ....

Peter Foltz. Personal communication. telephone conversation, 17 January 1996.

....its inherently high selection pressure. Possible jobs are picked at random until either one is found that is more attractive than the first, or a fixed number (the Monte Carlo size) of possible moves have been selected, in which case the job with the highest idle time associated with it is chosen [14]. 5 Evaluating The Heuristic The measure of performance used was the quality of solution obtained after a set number, N , of evaluations. Fifty runs were taken in each case, therefore summaries are given below and detailed results are available from the authors. Experiments were performed for ....

Peter Ross. Personal communication, 1996.

No context found.

Peter Shipley, 2003. personal communication.

No context found.

Peter Lee. Personal communication.

No context found.

Peter Trei, personal communication, 1997.

No context found.

Peter Deutsch, personal communication.

First 50 documents

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC