Effective Software Defect Tracking

From: "Saved by Windows Internet Explorer 8" Subject: Effective Software Defect Tracking - Apr 99 Date: Thu, 4 Feb 2010 16:19:40 +0530 MIME-Version: 1.0 Content-Type: multipart/related; type="text/html"; boundary="----=_NextPart_000_0000_01CAA5B5.DA51AD90" X-MimeOLE: Produced By Microsoft MimeOLE V6.0.6002.18005 This is a multi-part message in MIME format. ------=_NextPart_000_0000_01CAA5B5.DA51AD90 Content-Type: text/html; charset="Windows-1252" Content-Transfer-Encoding: quoted-printable Content-Location: http://www.stsc.hill.af.mil/crosstalk/1999/04/subramaniam.asp Effective Software Defect Tracking - Apr 99

=20

Effective=20 Software Defect Tracking
Reducing Project Costs and Enhancing=20 Quality=20

Bala Subramaniam
ISSRe Systems, = Inc.=20
=20
The costs of defective software can be as high as = 50=20 percent of the investment in software development. Yet, the = potential=20 to improve software quality and reduce project cost is = enormous.=20 Software defect tracking can be an effective means to achieve = quality=20 at less cost. However, defect tracking is commonly = misunderstood,=20 incorrectly implemented, and often seen as an impediment and = cost to=20 the organization. This article discusses the quality costs of=20 defective software and provides a working model to implement = an=20 effective software defect tracking system within an = organization.=20

EFECT TRACKING IS SOMETIMES = written=20 off as boring, repetitive, and unglamorous. Even effective = defect=20 tracking is often viewed as an unnecessary cost that impedes = schedules.=20 Yet, defect tracking is one of the most critical components of = the=20 software development and the quality assurance efforts.=20
     When implemented well, defect = tracking=20 greatly reduces overall project costs and improves schedule = performance.=20 As a critical component in improving software quality, the = potential=20 paybacks for such processes are enormous. A Hewlett-Packard = quality=20 program reduced software errors by 75 percent and cut = development time=20 20 percent. An Air Force systems group reports that every dollar = invested to improve quality has a conservative return of $7.50.=20
     To effectively track and manage = software=20 defects also improves customer satisfaction, creates higher = productivity=20 and quicker delivery, and leads to better operational = reliability and=20 improved morale. On the other hand, a mismanaged software defect = tracking program may indeed be an unnecessary cost.=20
     Software defects take different = names in=20 different organizations, e.g., errors, issues, bugs, defects, or = incidents. Whatever they are called and whatever form they take, = defects=20 can have an astounding impact on the development phase and can = continue=20 to haunt the product through its maintenance phase.=20
     The costs to fix software defects = are high,=20 especially if fixing requires developers to re-familiarize = themselves=20 with months-old work or if someone other than the original = developer is=20 doing the fixing. Costs also increase exponentially while moving = further=20 along the software development lifecycle. Studies at IBM = demonstrate=20 that compared to catching defects before or during coding, it is = 10=20 times more costly to correct an error after coding and 100 times = more=20 costly to correct a production error.=20
Software Quality=20 Costs
A 1996 study by The Standish Group reported = that=20 U.S. businesses invest about $250 billion in software = development=20 annually, yet a great many of these projects fail because of = cost=20 overruns. One of the significant components of project costs is = software=20 quality cost. One estimate put the cost of a single post-release = defect=20 to a large organization as high as $20,000 to $40,000.=20
     Software quality costs are the = costs=20 associated with preventing, finding, and correcting defective = software.=20 Following are three useful definitions of quality costs [1].=20
     Prevention Costs. These are = costs of=20 activities specifically designed to prevent poor-quality = software, e.g.,=20 costs of efforts to prevent coding errors, design errors, = additional=20 document reviews to reduce mistakes in the user manuals, and = code=20 reviews to minimize badly documented or unmaintainably complex = code.=20 Most of these prevention costs do not fit within a typical = testing=20 group's budget. The programming, design, and marketing staffs = spend this=20 money.
     Appraisal Costs. = These are=20 costs of activities to find defects, such as code inspections = and=20 software quality testing. Design reviews are part prevention and = part=20 appraisal. Formulating ways to strengthen the design is a = prevention=20 cost, whereas to analyze proposed designs for potential errors = is an=20 appraisal cost.
     Failure = Costs. These=20 are costs that result directly from poor software quality, such = as the=20 cost to fix defects and the cost to deal with customer = complaints.=20 Failure costs can be divided into two main areas:=20

Internal failure costs: Costs that arise before the = product=20 is delivered to the customer. Along with costs to find and fix = bugs=20 are costs associated with wasted time, missed milestones, and = overtime=20 needed to get back on schedule.=20
External failure costs: Once the software is = delivered to=20 the customer, poor-quality software can incur customer service = costs=20 or the cost to distribute a patch for a released product. = External=20 failure costs are huge; it is much cheaper to fix defects = before=20 shipping the defective product to customers. If a product has = to be=20 shipped late because of bugs, the direct cost of late shipment = includes the lost sales, whereas the lost opportunity cost of = the late=20 shipment includes the costs of delaying other projects while = everyone=20 finishes the one that is error-ridden.=20
     User interface defects are = often=20 treated as low priority and are fixed last. This can be a = mistake.=20 Product screens may be required for effective marketing and=20 documentation. This can result in increased costs in = nondevelopment=20 areas, lost marketing opportunities, and contractual penalties.=20 Unfortunately, numerical estimates of lost opportunity costs and = delays=20 are difficult to make and can be controversial [2].=20
     All the above software quality = costs=20 contribute to the total cost of poor-quality software to the=20 organization. In aggregation, the total cost of software quality = may be=20 presented as follows:
     Total Cost of = Quality=20 =3D Prevention + Appraisal + Internal Failure + External = Failure.=20
What = Is a=20 Defect?
Defects are commonly defined as "failure = to=20 conform to specifications," e.g., incorrectly implemented = specifications=20 and specified requirement(s) missing from the software. However, = this=20 definition is too narrow. Discussions within the software = development=20 community consistently recognize that most failures in software = products=20 are due to errors in the specifications or requirements=97as = high as 80=20 percent of total defect costs [3]. Other studies have shown that = the=20 majority of system errors occur in the design phase [4]. Figure = 1=20 represents the results of numerous studies that show = approximately=20 two-thirds of all detected errors can be traced to the design = phase.=20

Figure 1. Origin of software = errors across=20 industry.=20
     I recommend a broader definition of = defect:=20 variance from a desired attribute. These attributes = include=20 complete and correct requirements and specifications, designs = that meet=20 requirements, and programs that observe requirements and = business rules.=20
Implementing an=20 Effective Defect Tracking Process
Software = quality=20 assurance departments can play a catalytic role in implementing = an=20 effective defect tracking process. A survey conducted in 1994 by = the=20 Quality Assurance Institute found that a mere 38 percent of the=20 organizations had formal software defect management processes, = whereas=20 25 percent of the survey participants said their organizations = lack=20 consistent testing standards and procedures [5]. The survey also = reported that although 60 percent of organizations had testing = standards=20 and procedures, some organizations admitted they were out of = date and=20 not followed. Recent surveys, nonetheless, suggest that more = companies=20 are now striving to improve their software development process = through=20 early defect identification, minimizing resolution time, hence = reducing=20 project costs.
     Effective defect = tracking=20 begins with a systematic process. A structured tracking process = begins=20 with initially logging the defects, investigating the defects, = then=20 providing the structure to resolve them. Defect analysis and = reporting=20 offer a powerful means to manage defects and defect depletion = trends,=20 hence quality costs.=20
Integrate Software=20 Development and Defect Tracking
Traditional = approaches=20 place testing immediately before implementation. Typically, = testers=20 receive a low-quality product at the tail end of development = when there=20 is tremendous pressure to deliver, even if the software is = plagued with=20 defects. For early defect detection and resolution to take = place, defect=20 tracking and software development efforts should begin = simultaneously.=20 It will solve a multitude of problems downstream.=20
     Defect tracking must be implemented = throughout the development lifecycle. On projects I have managed = or=20 worked on, this has always lead to fewer release defects; = however, such=20 organizational foresight is rare. The Sentry Group reported that = 62=20 percent of all U.S. organizations do not have a formal quality = assurance=20 or test group. The report also added that a large majority of = these=20 organizations place a much higher priority on meeting schedule = deadlines=20 than producing high-quality software [6].=20
The = Different=20 Phases of Defect Tracking
Successful verification = throughout the development cycle requires clearly defined system = specification and software application business rules.=20
     Requirements phase. Defect = tracking=20 focuses on validating that the defined requirements meet the = needs and=20 the users' expectation about functionality. Sometimes, = system-specific=20 constraints would cause the deletion of certain business = requirements.=20
     Design and analysis phase. = Efforts=20 should focus on identifying and documenting that the application = design=20 meets the business rules or field requirements as defined by the = business or user requirements. For example, does the design = correctly=20 represent the expected user interface? Will it enforce the = defined=20 business rules? Would a simpler design reduce coding time and=20 documentation of user manuals and training? Does the design have = other=20 effects on the reliability of the program? =
    =20 I experienced the downstream cost of a specification error when = working=20 on one of two groups of developers who were programming = complementary=20 parts of a data-bridging program. Coding was well under way when = incomplete system specifications caused transfer of data on the = bridge=20 to fail. The failure was not due to coding errors but to = specification=20 errors that were translated into program codes. Had the = deficiency been=20 discovered before coding began, we could have saved the = substantial time=20 and money required to repair the programs. =
    =20 Programming phases. Defect tracking must emphasize = ensuring that=20 the programs accomplish the defined application functionality = given by=20 the requirements and design. For example, has any particular = coding=20 caused defects in other parts of the application or in the = database? Is=20 a particular feature visibly wrong?
     = Maintenance and enhancement phases. During the = maintenance phase,=20 effort is spent tracking ongoing user issues with the software. = During=20 enhancement phases (there could be multiple releases), defect = tracking=20 is focused on establishing that the previous release is still = stable=20 when the enhancements have been added. Figure 2 represents the=20 philosophy of defect tracking throughout the software = development=20 process.=20
Introduce = Defect Tracking=20 Early
It is not difficult to introduce tracking = early=20 into the development process=97it fits well into current = software=20 development processes. Today's rapid application development = (RAD), the=20 dominant approach in client-server software projects, focuses on = shortened development schedules. This software development = method=20 provides early review points, delivered as "builds" or = iterations of=20 development, to ensure that requirements are met. Such a process = clearly=20 lends itself to early defect tracking, which can shadow = development (see=20 Figure 2). Each build can receive verification that it meets the = defined=20 requirements; if not, defects can be reported and resolved = quickly and=20 relatively inexpensively while the software is still "pliable." = The same=20 defect reported later in the development process may require a = major=20 "surgery" to the software product; hence, it will be more = costly.=20 Front-end defect tracking costs much less than waiting until the = end.=20
"Quality comes not from inspection, but improvement = of the=20 development process."=20
=97 W. Edward Deming. [7]

Figure 2. Defect tracking = running parallel=20 to development lifecycle.=20
An = Effective Defect=20 Tracking Process
To merely integrate defect = tracking into=20 the development process is not enough. A clearly defined defect = tracking=20 process is needed to ensure defects are handled in an organized = manner=20 from discovery through resolution. Components of this process = are=20 described in the sections that follow. This process is=20 progressive=97defect evaluation cannot be successfully performed = if the=20 earlier components (such as describing defects and prioritizing = defects)=20 were not implemented.=20
Defect = Repository=20
Once a defect has been discovered, the important = first step=20 is to log the defect into a defect-tracking database or = repository. When=20 a defect is logged, it must be fully described so that it can be = reproduced during debugging, prioritized based on its severity, = and have=20 resources assigned for its resolution. Defects have a number of = other=20 attributes that should be recorded, such as=20

Defect number.=20
Date.=20
The build and test platform in which it was discovered.=20
The application requirement or business rule to which it = relates.=20
Any supplementary notes. =
     It also=20 is important that the repository offer a means to track the = "life" of=20 the defect (the resolution status) and historically report on = all=20 defects discovered and logged for the project. It pays to have = this=20 system online and available to all development staff so that the = assigned parties can update the resolution progress for the = defect=20 status.=20
Defects = Described=20
Your organization's defect reporting procedures = should=20 require that details about each software defect be recorded when = the=20 defect is discovered, including a description, symptoms, = sequence of=20 steps to re-create it, and severity. Defects are of various = types:=20

Interface defects include incorrectly working menu = items,=20 push buttons, and list boxes.=20
Navigational defects could be described as a window = not=20 opening when moving from one interface screen to another.=20
Functionality defects could be incorrect = calculation of=20 salaries in a payroll system. =
     Do not=20 merely log, "Adding new customer window does not work." A = detailed=20 description, such as, "The `Save' button on `Add New Customer' = window=20 does not work," would give the developer adequate information to = go=20 straight to the specific problem and repair it. This saves time = and=20 unnecessary interruption for the developer to research the = defect thus=20 reducing the overall project cost.=20
Defects = Prioritized=20
Once a defect is logged and described, appropriate = resources=20 must be allocated for its resolution. To do this, the defect = must be=20 analyzed and prioritized according to its severity. Each defect = is given=20 a priority based on its criticality. Usually, it is practical to = have=20 four priority levels:=20

Resolve Immediately.=20
High Priority.=20
Normal Queue.=20
Low Priority.
     A = misstatement of a=20 requirement or a serious design flaw must be resolved = immediately,=20 before the developer translates it into codes that are = implemented in=20 the software=97it is much cheaper to amend a requirement = document than to=20 make program code changes. The wrong font size for a label may = be=20 classified as "Low Priority."
     The = critical=20 path for development is another determinant of defect priority. = For=20 example, if one piece of the functionality must work before the = next=20 piece is added, any functional defects of the first piece = will be=20 given the "Resolve Immediately" priority level. On one project I = worked=20 on, a query engine retrieved transactions matching = user-specified=20 criteria upon which further processing was performed. If the = query=20 engine had been defective, no further development (or testing) = would=20 have been practical. Therefore, all functional defects of the = query=20 engine were prioritized as "Resolved Immediately."=20
     The urgency with which a defect has = to be=20 repaired is derived from the severity of the defect, which could = be=20 defined as follows:=20

Critical.=20
Important.=20
Average.=20
Low.
     A defect that = prevents the=20 user from moving ahead in the application=97a "show = stopper"=97is classified=20 as "Critical," e.g., performing an event causes a general = protection=20 fault in the application. Performance defects may also be = classified as=20 "Critical" for certain software that must meet predetermined = performance=20 metrics. If the user is able to formulate work-arounds where = there are=20 defects, these defects may be classified as "Average." An overly = long=20 processing time may be classified as "Important" because = although it=20 does not prevent the user from proceeding, it is performance = deficiency.=20 Defects with severity "Average" will be repaired when the=20 higher-category defects have been repaired and if time permits. = Certain=20 graphical user interface defects, such as placement of push = buttons on=20 the window, may be classified as "Low," since this does not = impede the=20 application functionality. Although defect priority indicates = how=20 quickly the defect must be repaired, its severity is determined = by the=20 importance of that aspect of the application in relation to the = software=20 requirements.=20
Structured = Resolution=20
The defect tracking system also must ensure that the = defect=20 progresses in an appropriate sequence from discovery through = resolution.=20 Each defect is also given appropriate status; for example, a new = defect=20 is given the status of "Open," and a defect under repair would = have the=20 status of "Assigned."
     As repair = work=20 progresses, the status of defects is updated to reflect its = state in the=20 resolution process. A defect that has been repaired will be = submitted to=20 the testing team through formal change control to be verified = again.=20 Only if the fix passes the regression test will it be accepted = and the=20 defect assigned a status of "Closed." Other defect statuses = could=20 include "Deferred," if the defect is not to be fixed for the = current=20 release but may be resolved in a subsequent release or = "Enhancement," if=20 a feature that is not part of the requirements has been = suggested, and=20 may be reviewed as an enhancement for later releases.=20
Communication=20
An effective defect tracking system must allow = communication=20 of the software's defects, status, or changes to members of the=20 development team and all others concerned. This has become an=20 increasingly crucial element because people working on the same = project=20 may not only work in different parts of a building but also may = even=20 work in a different state for a different organization. Without = an=20 effective means to communicate defects, defect tracking=97and = consequently=20 achieving software quality=97would be a nightmare.=20
     E-mail is an efficient vehicle to = expedite=20 informing software engineers and all concerned of defects as = they are=20 discovered. Software engineers could then perhaps access an = online=20 defect repository as they receive the E-mail on new and existing = defects. Similarly, E-mail also serves as a reply medium to = inform=20 testers that a defect has been repaired. Some defect tracking=20 repositories, e.g., one set up in Lotus Notes, facilitates = built-in=20 communication features that can be used by both software = engineers and=20 testers.
     Commercially available = defect=20 tracking software, e.g., AutoTester and SQA Team Test Software, = are more=20 sophisticated in communicating defects and their status to = individuals=20 or as a batch. They also automatically inform respective = development=20 staff and management of defects as they are discovered.=20
     Although E-mail provides a means to = convey=20 information about defects between the development and testing = team,=20 regular formal defect tracking meetings also help keep a close = eye on=20 the number, types, and nature of defects found, which may = indicate how=20 software quality is progressing through the resolution stage.=20
     Defect analysis is discussed in = more detail=20 in the "Reporting" component of this defect tracking process. If = the=20 testing and development teams must work hand-in-hand toward = achieving=20 software quality, there must be continuous communication between = them.=20 Informal or verbal communication between these teams is = inadequate.=20
Continuous = Defect=20 Resolution
It costs much less to resolve defects = as soon=20 as they are discovered=97do not merely accumulate a list of = defects to fix=20 later. For example, in my current project, the software product = is=20 undergoing two transformations: The entire application = architecture is=20 being revamped and enhancements are being implemented for the = next=20 release. Revamping the architecture changes the fundamental = "backbone"=20 of the application in question, which is in itself a complex = task. We=20 have three categories of defects:=20

The existing list of yet-to-be-resolved defects from the = original=20 application.=20
Defects that would come about as a result of revamping the = architecture.=20
New defects contributed by the new enhancements.=20
     We have divided the project = into=20 smaller deliverables and implemented defect tracking for each=20 deliverable. If resolution were to be delayed until later, the = mere=20 complexity of the various deliverables would present an = inordinate=20 amount of challenge to resolve defects. Moreover, different = software=20 engineers are working on different deliverables and different = tasks=20 within each deliverable. At later stages, it would become a = mammoth=20 effort to merely identify and assign defects to the respective = software=20 engineers. Additionally, when they have been assigned defects to = repair,=20 the engineers have to remember what they implemented in the = codes=20 perhaps months earlier. This will incur expensive investigation = time.=20
     The best time to resolve defects is = when=20 they are discovered. This is especially true in a RAD = environment, where=20 the application is developed through several iterations or = builds. Each=20 build has an incremental amount of application functionality and = related=20 coding. Any defects discovered in a particular build should be = referred=20 to developers immediately for resolution. The functionality = added in the=20 most recent build and related program codes are still fresh in = the=20 developers' minds, which leads to faster investigation of the = root cause=20 of the defects, and therefore more efficient resolution efforts. = To=20 defer defect resolution until later in the development cycle = wastes time=20 and resources.=20
Defect = Evaluation and=20 Analyses
Most organizations consider it essential = to=20 constantly monitor and evaluate their performance, and this key = practice=20 is especially critical in defect removal. The overall success of = your=20 project largely hinges on effective defect resolution, so you = need to=20 know your defect removal status and the cost of achieving = quality. For=20 example, a defect trend analysis will indicate the number of = defects=20 discovered over time. This analysis may even be further = subdivided for=20 defects by status, functionality, severity, etc. Defect age = analysis=20 suggests how quickly defects are resolved by category.=20
     The type and extent of the defect=20 evaluation and analyses may be determined by the organization's = cost=20 objectives and delivery schedules. Following are a few suggested = analyses that may be applicable to most software projects. The = following=20 measures need to be determined to analyze defects (or those = chosen as=20 part of an organization's defect analysis strategy).=20

Defect status vs. priority.=20
Defect status vs. severity.=20
Defect status vs. application module.=20
Defect age.
     The above = information=20 will not be available if the earlier steps of adequately logging = defects=20 were not implemented as part of the defect tracking process. By=20 comparing these measures from the current iteration to the = results from=20 the analysis of previous iterations, one can get an indication = of=20 defects trends, which are discussed further in the following two = subsections.
     Defect = Evaluation.=20 Although the evaluation of test coverage provides the measure of = testing=20 completion, an evaluation of defects discovered during testing = provides=20 the best indication of software quality. By definition, quality = is the=20 indication of how well the software meets a desired attribute. = So, in=20 this context, defects are identified as "variance from a desired = attribute."
     Defect evaluation may = be based=20 on methods that range from simple defect counts to rigorous = statistical=20 modeling. Rigorous evaluation can include forming a model (or = setting=20 goals) about discovery rates of defects, then fitting the actual = defect=20 rates during the testing process to the model. The results can = be used=20 to estimate the current software reliability and predict how the = reliability will grow if testing and defect removal efforts = continue.=20 However, because the field's current lack of a scientific model = and=20 resources dedicated to perform such evaluations (or a tool to = support=20 them), an organization should carefully balance the cost of = rigorous=20 evaluation with the value it adds.
    =20 Defect Analysis. This means analyzing the distribution of = defects=20 over the values of one or more parameters associated with a = defect.=20 Defect analysis provides an indication of the reliability of the = software. Four main defect parameters are commonly used for = defect=20 analysis:=20

Status: the current state of the defect (open, = being=20 repaired, closed, etc.).=20
Priority: the relative importance of addressing and = resolving this defect.=20
Severity: the relative impact of this defect to the = end-user, an organization, third parties, etc.=20
Source: what part of the software (such as a = module) or=20 requirement this defect affects. =
    =20 Defect counts can be reported in two ways: (1) as a function of = time,=20 resulting in a defect trend diagram or report and (2) as a = function of=20 one or more defect parameters (like severity or status) in a = defect=20 density report. These types of analysis provide a perspective on = the=20 trends or distribution of defects that reveal the software's=20 reliability.
     Defect trends follow a = fairly=20 predictable pattern in a testing cycle. Early in the cycle, the = defect=20 rates rise quickly. Then, they reach a peak about midstream, in = an=20 adequately staffed test project, and fall at a slower rate over = time.=20 The project schedule can be reviewed in light of this trend. For = instance, if the defect rates are still rising in the third week = of a=20 four-week test cycle, the project is clearly not on schedule. = Other=20 instances where the rate of closing defects is too slow = (experience=20 rated) might indicate a problem with the defect resolution = process; for=20 example, resources to fix defects or to retest and validate = fixes might=20 be inadequate.
     This is an important = aspect=20 of software project management: to ensure that software quality = is=20 progressing within the planned delivery schedule. Figure 3 = displays=20 defect status by software module. In each software module, a = discovered=20 defect is given a status of Open and assigned resources for = fixing.=20

Figure 3. Resolution efforts = for an=20 accounting system.=20
     In Figure 3, resolution efforts for=20 accounting appear to be good, because there are more defects = being=20 repaired than left open. Retesting efforts seem to be adequate, = because=20 defects closed are not far behind defects repaired. However, = although=20 there is a similar retesting effort for the claims module, there = are far=20 too many defects open, indicating that additional resources may = be=20 required for this module. The trend for the diary module = suggests that=20 both defect repair and retesting efforts are progressing well. = The=20 defect trend for the contracts module also shows that defect = resolution=20 is progressing well.
     An = organization could=20 set quality criteria for how the distribution of defects over = priority=20 levels should look, e.g., "No critical defects should stay open = for more=20 than one week." It would be expected that defect discovery rates = would=20 eventually diminish as testing and fixing progresses. A = threshold can be=20 established below that in which the software can be deployed.=20
     Defect counts can also be reported = based on=20 the source, allowing detection of weak modules and "hot spots." = Parts of=20 the software that must be fixed repeatedly indicate a = fundamental design=20 flaw. In my current project, this type of analysis helped us = come to=20 conclude that the application architecture technology needed to = be=20 revamped. The originally chosen architecture, although = technically=20 superior, was more complicated and made the application = extremely=20 delicate to changes or defects "fixes." =
    =20 Defects included in an analysis of this kind have to be = confirmed=20 defects. Not all reported defects turn out to be flaws; some may = be=20 enhancement requests, out of the scope of the project, or = describe a=20 previously reported defect. These defects must be reclassified = as such.=20 However, there is value in analyzing why many duplicates or = unconfirmed=20 defects are being reported.=20
Defect = Reporting=20
Defect evaluation and analysis have to be reported in = a=20 useful form to those who make decisions about resources, costs, = and=20 delivery schedules. Although each organization may want to = produce=20 different reports and different forms, there are three classes = of=20 reports:=20

Defect density or distribution reports allow defect = counts=20 to be shown as a function of one or two defect parameters. = Using the=20 priority parameter, defect distribution may be represented as = shown in=20 Figure 4.=20
Defect age reports are a type of defect = distribution report=20 that shows how long a defect has been in a particular state, = such as=20 Open. In any age category, defects can also be sorted by any = other=20 attribute, such as Owner (developer assigned to repair = defect).=20
Defect trend reports show defect counts by status = (New,=20 Open, or Closed) as a function of time. The trend reports can = be=20 cumulative or noncumulative and help management identify = defect rates=20 by status thus providing an indication of how well the = software=20 quality is progressing through the project cycle. Figure 5 = represents=20 a typical defect trend report.

Figure 4. Defect = distribution.=20

Figure 5. Defect trend = report.=20
     In Figure 5, the number of new = defects=20 peaked in February. Lagging behind new defects by about a month, = the=20 number of open defects was the highest in March. The = defect-fixing=20 efforts appear to be consistent throughout the project, closing = all=20 defects by June. Before an organization can produce the type of = reports=20 discussed here, a defect repository must have been established = to=20 support such analysis, i.e., provide for logging of defect = description,=20 status, priorities, etc., outlined in the "Defects Described" = section of=20 this article. To be successful, all pieces of the defect = tracking=20 process must be embraced.=20
Making This=20 Approach Work
It is unwise to try to achieve too = much too=20 fast. Change is the most difficult concept to grasp or = implement. The=20 effective defect tracking model discussed here not only may call = for a=20 fundamental change in your software development process but also = may=20 require a broader concept and definition of defects and the = tools to=20 manage them. Depending on the capability of the development and = quality=20 assurance process, an organization may not want to attempt the = total=20 defect-tracking model all at once. For lower maturity = organizations, an=20 incremental adoption of the recommendations would be more = successful.=20
     A good start is to set up a simple=20 defect-tracking repository that implements defect description, = status,=20 priority and severity, and communication. Expand that list later = to=20 include defect evaluation and analysis and reporting. This = ensures that=20 the required defect data is captured as a minimum so the = organization=20 can build on this model.
     Next, the = process=20 should be widened to embrace the broader definition of a defect, = and the=20 concept of implementing defect tracking across the development = process.=20 The implementation of an effective defect tracking process = should be=20 taken through levels of maturity, which is a topic for an entire = article. Once the model and process is applied to one project=20 successfully, it can be implemented across the organization.=20
Conclusion=20
Effective defect tracking strongly contributes to = enhancing=20 software quality and reducing development project costs. Using = the=20 broader definition of a defect ensures that not only are = resultant=20 errors
or nonconformance to requirements discovered but also = variance from a desired attribute, including incomplete = requirements,=20 takes place. Searches for such defects can then take place = across all=20 software development phases.
     By = "shadowing"=20 the software development process, defect tracking helps you = identify and=20 report potential software problems early and acts as a catalyst = for=20 problems to be addressed. By facilitating discovery of defects = earlier=20 in the development cycle, effective defect tracking is a = critical key to=20 lower costs, enhanced software quality, and reducing overall = project=20 cost. However, to achieve this requires a fundamental change in = the=20 ideology behind quality assurance and the software development = process=20 as well as the introduction of the necessary tools to track and = manage=20 defects. The defect-tracking model discussed in this article = will be=20 useful for organizations moving in this direction. Careful = planning and=20 phased adoption of this model can make this approach a powerful = software=20 quality strategy. =20
About the = Author=20
Bala Subramaniam is director of = quality=20 assurance at ISSRe Systems, Inc., in New York. He has 15 years=20 managerial and technical experience, during which he has worked = for=20 medium and large software companies including IBM. His special = interests=20 include the definition and implementation of quality assurance = methods=20 and software process improvement programs. He also is = experienced in=20 designing effective automated test methods to test complex=20 mission-critical software functionality and business cycles. He = has a=20 master's degree in business administration (finance) from = Birmingham=20 Business School in Great Britain and is a certified software = test=20 engineer (Quality Assurance Institute).

ISSRe Systems, Inc.
200 Business Park Drive=20
Armonk, NY 10504
Voice: 914-273-7777
Fax: = 914-273-7796=20
E-mail: balas@issre.com=20

References =

Campanella, J., ed., Principles of Quality Costs, = ASQC=20 Quality Press, 1990.=20
Juran, J.M. and Frank M. Gryna, Juran's Quality Control = Handbook, 4th ed., McGraw-Hill, New York, pp. 4.9-4.12.=20
ANSI/IEEE Standard 982.1-1988, IEEE Standard Dictionary = of=20 Measures to Produce Reliable Software, Institute of = Electrical and=20 Electronics Engineers, p. 13.=20
Perry, W., "Structured Approach to Testing," Effective = Methods=20 for Software Testing, John Wiley & Sons, New York, = 1995.=20
Perry, W., "1994 Survey Results on Software Testing," = Effective=20 Methods for Software Testing, John Wiley & Sons, New = York,=20 1995.=20
Automated Software Quality Directions, The Sentry=20 Group, February 1998.=20
Deming, W. Edward, Out of the Crisis, MIT Press, = Cambridge,=20 Mass., 1982.=20

------=_NextPart_000_0000_01CAA5B5.DA51AD90 Content-Type: image/gif Content-Transfer-Encoding: base64 Content-Location: http://www.stsc.hill.af.mil/crosstalk/1999/04/d.gif R0lGODlhPAA7APcAAAAAADMAAGYAAJkAAMwAAP8AAAAzADMzAGYzAJkzAMwzAP8zAABmADNmAGZm AJlmAMxmAP9mAACZADOZAGaZAJmZAMyZAP+ZAADMADPMAGbMAJnMAMzMAP/MAAD/ADP/AGb/AJn/ AMz/AP//AAAAMzMAM2YAM5kAM8wAM/8AMwAzMzMzM2YzM5kzM8wzM/8zMwBmMzNmM2ZmM5lmM8xm M/9mMwCZMzOZM2aZM5mZM8yZM/+ZMwDMMzPMM2bMM5nMM8zMM//MMwD/MzP/M2b/M5n/M8z/M/// MwAAZjMAZmYAZpkAZswAZv8AZgAzZjMzZmYzZpkzZswzZv8zZgBmZjNmZmZmZplmZsxmZv9mZgCZ ZjOZZmaZZpmZZsyZZv+ZZgDMZjPMZmbMZpnMZszMZv/MZgD/ZjP/Zmb/Zpn/Zsz/Zv//ZgAAmTMA mWYAmZkAmcwAmf8AmQAzmTMzmWYzmZkzmcwzmf8zmQBmmTNmmWZmmZlmmcxmmf9mmQCZmTOZmWaZ mZmZmcyZmf+ZmQDMmTPMmWbMmZnMmczMmf/MmQD/mTP/mWb/mZn/mcz/mf//mQAAzDMAzGYAzJkA zMwAzP8AzAAzzDMzzGYzzJkzzMwzzP8zzABmzDNmzGZmzJlmzMxmzP9mzACZzDOZzGaZzJmZzMyZ zP+ZzADMzDPMzGbMzJnMzMzMzP/MzAD/zDP/zGb/zJn/zMz/zP//zAAA/zMA/2YA/5kA/8wA//8A /wAz/zMz/2Yz/5kz/8wz//8z/wBm/zNm/2Zm/5lm/8xm//9m/wCZ/zOZ/2aZ/5mZ/8yZ//+Z/wDM /zPM/2bM/5nM/8zM///M/wD//zP//2b//5n//8z//////wAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAACH5BAEAANcALAAAAAA8ADsA AAj/AN8oEUhwoMGCCA3uucawocOHECNKbLhniRuLGC9qzIjxzUaNAyeKHMkwocmDKE1erEOyJcWT b+rU2bMHlKmGpkDtqQNTiRKXJE3xTGmqqNGjSD8NRXjxJ1CRKZUgnUq1DkeNbxY+jWiSqlGHrI4q jfomytaHdaJ6XXu0zkeLSliebdiVrV2rb91onauWbUSkeDnGnXutLlKgRZei3PsUocCpZ1m9OQnq bN+jhMnWudmY6GHCFxPu4ezy8lfCcB0XBWr4NGqTo0mPNL2a8LW8SzbLFtm6NuirFl1DTQjZ9m3g Nn1P7L37LHC9wiUORFjcNm7dyiPSbr5VM2benr9b/weOPfvD7cav7cFq0M1niczTF0655L124vZR 560OMX765/ydh594470VoENM1ZffXLgd2NB+C261B1kOMsTRYxE2ZlJ55tEVXnSWRcUhdwyhZ9xk KY243IcdPuVGVMkRKCBKFbIGHFUrUpehS8+NSOI1Jv721ic4wsfijy1RWOR9NO5IkndL9ndkgRyt NdFAH9U4Em5ERgnRjU5ONCFzIDoU2kFa5njQG35dOWCZI50Um1fg6Sjjk3nx0aabTd5ZJ0ddWlkn mmE2FEhBGwVK52xvtnjehXHFuOhwU0KFJUc+pjkfQQjJVBN3e8CUqaY9cTpdVCnRZJefDIVK1qul 7jK1KqsizWTrHprspGtQs8IpUq+1ATspYsKuipOXhNnlkqDyzSrRnvJdU6yx0U5LrW0BAQA7 ------=_NextPart_000_0000_01CAA5B5.DA51AD90 Content-Type: image/gif Content-Transfer-Encoding: base64 Content-Location: http://www.stsc.hill.af.mil/crosstalk/1999/04/subramaniam1.gif R0lGODlhKgHDAMQAAAAAAAwMDCAgIDAwMGkAaUBAQFVVVWVlZXV1dcZ17ICAgIGBgZWVlaqqqra2 tsDAwMrKytLS0urq6vr6+v///wAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAACH+FkdJ RiBTbWFydFNhdmVyIFZlcjEuMWEALAAAAAAqAcMAQAX+ICWOZGmeaKqubOu+cCzPdG3f+Ejsee// wKBwSIztCK5jcclsOp/Q2TFBTaiUqEWhdEBQBg1KIGzYAiCQ83cRAUjWpLSIEaAUtg4A5XDYH+QT EBQCDoNhUYiJikNTVVc8S4EiBYeLNwCYmZqbnJaenySNVI9IKA4BEwsDIgcGdl4jcnsCP5I9mAq5 uru8vb6/wMHAmKDFN6JWKVgnDl4TbmKCDKuxeiQFDBQQDbIUABEUCGh1FAsSWiJn3S0Awu7v8PHy vtbG9ibIpPcw7fP+/wD/AXhAsGDBevss5VMGKeGJfgEjSpQ40KDFixgROmyyEMWyjRAnihw5DKPJ kyj+H2jcKKTjiY/3QpKcObFiyps4M7JktKNKMo8N7cmkSfSdzZxIk15cuROHS3xBix1VepKTJqJT qWrdqrLpj6clYILKyrWs2bNocTL1SgNsqKhS08qdS7esxm4FtCxIFwFBAEF2FshCpwZdiXPqrOU1 nDeIWx1wjZGtS7ny3LU0VCGAtvGxCLEJJ1seTTolZhEN8vYZIQGciNYi1o0I9IaEBAgTShiWcwa2 CFsUaLN2zcIzBdAsRZdeXvd0CQF7tbkpvGWAgAIBHHSDTkFCAAQCBJ05BT73CMMjqIcLUEAAAu/g BZ0a4Kp4T0cMS7F9yLy/WecsREBcCrdBMGBTxiH+t18KyvlnGYAxNBaYAdcJMAYJ6JUD3QKCMRZd OgdwuJd6Ejp23yj5LfgDJg4iRYyKCJ7400uRwWjjjTgeJ6M+Ofbo44IJHiHkkEQWaeSRSCap5JJM Nunkk1BGKeWUVFaJpE8/Zqnlllx26WWWCn4p5pg9BKnfCQgMAMEAH55AGGAupMHhd4GRaadDZqbQ Rm7PeCfIAgfsWY4ZaCi2QAPUlChdCY3Jkd2gI+QhAQDm3WkpEXleyiAuRRn1oqZQZMrgALmlUYAD 3FUDgQMDbEEJBQYIRilfFDQQAQS0tFEONYnVwGmnwAYDIahJ7Jiij0MFq6w/FQ1LLGQEYHnsgsn+ LmutUUk5e6eoO1V77be7NEiVtl5yu4+34C4r7mXP6hgtfkCdKVS63667HLk+misVvUXZ2+JBmur7 ib//FlwZvjkKPLDBDLeIkCyT1paIomUaG69DBDessVYryblZGJM2swo2EKDSWAEGrEkopKwBwAAD vQY2QQALwNxmxe+ieHFyG/ecFsI/Khyaz0TnBOCkIQIQhnoTHDBAr32GB2sA2XiDmwBSY6gmcert SvPU2ci5QKqkSLszjBkXfW8TEmB9YEJCo602c0BzcfME0G2Bod5yNIB1GA047coCAQgQxgQGAGBA bpRY6EADhcMCRNxasji3i0xM8HKl3R3o22z+DTCQ2+cSMFBJBILFNrprDVTylcU0ytuuCZpQpsns iVCO++5d6s7770HDroJPxBdv/PHIJ6/88sw37/zz0Ecv/fTUV29988Bnr/0KVnbv/ffghy/++OST X6zs26ePo+/qt88S++7Hf4/QEEcHXHfn6E1CBBJHwHls9ThDXu6Hm1gQZwJvk1+7hJYhqw1qMISS xWL0pqg0oGEEjbKGKq5DAZghIBAA4NAdFLi7uEWAPQEAxwkHQKoOBoCFETTULAbwtWowai9y8AsL V0GfAWRjPgL4HwlBBT8b+I1qbLkKVjIxRJyZLXbt+hW/hFU39+nuFMFZSzeoU4hzNMAVE0D/wKlE EAEDHKBUDXBAYw4QQonNQIpTpEgT4/VEqKDvMG6QxTpkA4BWFUAc1lgH4e5QANTpLWYxQFccZ1LF Z8VNjQeQmF4Oc7NBgexDTRsjBcoYSW0cwhwUUCOcGLTIRTbSUkUMTSlLOZVTiimVklllHNflyi3B 0hOKlGWn0tYV4N1SEbnUJSN/xrtfRiGYwqzJwXBnTCcgM5kA4eW4ZtfMJTwTmvKQ5n8cKbyw1Ghg 2Bwmw2oJpG6+5Y64DKcyfUZOtlRzCNdUpy60WZp2xihnM7LjuapiFTiukp79sedO3hkEgF6mn8ky qIME+j5zQmufl4voVhjaGYd+5puWUKhE/nv2MEwE4Iws4MYQRDo5i7praBtN6U3uUo+8eMcBZhwE AyJQgAmcLGWt8sapBrCaEbSCDtI5wCkaMLOahdBE+OQRRFXK1KWUYIsQoFit3JAXCDowQ88AR8wm OIlKOsWkYcJlU8dKEIE2cH5gxejCyJpSit6zjt5Ep2TYGlHnACBrK5DNQ0apG+hsphBqaEla5SoU uhINQFW1xlHFaIg+yYEP5YjVFmAjwL3wb28iMIArFrsFyIIBHbAhRJ/sk9RpeUWjhuUKhFQhAsZS R1dWk8U0aDEC18KJTZglgXr+UqfWFmBSQswPXM8pt9T6x60lEFRDS3u24hp3mekj6Fqf/rtNG5xw AhEjIwD0JwLDaPYHJEXqcB+6JdSqlJzYiYCkBpG/vgIKAA5glQjCYxgErAY3AyhEaq7qKq/mQLoY M+/GKIqoLMKpgRkSlFANw9pnIM4LENjsn/rLE+ZC8VKWY+qn5hi78V6UsJbKhMGYyGEYAPhOCE3x hkvsgxOz+MVSGCyMZ5w7GdP4xk4wU/l2zOMe+/jHQA6fh0twvSIb+chITrKSl8xkJuP4yVCOspSn TOUqW5lLQc6ylrfM5S6H73xeDrOYx0zmLJ/vymg2glrJm+Y2c8/Gbo5zh+F1YTnbGVpDDuudrUy5 SQrhHBxK4J5fCWcTZLVWtM3CFgr4/wI9Ukqqg8ZyoUvAAO5u8oUthBwMYwsAp31jhjUEYHqiagAa usI72BFED70wH2pE2isMtDQhwhHJWb1JgodKVJtMRcMwZPA3g5rUISbVWsm9erl5/qZ8ychfq25V hF2Ng0Z+naZU8O07FkSDoI9tj7gJIBtjg5UXfCgGMsSwq214A25teA0c6gG+4ajpD+tgOD9we6CT NjSjgzPKywZIApBegQTcCAGJ3e/eeMp3EGZmIM7sZxMB2cSgXdyjTNRrxVWmOLX8OUvkMlPhtb2r a2y6Xc6lARMG2LZU4skvjM+YfmGQBW8/t6gsqjwGeqUBy3Xpcg5TTgK4TQNNF2Bya/5EDFGnuI3i CtGK9X4bsoQzqgTYCEqdy3MePSch5SKQ0zS8gQGJls4dRHfVlAUSIYb5RtoLZfWrR9zj5QJ5CXpz dmnPHU6Ozm13D8l2fuzc7fSYo9BCdgDaqgJXu9YIonClnZYKFRrfZu3ahX3zdAAeK0PUeCL+fnlh kDh+mocC5zvveYBZUe6nJX2/dKK+0DNh9Kqnh2mii3qQxF6cK92e6+F5+5GkDe772T0QYN/7XGgT +LCu/byKHxGDIh/fFtZn6pkfTbk8v6LRj+v0qY/15pRQ+eDkfjYf9PHsEzc54o+HgFnPTfOzGWPp d8f6q7JA8C+C+ICf/+yJJXxfxf6/JP9yfcbQf233f7ygf0bDf/aHCPinTgiIOUS0gMdkgPPEThHo fh+mShT4gBxzgckGYptngBw4UR5IZ9IXE/E3gnYRMBL4BClGeipYXajUgk3ASyqWLjF4FgLoCQTI D8f1gt3XVDuoEDT4eufFCWw1hIvQgy+Qg9Qlg9tShNb0hOfFghh4UktFhZejhIrAhE2ohRLVUd2R UyuAAK7zAwFnA17oAk4IhtnyVPUwazD1QZtUBvolCE1zAKZTDhHQCgfSNAbQRWxgRrlRRn72Oleo Z5vnhnPDUhjEIX2gK/AWbZ9GGIWAYFpVKIwjGOSQhjWwhmzIiEXjiJNQaddxB//+sxnZkFh7x18l 4Dd50Hdc1VuI+IEYI4oWCIeoUQcLUB9VZzWNUYkrkyGy8GxYRIstJoU1iIsDZgKA9osw1UlllDK1 goetUAbl8AbbUAJqFDr8Ex0iFVWHwlf/pYxGyIzj9APUGCLIZoLaBxLoOGJA4AC/mHCJuGZiFY8L VX/3CIIZpY/HtQIN4F9CAAEcwgBuVGG22C0AuTYq0AbY0AI591TB1Yq74kCC1Y+n1ZCjASCAsl/e 4GnzxkGOAjgCAB+PMh7fEQButBvvJpLd8UKogJKFcAru0QKg+EYcaTstwFuf1jW50St+ExiQwzUQ 4DRCNEhu44rBQZQp9BrWkF//pLWQG7mT1tcCecAhN6keA+AFW5UH5gEzgGWNLKl36TBh8WZT0SGW 2VZwU+mO55dEVokWw4Iyc/BuE5YGE/AMfTcppxYGCCBhF0Ru52Fp6gFvbCIBgOkKs1Zvb6aRKtKG xmVP5HiWv7Fvr1GZLeA/s7FvNFc2cPl+DzeX09QEqVAf7agzdeZcpLl/2pOTt9CaKMGF0EeVOCKZ W3gDhScAPfUFljZI1lF5xQCbwyebvWQDvWgCBzBCuTWRNOCJamiOWdiQtfRHNFQ1DIAACKY3qjAz t4IKC3AGgcCSJRNZdaI4iGMAQBcoGWmbWZJh6NhOfzRfDKBGLHMe0IGQu4IG1mziIbrFABAAM64I nZ8onTwjillXAwgAC3lxK0c5AC3JXZWWW7FCArw1apBCoG1hoE2Bm7aDXPRRamZ5n5NgRkFkGPCB jSUzNtmgHmNjbMkImShmWAmaeRwamR6aLbTZOzd6IyLGUTtKaDL6Oz96LzX6ZMR5WrUzUbezZ0mK cFE4pFBKZU86pWNSpVb6JViapTwqpVyKpD36pdmzpWIKJoNVZmiapmq6pk4yZCTQZHAap3I6p3Ra p07WAnaap3q6p3zap0nmAn4aqII6qITKp2V6qIgqZwAAOw== ------=_NextPart_000_0000_01CAA5B5.DA51AD90 Content-Type: image/gif Content-Transfer-Encoding: base64 Content-Location: http://www.stsc.hill.af.mil/crosstalk/1999/04/subramaniam2.gif R0lGODlhmwIOAbMAAAAAAGkAaX9/gMZ17MzMzP////////////////////////////////////// /////yH+FkdJRiBTbWFydFNhdmVyIFZlcjEuMWEALAAAAACbAg4BQAT+sMhJq7046827/2AojmRp nmiqrmzrvnAsz7QV3HWu73zv/8CgcEgs5m4BD3IDaDpNz2C08HQCoE2jdsvter/gMAo5KA80S411 nWW33+64/Mqmru307ESOr7r7fVZ4g2KFhoeIiYqLEmRmaDiMkpOUlZaXmJeOZZBJmYxzoaKjpKWm p6ipdRSqra6vsLGnrLK1tre4czqbZxlpGATBwsPExcbHyMnKy8zNzsRqeiDP1NXW19jNElPb0h7Z 4OHi48V73twd5Orr7MvmuzdmvRi/F+BswfhzBPr7/mv8SoGLlspYE2JyAjpRKGpYPysMFwpbKLEh wDgBIxa7QwvVMXz+EwEqRAiAJMmSGBMedBMRpcqLoQxy7OZx3L+KLl+eatkSpLiZNHh10tBOmcSi SJm9KOoz6b2DA1s4nWq0JNVhLq5q3Yrsh1BfkYBxPUkx58qzZiGODXlrLVucGsumPboWl9u3aOOy ZHn3VoG7+WCm7TnXqluv8R6B9SQWsGN2Sx9LJpdVnIDLmDNr3sy5c+dinkOLHn25nNTJqMMhDiBv 6KfXsGPD8GxBwITLEmwX0K2bQufanC/w7l0b+OYf6GSL8XbhK72wyqNLn66BtPXruCtg3y4aOXM7 GVKLv0Zrg/Pm0KmrX8++vZE/EsJBDRk4I+GRGvG6rE94f8b5UX3+cJ4N6bln4IEIJjgDYA+N50wI A1ZQj4IUVmjhhRswCCBOAtE1FoSJcbIYhiSWaOKJKKoQIQUTViBIFLoIQhNQfHDUSiBXmKNHHIHM CAeOeSSX4pBEFmnIihO0+MF375QnwndMMrHKkzk2yQ2TURqp5ZZcAoFkIwV2KeaYZJbpxZcFKDmB g2w+iIGQGbYpZ1eE6BjCnHiaRlNHd+bpJwFO1oCmmvFxZdF+sWyVgZA8FvonnnvyaQEfH8V0Vn9y JaRfYIf+59FeImFl56guwpcNgPQ99A9PTaUEEVRoeUhNoEGFOA96jNnzaJ6VyVeLffSRtapgamXT a56oynpXZFf+8eWMs6h5aatrjWED0rBWxTpfP/Vh61+x2IwAlKOPKUtOqOuIO+65GHk60oaw+ucu Re/mZF9T4VIpIwWGxsupXA69eqmmVEnLmmLP5WrBrryGsG8FDM+pb5UQR2yTYeOoS3HFFjODccGr tTaimSSfSJt2v+2WW2YrH4eyyy9rZlx2KfeQy80453xjJyInXPLPJJ7sG3G5Fb0b0TXfFhxwHGSn NMw8wBlexw7SimvPuAKt9dZFqtIvtvgJ/C2+Tk2ZMNYEKsz12myLSXXVIAwaZtt0132irya5W29e +f176b2udvtxvgJOO7LdiCeO4dvjgXiwiD4rLvnklCsot9r/lWeu+eaYoEmBPKCHLvropJdu+umo p6766qy37vrrsMcu++y012777bjnrvvuvPfuO+6cBy/88AkSSvzxyCe/w+VS5ujKIHVEX6MpkgIp C47KZ6/9FswTIfUW328v/vhGdE/++einfzXCWXMePuLovA//OVlKXn8L5utqLbrXDL4rQZN62Jo2 Ai2jTEVW5kIGqirljjpFqlToqBS48OM3ZbWrf/4DDKkc6KOFXQxj5log4wDVDXg87lZpI8qpDPOr bcGEP9oaG6aMNbFq6SdeYhuYC/lTQbjwhS5l6aFaOBTDBWpsY+QK2KZuyL/+VJAhTgSVwGAYk735 74j689XH/mKotyL+cB+eqqJqfJA/D/blggxTlwrdAqqO6WtqZ0Tj/xyGRH5pSI6PChn7UghHNm4x W4XJoQi1QgKz/SWOefNipgZ5lSNywzHJygsYm8iViUXwjgED5Ax5siwyGi5y6gslCYTWstCUcjhQ OyVmZlYzmR0tlaIU1CfbF8taeuBkRCuOyna5y6Sp0mm5/AAqXWmz+5lxhJKxWtrQJqG52fKZQ+OO NIn5tGla0zs1WaFJ/AUvv2Xym1Dk4RDHychnjWuZe2wm5qDJzvMhM1pxmyUf20lP9N1xJ8TKICHj eUJq1fOf23vnZBzHTBY5E3uLkl8ABdi8FDCUoXSsI0An/oq8MkIwSM/j4B82uqM7jMKBMvKDKHwU Uj5Q9KTBsyhKV8pSDKl0cgqlW/yMOT+KxVSmEn3BS5NYNnj9EVjBYlNCc9pHgXZypkTNIqfAmao/ 6tOcT2VgORlIng3e9JD7I1s1piqTqE5FmTHYKVb7JUREwUJRlpyCFpkq1QQ60WNezZgjBaFAHHJr FJgCJAhNMS9L5fNvRiykqVb4RcGx8K6atCBeVcXVZBisoEk6qBbdCte42gQyp7kYEw8LLAQ6y696 WWqAVsCuCVLDX/sLG1KY1dMCLsOnqXlsOg26Tp5u9bCDyc9fMwnYxL4wjFp9hl2MCsm2aJa3vfXm tcZp/1jBQJGywh2uVuDyxBdy6z6N1CPkaNlS9Vnzu7kE7zW7awKxkjd7pGRaL41Gs6VFk5rqrc4w WeaDq7LtfuY9b/LEO01W8tc62FzoJZu1VwJSkIdwKyHPZhvZ2ur3wVqCaF1w+1vGNmSfhevn4SDM Ya3hrSptaqxS+AlZMDm4wyjmEnEdQ1AGmzjFMI4xI/Ir4xrbWAiem8DvdszjHvv4x0AOspCHTOQi G/nISB7djZfM5CY7+cmZQ4KUp0zlKlv5yljOspa3zOUue/nLYA6zmMdM5jKb+cxoTrOa18zmNrv5 zXCOM5qVcFAo2/nO5ZWnOjGUVDz7uUzmbdQbdXAlaf6I9AT2/bOilRPojuoMFtCjX5VEqgvrOQ8O EF20pmXT6H3ZiNI/kt6LPIrpIEXapjwC9YwAASSEbvrV0aFxDRJdX5rC+tbTkTWud31jXfP61zD2 9X1tjdON0Xprhe7z/ExY4jTVuabBQ6rwkj1tZatIz7Rd44rXAsA+bZvbGiW2bb+dXZCKe6zkRit4 joDtBmvbGqg91XQt+4xuX5QjMmFrMuidVWFFF64NlLaA70BVIB7l4B7yKb+VSNZjWBVLPCptrD5S VzcqWJYaBqVSb/vcdlm4sxX27RT1Fk7CRfRh+WZLXoUIOH9zU69MpGJo8wlcYwh2FW21q9j0HcnF Bv/Om7rVq6aC6PC56qG0ifShb8sq8uQqHLra8GTGuXtMAi/8T6xVh2vHA3XHZnbFIlbHsdKNYR6I leyVHDvanaL2tSel7W4vinZRuOeijkXnOVSuadOeVo5NWOjkBHzfDtN3O56xrEHXOYtPPmA/BjK3 gb96uqTebOMVN+n64CQ/7vkrRDLcVT6RvMRlcflv3mSIxbVF6sG5E6B+iPIudraD4w6yiNqd9pgF gdRw/1U6vpv3uTd7u18MbK39lzv+PX537Czs4g8Jl9V05Xx/GV739saXr2wvLHPw6O57//tJbb7z UQR99fKGvfNdZczoy0ulNS39TovauYdtnuHLfvz+P0vv0W7D3pWtX/3vxX5P837/F387cFXAt1pN shiV92z4VyTKhx3JF4HYx30Qh3JPQUlhRzVgRVsNeGIPGIIl4DUZuEhzoRd+JUi7JUZZFT7iJ4Iw OGhVB28eNzaTNBjX9XIjpy2j1QEvGINAiAIJKHckFnvGE4RICHdWp3lpVITbNU9JGIUxwHlLlXcq KHpylWEfKIVcCAOYpHLAtVxdp4BaaIQO2IVo+AFD2A4t9oR1l4ZwKEpiJWgQ1AHJdlPUdnJ7cmwb FId+OAJzCGqkhgodMVKhYGl0WFKmhjN/2IgcEIifNmoYdYiFCCODeD06ImrYY1Kl4IieeDZmCIL/ n2BIDjV/n3iKHhiKqLiKwvODQAN+sBiLtnBvsliLttiJmXiLuriLpBhW9neE0OY+zMGHryhp1cZs qnh7azhGF2Bfyzh5AqeGz0gZD2dtAzSN4tCBOvWLkrUVHGJgm5c34qhBbwIlEoaN2RhuSHROXSVa n3cSKfdaY9hUzbAtJpeHlTiD0bVXG7gRNIiFWbhuGLeFythagqdDTUV0W0eGA8dX6DhZxlYQEnSQ ZlWFOxh6bQR0CPl0YSMvIrFA8EFSpEBY8KiDoYKQYViSi9RywVWP5+SLUweF+viQ1WBJG/ePFkmR YshwYIhwmlRdACkqjDeTbNJz8GSTfndALTmR//NYbj1wdiU4jUp4W4rEWYrHk0zXcaF1YFXFAqMX lOyibl5JFQTjkhIzd/5ElOZEQR95kXl1hVf5V3tXb75XkH/HFHwzebpHVF+pdEQ0dJGHg48Hcz3o AUn1NVk5czTHhG+Hlhumlma5N/nAkjpULPaIWidJmSdocktCKTRJQ2n1BGGpcgopLx15miZpVh7Z lHQSmqLpjfHmXHhRL80Flm4Ce26YbXaJFP2YR6FpeJ9pDUZXJUPIml7HeI8Edkspdo6pcZAZnMow ldB5nKRlGRRIGqBxnaMBDdI5nUWHm3Snm6x4IOlFGqdEfRNYgDAjfe71ZK44npWgf9d3fr2ENP/t iZ66ZIC+kZ61Zo04VX8x+Ybw2R64FF78t17oB0utZKAboH3wRWjn5p01uYCgmJvuNqDuoZ2jMYEa CoDyt46/J6HRSaHrY6HEh6Hs0aGmpJ4qGkyzdoELiZeA93JAyYOU1JgXV6HheaEoKoUheXcm2Fcz 2lssJJY+yI2i2KNJKKJ0WYYmen9K2ogQKUK9GTFtuKMnGqVxOKWsQnNSxEWnV5VgeaVpqaVoOKVK x3Sp2aUyl4O5xZmPiKRm6odMept0FqDiOad6Oibvuad+eiB9+qeCuh6BOqiGGmtIKmeKuqiM2qiO +qiQGqmSOqmUWqmWeqlT1myfk2Sc2qme+qn/oBqqojqqpFqqv3OoqJqqqrqqrNqqgIqpsBqrsjqr tFqrtnqruJqrcnanutqrvvqrwBqswjqsxEqrd+qqyNoewJilydqsjCanzhqtnAatQ+Wf5WitLsAo fEiM0qppnZZpdmiKI2iMeyiufditMNhprMZRl/Yj66qIUyKJlFY9qTYFMTIT4IquuKauiPgK/Qpp pMKIIamvwKau1PNRs4Cw7vpAgggfnFhq5kqwTvatDQsI9yqv03OIh5aLGHuxk1ixEluw1Hqtw2ho 5Gqvkwav7mqJiyhpmmhuGMWx5xqy+zqyUhCxB4izNAtrhbqzPlsreMqjPzu0XdCzRHu0K2C0/0i7 tHkWtMxaObwYtVIbkVNbtbZYiVabtd2HjE+6rHWjtWDrfbQYtmTrFxxbtmgLsAOZjJvDrSQTjW1L rsLItVgKpXGLrVwDt5qDj3FLt2UKnHVKnQslAoE7YnrLAYU7YjDrbYnbmjrrpHVreY3bms04f5NL J4cbJ5drcw/EQYi7uZwrkEBLkDYEuiR0rYNbulRTpcx5uHCCQcFZjan7nIlLojIAlS3IumBoMfZ2 thTTVcvpjjJKFqcVV1zVuZ7bauMGLZaJR2VZvOIhuw1JnB+EW+1IXLYLk6R7k//mk3gXpF6ad+EU vrobDASRTVgJOCsppgSEd0+EXLM5voGpmv8gKUAkCI7ei0M851TYZaOC16U1CL57Zzb3i6YrmZLk e8AnKUij57ePmZRU+Zcr6FwZKUnWRaTle7qVS4jxqKYrcR95GUYAfGCFRcGCiRLYhVxji1fA60Jt +VPpmyh/WaNvaS+MRMAS+RMW3EWqqb5fasNA/HjVi7ckgLuwK5koKMBBJUMkJ0P0MpdRh5z66ML2 snkXXJrI5cR56ZM+fJEfDDCmdXPJib9bXMaRZHqXCTAd979t+b4daZrcKcWAS5JK1MY0ypYS7HT8 iGCgKXxOa7fcu5Y5KcAZqXcyt5WKVJh2qLlZzI/+66bCQl07Cb9h+I1JHMZUooxU7FR4lML/yxVz AIzFVkROJSeUe7mb+4h5KCmmftnGl2ybGrw8NgvBSgnLjdOdGFTFcmKcpqwC5JbBwvl1pmu+4Pm3 1zhvthy9uNxveMLLvZwCv5zMUVydwywMzUl1tAybI2elyyyj29yEYxlHMQopwgyknewn1yyT2Txv QMfNQ0m7vHmV7ryXTFJ6gzdH9CxRq3fPvvmUszzHX2OCjMWYfIec0sB5rPyW5GjQR4fQPhekC53P +5J6rkDQBS3Lfyy5h/d5J/y8Ad15jpd0SdwTDv0K9pyTIz3OtQzSId2THU3SneTPGd2NGx0sIxm8 OMrQOfKFNl3Ar6fT48bOaJzDhMfQx3yX/y5N1D/tx9sLz3XqSIE8uVjk1E/9zut8ueksoFQtolln ut2sDC36GcQQ1pyhJ9RczbHswM551Y3b1aD71clA1puRnXKNGWbty2hNzDLd1EzbA3UtgNTXonem tH0dAvKZMtOXfYCt2B4afY2t2PhZTGk72Wqro8Zc2DVQoEazfwjqodjXSgeqn+4XgI9tgUSMbOH3 z5id2RU42vRZfQp6n+iXAbD9oKZ9jAvWtWe42i1QoPbZf/UZ2Y4NTL/doPDnojTgtm+b2jOdpLzN AvrXfsBd24vd2h1A3aX9otaa1xqMX6r93DHw148t3qKt3Yw7zNpoYnwN3jJA3thN1gF2UcdR7ZIa KM1Clb2pqNvOzd4o4N4cqqLxvcJBLcj7Zt9aF5De3dz8PVE+/Y+EvL4DbWEquHIlmODrveAn5ZmE hcV2JZmtEnIztMPpCLmXjeH/xI5PIdKKxzeZF0SYCcSamcFk+sAm/mAGvHSjLMkz/OKG7MOKnNuR u9s1TlHcPeNrPeTnVeROGOT7jeT/VNLAHNN3euFOzlJfeEEIp2/gRuI0XuVWLs58dWFFPeVs6+Xk xdNvLJc36pRHquBmnuSeJ1o3vea1R+ZP/ggBRcpQ58VZb979B0NxJEvzRFN1ZVv3hWP5I2r7xnN9 53v/BwaFOFIlMEAOOsZZ0/mERqVTatV6xXqGW27X+w0WK0klh5lFp9Vrdtv9hnPAc3rdThwZyUtL 3P8HDBQcJCw0bNNL4gtwAXCscgSAilShPLzEzNTc5JRJRFrsiBzlIC0ZfRSxlFhVtaQ0RWntpK21 vcX9+yzbONtAnZ0I1piNDWkd9jCFTU1JzoWOlp6mFjtSNOv7ZQbuBi7wlhQORwUnPzc3Xj5/7E5n f66Wn6evJ9wNLeWGD2dt9oeXDmCqgMhetQv4rpzChfYcPoQYEQu+bIz0FeRnjKG4b+gy/v7rx23j R4TiJJ5EmVKlCYq9tBH7xlBgR3cXatIkuQ+kt5nlfBJ013DlUKJFU7bU4EtezRcajT6FGvUk0gxK 5zllEU/qVq5dbVHFYNXrWLJlzeICe0HsWbZt3b5FNAabS4twneXEm1fvXr59i/kFHFjw4I82CR9G nLjvlLQT1todoVjyZMr9DFfGnDnxZc2dPe9lLBdURch3TZY+9U8g6hMGT7OO/NpJYwovYR9TffvD X9m6le3s7Vtf6Guj6Xa4k1z58h8XSzCHHl05QAxaM0jHnt3LZe4jtH8H78Mw8T2k5YCPdCO9egA8 HIXn4pwEfPr0qXcXkeM9+/bsd6wH/mO/LQSED7/V8qsDwPoUrG88KWgr4DEJ4COQwBreY9BCG/qr 7wb5HNygQxGzu28c2ZLRT0AGCcCQwws1vDCIUbhoUYj+YAzDRAP9Oc8LcubAsYcVCxyHvLmSsg0D +vbzhsX2mPwRFSeHlO5DEzUaMUvoSuTyyn90WK+bF518MZwpW8xQyjPRDPPJG80Uc003Z8SDy2CY KiBAC9NTE5g1x/TTzw3Z/BPKMl3sAkQoIJQwTwrF5NPNKQ+dFEo1v7soJy03TU6okdj5D9Ia0Sy0 Rjk5bNLSSCU19dA5AVW11UGFykvPFFmV0tRIydx1T1GfJJPXS7uIKQpGk7xgwVV7/n31VEknDRbT bfCiQ0EqwQzyVE5vVa3WUEkd9VU433xzUP5iFRfXc5sl9c9oZ1WH2i/oPDNYNp9lFtY07620XBX1 xGqGY+vSoEMVRXV1Tn7CS83TBC39oV33UL0Wh2yTa00oISkOlNxlFZazTIt/bdPeKPvMtdw6ScCz WkFNFhRlk0OGd8aU6Rw2volE4wVJgq/bNmgwUkPuYX3vPdlFXxVGGFQ6TEBR6I1HRnTEjHvULk4b X942i4GLljpsIWK4IyNzYx2Z6nOOxvliIMgWG1uqu4Y7PLd9qJruK77G2secacw7bh3qdhncS1td 8WCk821b5DkIB/Jvv21lUTsZ/74zcwhfg/aa53xCjHxMINELfGgYOuVYYl1h1JrSeteG1vQXyi5d c0PnvVF05i63+2651RO68+J6rgrZCQpv0+aU2U6a5MUz/6Lh4LatOHLJn24Y9IfjRNrZm5UHP1yn H5dee8zRDf95pteXdm/PzTN/8uRjxpd9bW8fF1Z4yWeZKcH/3wHU/GeHChkKbbPSXwL5Zbj91UGA xXIU6VRWKrWJzID0ItHOhve5goXOe+zrGL06JrruXbBJuBMMAFW4Lr/QTm2r++DHYJgqQlXPdoN5 1ARBlT9h1W53GizPcfp2vhUmSnoLKeIKM+aOJDLHdzk6oimauLXrLUd4QfQZ2P56N8WxHRFoXBTb A3MDRgL6UGf9GwYZ8ca1KgHxSMX7mZLUGLZTaGGOweufFsEgAD720Y9/BGQgBSnIHAzSkIdEJB9X pgo73lFvVuBb/BypJchN0mqnq0MiNblJReKAk5885CJbYElOXfGNYTGecFR5C1C2MpAZcGUs+7hK WkSSlre0xSBhKUgJCKCXfvwlIHcpzGH+sZi+LAAvr/AZZjazMIvAIhxxOc1a6DKZyERkMPmoTWBi QJnX5CYxrznLcRpzmc5EZzoh6LNoojKO1ISnIax5AWSCM5m9xKc3X6lPcdLzlxyo5wS+CYncxLOg anmfEOO5UHkqk5wCxWY+sf450X6Gc5v/pGdAN3DRgVLBOrQ8qGMSmkWGlnQQ8/SnNyV6zIdCtKIj 0Gg5u2mFjyaLlJQsEjRPidB3mtSnbZClLFka1E1ioabHu2mWFMXOnYq0pz+FKhqI6sqhTjWb5zxR Gs+ntM1VLm1xWyoc28nTqJaVDVYFZVXRilKPFnSAW3Wcf7wq168qNacVGatTzbpXvs7AWxLMn/o+ 0ruAOTWvtXlqXxW72BX8KIceU9fBKKgt8FjjsBFKJWM1u1ncYGlJ6uvhXhiWh5FKk7OnRS0IgjOh pF5SBLZMbWxlS4PWisiyTUXsbHW72w7WtkGk3SD8eDvc2SJvSE9U422NQ/9S4jY3tS57IQjTB1nX 3ayB3zNjc4B72UY517tmNa5012c4nIXMhOZNlx2USzx3fte9jIWuXFcXQlfRtX4JrNwF1btd3GI2 se8FMEPDWzL0yQy79w3tDDFoxNeWtr0BhjBUfduh9XIwwheO54R/2+DgKhTDHwbxVBxM1hCX2MT1 gO2JVbzir4xYryyGcYw3kWIZ19jGfqDxjXW8YzTkmMc/BvKiXJzbIBfZyLMZsn+NsGQmN9nJT4Zy lKU8ZSpX2cpXxnKWtbxlLnfZy18Gc5jFPGYyl9nMZ0ZzmtXsZe6u2c1vhnOc5TxnOtfZznfGc571 vGcnX/YCZAB0oAU9aEJUF9rQh0Z0ohW9aEY32tGPhnSkJT1pSlfa0pfGdKY1vWlOd9rTmA7Bp0U9 alKX2tSnRnWqVb1qVrfa1a8e9JFlPWta19rWt8Z1rs8SAQAAAAAAAAAAADs= ------=_NextPart_000_0000_01CAA5B5.DA51AD90 Content-Type: image/gif Content-Transfer-Encoding: base64 Content-Location: http://www.stsc.hill.af.mil/crosstalk/1999/04/subramaniam3.gif R0lGODlhQgLdAMQAAAAAAAwMDCAgIDAwMGkAaUBAQEtLS1VVVWBgYGVlZXFxcXV1dcZ17ICAgIGB gZCQkJWVlZaWlqqqqra2tr+/v8DAwMrKytLS0urq6vr6+v///wAAAAAAAAAAAAAAAAAAACH+FkdJ RiBTbWFydFNhdmVyIFZlcjEuMWEALAAAAABCAt0AQAX+oCaOZGmeaKqubOu+cCzPdG3feK7vfO// wKCNQBQaj8ikcslsOp/QaJJIcFGl2Kx2y+16v2AnlUFmqK4oS0AAEAEKmkIb4IjDUY32qRHu+/+A gVt8gkpjZWdFKgt3FwcBBxcaFwWQGSV6K4SFnJ00AKChoqOinqZIm6c/h2SJVT6ZKqmqtIAACLi5 uru5bwW/wL+xtcQ3s8U4rGYpaD3DJm3HmLzUu6DI2DYADdzd3t/cvsHjz9nmmufJRGXLKM0sF6EW FgAWGnRupZqk/Pzj/+WI3apG8E1AQdvAKewm7p+wdBDRRaShzNWPgyTyLNzY0KEcVf1ChvIIDKOt jQv+OwKcyLKEtJYrKjJT5KyFRpTgVK48lRCnN52gFAgdSrToUFAGDJhk0tNnOJIPYbZ8CaPpzxH0 6jigs9Xe1gCSuo54cw8OnQAJ5ISyI2KrvI8fWch0R5NHuWsjbjp9CjWuKas+AUQYTLiwYcIAkipe rHjpEsA4dQZzzGkgQWqgSp6iKvXEXM91V5CVQ88ePhKO9e6VPBnkXoaHYxtOzLg25SSQUbKOyvM1 X6i3vXDuXOKziXc7Uufm2NfvCVAPokufTl06KF3BYyxPKbt7BNq1F2c3sl3hbueeyudsPn4QDsGx M2XV0LV+PbXXyJJlYrx4aB3K+ZYHe6hVYOCBCCb+iKBl1bTngnrfwOfdYeCFl5SDQED402RGdUgU dNWFOB2DuGDIgoYMEWjKcMSN0B8JyK2gBh30SIDBaWONFZBqTp3nHAAKBqngQKBQYOSRSCZ5JGC3 oRjOhPFZKB4XTuYB5WxCZlmBZQAo6WWSoOT0QpU+migFiy1q8KKL/51YxwCgBIABJjaRqaIbWmaZ XXlNCijhlYNVaKGZzvgJKGJ5Bjken2P6KYyHkAolaGNaoCmakPLhsxUE+LXxUVcJAMCGA2Kd1mlZ cZi5pggx5hCgbwYBEOmHiSqqXWSNwnooYlJSugWZu35Xa4KL4vqgobtOaptIIQFhKXGrqtnmeyf+ 8hhYsEAOu+CtujE7CrLJ9nqht+SWWy64gGar7ZYy8GkusOGCKKKI1/Hy7r3bpCnXOojM9MoLGaxR 2j11oLaUtZFhu+6BexrbArzhikuoXeheqa62xeqW62t/WrywgV1+KXKYKbnwbAoBZVqHBAVc0FVW 80kQAAAT3BPALwvMt98q/Lbi70V1VgzlxRhzy9HGqwX7nZTyzlvdLROjAHG6H2d89LG6Jvsxu1U5 LJG+/rJjESxBZ30o0cM2rDHWHCuNdqIk9iLyyOEg3aPCC1tdMttJa1210XuzcLJU0bZKLR4/CT3h 23ADnrifii/+d9fmCRhO5N0xrqXeDFWJsG7+eOftOEM9gm1Fz+2A9m+7dfxCp+nPwQ6N7K/T7obt MxQ+Le6CXDAP77L7LgnwLOm+ehdbORCqBKiOdQd99WyV6vT6mWU99BZ0VX0f9MCRfFcYBBDA9/Wg OpoeEgAgJzHzkW9BPAMcEH0mELwxgCft3+fp/tO77vokAIif/kgglgvAqWDE4w/qxpbABjrwgRDk GQHE5i8qWPCCGMygBjfIwQ568IMgDKEIR0jCEprwhChMoQpXyMIWuvCFMIyhDGcYrQja8IY4zKEO d8jDHvpQgj8MohCHmA7jMSEPIhgcEZfIRBsosSVGVEFczkcwtqDGJk3MohZ98MTiLfBnZBP/3BZZ ErfLlEgzY2xCFycSxZqIMY3FsBNw4HjEBLaRBQsIxQICJqryXfGPLsFEGS+TmdbQ8TGOmuMhk7DG iNxxBWvAwAACADMc3U5wclSkQMyImZ30BlZ9idoi3wi8RwKobBy7U3oSqcm/sNIjohylLN4TuPw5 AAMWuEQGJCCBS2ggAxOwhwguMIE5+e6XFsAABiQgTN9lQBLAnIAvJUAJx5jSVahcjSo7IcegzMoo ylIKlV7pkFjKEg+0NA9WQhGABWBPA3ACwAF+yQYAJEAD6VOfJOIBhwvMDADM+18G4nnP+q3Bml+k y/FUQJZLDMyS94ioLDJJEnMyBXOZE5c4/n9FTk+ChJP2quhm7JhQ1QENkx316IkA4DR6BWFq6dKo RbWRUkN+MpWhHOkNOoaYdWqlHtm7wAJqpgEIHAACl3CABCCwxyZc83D7qKlN74GvUBDypRiVTTin xFFQwvKb4GxpiKCGVa+2khONzEc/SCABNgSAeVtpqzszIL+WxcGdBXCnAifYL4We9Dl6+BxzcjqW rQ2ySHNT0lr4djdsyXScsKpqVRuUWC9to09mFemKSMpXn/nVBWkh2EMReMmoZhaWBZqc3cTE2Gs5 VmKQbZvf1sU51j4spbICa1GQslEspPUemPIp9FaWAAdcQgIOsEDOJrGABNQse3NywASo/onP4l7i As5FjwyeutNs9qickNLc5kYXodUm7LW9mmkNYDo01Zr3Kq0979k02tszvadZwywA8yyQV7oGYACS kIMD2qmBAwiAZRBgWYCZ98tHABifAkgAI27AXW1417VnMyx54Xtbs80Xtl2VbYZFt2GGvJcj6BWP WJ/GJWdxloKfbcEF2CAAe+TxDb7MUe3ywl7JubfDVwNy3yKW3tgOmWokplyQq+Vhj1UNOyGrLJKa Qpnfpky49rHHBAowgDrQ9Q37NPABkiqA+j3PBxUuwSR9SQ9pQvRVIkZykuPLYSbHOaYg1kKPvSPe 4Ja4bnRG8Wxp+2ckYnGn+B1mWhIg/wnoTiIBB6hZBjhl3F86oAAQEMEEEpA9BksQxiYNo2nv3N4f 2zlwp24skZu24gfUCztlJbWPCV1ozMqazxpWMqpnSbw0f+LC8pUzrU8MaCGresS0lZuUp6zOQHNn 0Glrl9dEk1UK5ZrYhiYl73yNstadWaIaiEy1sWRqaufBcxEaN7mHzTZ0p7vJpWY31tz9uFtn7trz Ll0OuZ0CC0BgeOcMuMB9yO8nVOJmmp5ZAGomM1FdYKCgaB3/RDWz+mWGLFuB0/3kB6dvd6F7wxVL qAKYgVIJYGYLQDn0RJVjVdhSLGrxHh30S7Mb6bUQ+XufWwMaQAAwYiTxkzhZMFBPUv6FAgL/JOrA 2dTZ1B1nd1h4nwh+p4GgjmACDhgeBiBA1F1i4JhSl3oG3id1XGbA7GO/kce9gFykasDsk+jl++Ae 9n3aA5gQmBMxxj4Preey6nJXpjBFkMdT7FKpvnQZUY95TL5fAOAWCPwkuD4CZ7ps8EtnVUmfvtDM e/7zS6zhCNhB+tKb/vSoT73qV8/61rv+9bCPvexnT/va2/72uM+97nfP+977/vfARz3oh0/84hv/ +MhPIw2Xz/zmO//50I++9KdP/epbn/qn63zyt899fRW8++APPzK+zwPLmf/86E+/+tfP/va7//3w j7/850//bry4r6FOQaiG6VZJNP5cAACXAtmWROJXgGnyW+eQZlMEBx+BD/+TESaBgAY4gYEggdmg gHpART4HT9+WGhT4gRNhgdhAfg9yaCCIBa8GUgiwWCcYSL22ef6hfd2lbS2oBINESCpVgyI4fjAI I1C3XiZYg+QhVbwhhOF2f56Vf3YRhEaYIUSoXS24g8VAgg8jNRkhNSqIHWjUhPfwhOo1cFJIDAXn PwUwWq+DEYSQgiooGV+YBZJFCxRVTlx4hC/YdAx0IgxYD+XzZsCmG9tUCzdYEOLQhkgQhzkYhUjo dDEIAy4zPGOXd7ODUqclh3AYiJjBhnDohXMYhrVAhSfSh4N1VpXxhFBYCIZIDv+biAT0EAoDUHJ+ pASlGIOgxnl/NVGkSIhDOIkcolsfgou69l2iiIg7VUuawhVARQf1kz00k0dqcD9vkD7KZRATACep kjP1ozz6Y0AIZYdg5Eajpk2E5UqnlVu8KCmM4YvYth7BeIJWRow/BQEvMz/KOI/bc2PVpAET9j9d oWBwsY2zuIiiBljhdoqoeFOrQY68yFuDYmTAqFnnYImdhI4/wIm0UHCTBAECAGBvsCk7hhcuQZBT xU3qVhhb5St6polxBJHWoBISyUWJeIeiMT4zY4ZU5ZG2qIsFKY72llF55oYoyRP48oeCQJGqUHDy kwCRFEDy42mldZM4tY6BsGf/fPZYIQaODqmTVnmVaIUENxIrehcGnigaoJgSQgkNLNVq1lEisXZk eFZkVdmQqOUaOBmShWBlWoVl8tgVHzFJ5rN2RhCWDDWW5iGUrLZiaqiWTghvQ0OVJzmXwiBZ+LKW cBmXnmCX8bFO7sSMWQYXAdCXTtWDTFeLAgiSWwhcT3ZVibmTWsWYPjmJCKlbZ4mW0dFiqZmVlNkJ RKkC0gMGgIkygqmOwDEKhmVGkHkuirk4MlWcytkPjvKa31SYhnmJyykStzid1okvuWkK5Id19oAB pIJ5TZkPH3mLqZU3K7lsS9ITtsaWFsOaKDiSgXKa2IGeS1Jn5vaUWlmXL9mN/y3wVjSyh6RVk984 mZSIJ1UTZei5HOt5bG3JNAyJYcKWNoiVoAqajhECHOV4FC3Ja0awiqEQgC0gFi0Qi9n3jz4og2Yp DwC6Y6NJnoVVbgw1bTF6nFPZk+9Jo1oFozO6ZPdpm875nIKyoQL4Hn5WdaLScRkgB0m6PxmQR+mT XMZoGjW2pEt6D/dDjUMAmpqHokAoifgJEOCEb8ZmW6kGoQ26kG9ppvFWNLVmoT/xo7NSkvUVBVZW pLbUj9OTR8mTFaVSMHhKFlhaA71phV6alXBKKzoqNTKKMvAZKO4pBVKZo3PmbAPYowy6mI/6BNnp AiS6VyYamgHZopF1L2Japv/lRak54TaZCgWRGh+JqqhrM6bB1p6DEptoSVYTuZ8xtoSFeqmzJm+m ap87yp6Y6paNqZrWNqmyKqyMiqNR0hiymZbW4GJ1+KlbKpqAJFjPhmzAaqlk6q2zSqvG2prIum5s +ovfOqy+ymeSdYn0SQFhUmVIEA8yCQAu84rt4pc5MIZ5SJM6Bm4moK3mETrdqq7p2qzlOhur+gSt mqwFa7CnuqyCxq0YUyLvCq/NRoMzcGVTZ0nxCKVGhwGuUz06o6/qwI27ygLyUwf+Gp540LDmeq6U uqBqWqwOmqbh+qvR1qaoGiEEK7MSW2xf06WRaKT3JD/zWEmX5ilm4Yxl6HP/ahcE/AoM+FQJN/ev TkmsuPaq0LCosJqwJJmckVmb63pvyhqsJtazP/GzO4tt8lqt+EeLofqyjdpnRTqzbgobKTYu0epq sEa2Nbu1D4uwuwaxoANtbYu3TGg7gxo7vRq4Znu2hpu2QTuw2AKdLXWYK3ixNgmuh0uxtVJbEYu2 2xqhiVu5b1tKWiotXPprj5uzgju4z+G1s+usrmqe8/muVkGzsBu5QIu6efskiBu6PKuxtNO4RSuA Apuqw9s46PqmHNOowpI3CLpsJJMTsFK3XNu1HJG9thuzp0u619stuqqEMSAHo8Ki4VY530uSpWqp 9Aaz4Ou8djO+kNO+iCK5/+oav9q7vajxbrSruihrvlXhbWp1nQicwAq8wAzcwA78wBAcwRIMmT0U ljiGcICEO0LqBxvcBx0cBh/ciatrOHNYwib8aXELkGCQT0pnFwx4tZ1wIwIwjTQjomxlV1LRPQNW DwoWEfRQY2wAnifQqSZgw8SHvD+QAfYkP27AgNezkfPTP3PwxNeTZdsTBiA3AuCTXE9aKtfzEWuQ pEx5Ci9XD96JAaEyjyIgAEdlr50wHyOAp3KchyJwxmn8KW1gHxOwADA8cEjsAwKWR3DwX0rctIXc FUwsB0Q3xeYDgBgnjwHET9wjKhYgP1h3jA5wjV6MKg2YMxBlCjoMJ/O4Ff+bFsUjUHH48wYWIMry A3FSnCqUcAeknMatTI3Js5cmO0p/nA03RkknrAKr2MK/nIAj/IPDfMzI3AK7nMzMLITGc33QHM3S PM3UXM3WfM3YbH3WWgLB183e/M3gHM7iPM7kXM7mfM7obM7NvM7s3M7u/M7wHM/yXJTZXM/2fM/4 nM/6vM/8DELZ188AHdACPdAEXdAGfUHZN88KfcIkrMIL/dDOXMytC9EU3X3LXNEYfUgXndEcvUUb LQsa0dEi7YICvM0N3QPFtqkjLYQqzQnc9WV6MEltJQAasAbpM0/K+wwtvdLsWL5yewIZUKWcPAcG HLCVeoU8LdI7LQgY2MT/1FPUosAiS53UEzjVgNDUQ11FD+iye0DVHW3Vf4DVk4SUcHByeXRPGTyk Xp3RYO0HF+2Baw0NWdgLqQi3SfjTvDq0a62S1lCELO3TDp0ci7vSmktIDRHCWdTWffDWg83RhS2I BWqEig2WEo2tXU3YVQWVHzjZvFnZc3vZI02aRGyAnP0FjG28GC3aiN1Epe0Fpx2jExzbsj3btK3a tX3buB3Brd0Fr22WBCg1fN0LBlHCql3XJZ3CJwoDbXcJ3plcySueSD0Nc72CdFmDxc2Fu80FY3il T2sjfAjdAYsa003dh9iC192E2b0FRikARPe0phGgLIhOwS3c+YkM810i/6s9jI452gWY3lowhkBc Yyua1nS73/xtivdN3vlNU/u94ETk31lQcOfTsgArql8a2YA43h3h4J/wk5IN2MntAjeGD/G0PgS+ B6J94Bys4Rgukg1u3Nvm2d6YtQRa3gjB4vVtih7+13atiCE+4xZumy2u4wbO4Ru74zqIBHxHdXC3 BMe0LwOM14L9un4YjqtU5Jn44th9X2vVsZwGJ8F0S0ww2r0d5DWek1d+4eTwhjhb5TkOgu2YsWVs D/9E0+EDCveTR2ywxwDgim6xgVwmAMb1T3keJ/6I3KAK5AWu5mju4owuBxlqjhx+3h+u37Y15zcC YLdkxdYTDywzAHKsB/+byT9QbtLGvLG/eaFW7uhCfqiRUiGTjuTCyOCXXoxGJ0wyEwCumEy5tD3C wDJAZQ/6CAAQMA9VSuYyntc0HhhlGZWv5OoeopAmSa5C3uhJbukR+8OVYK8fGyr36gDyM2MzLE/4 pD4aECoOcHJWlD5pYa8CNjPKnOxTDtvVmdvK2ZyRrgBbZe+k4IX8/u8AX5xWlmgsoEwkcAFfOXYp 8HAoYPB1DANlnq0pbuSFlhNVtbBOQOn2neC+QPEvAOFYUHD+pD7Mc2MF0HIVnvLLq+qa/QfyG7Y2 Cqmy/lHTveE6ddx3HdgM1a8DrmOdC4EuipVly5PjeqOP7tc8geM2/gf/IC8FEn5yPkfhPx+wEy+X YMsrMR8FGs9N911It7mVPQ6TO58qUp/qG7Lqo4i/WF/0Mq/lBnnm1V2BIJ7oeEj2NKPE8G32KYL2 RH71jpr1rDrzCBH0uJkElVwAByDEri3vKuAIN+N/Vvvci17t1r5S9yKZkEv0N3uslF+aad75nq+f iNblRnvu+GrajI9NVB6Kbw4NfRvfhaL2f8/2Wi/4zo7lm7VTd9mxP6WMkwQAGDDi94Tn9vSZUa7z qj+gzM73mNC3fouYFyG9cjqnbX/0Kv4FVZ/72rD7Rtr7o77KjLwzhpD6ULXskdHstnqr9oL5vav5 aMr5cI/0aW/9188F/pZJIXg5ygPkPM2zBBGP4iDQiCNZikCRqmsKaC8MPDNd2zWA6Pvuwj8wKBQC TEYSIKJcMptLgCEqnUZ9wys2qy0ejyjWCqAYk8tmsuymxlm1biC3a/qCw+87Pi9s6INJZ4TVBUBA QQDAhQPAwKFFAsABpAbKZEHfJQyBJgMnA5YmASZW21VDnBxSHUsbQIXrK2wsbA5PLQKpqB/qHGDv ExWwAW5u7umuaR2A2BkzWtraGu0w8YvxLp1qwTQ1NxZfN54i+Dhop+cV6Hi1m+lxanbLT6ssfSyt bY96jPvIny8glGBTtunbwu/EqmYKFTyDdkMZD4LFDiKDp60gxh/f/jIOseCAo6hynT5p0ifxRbuD 2FSxqufS1T18tzBaQ+XvX5OAAquAxFNTzhdlC5k1dGgjZr6CP7usTNaz4ManUnOJ5EQylDqJylLy awqm5ct6SGsBoGD2LNq0Z0/uU4kT4E4pbKUu9bJsaJmiRmlAtFVWLWC0yu7UNeJ11dRxURMzxlP1 3JB0WYe04erusJ0YYcXi+xs4MJe5kyjefKtE507RPQvzMv1ks6yxtzx/TrvVFGHSFi82pra4N3B0 m0YKx3pHAAAHGQ4htwDneRDLxzCrAAt7VsTaoEuIZs3dNRPUAlWD9I4E/Onr2HvQ1r6We26Vu8kH j17/fuThVovj/0kQyYEFADiXHHQESXfNfPKoh1177lFwG3xvmNcPeqfFxRN+1ZBW4TwLSuOgbV7E 15VFyph4Ioop0gdVTxg48KIDEujhEQZvSPBRSPpBJoRkbkwAQAbJBTggjihqcaBNCWq2IEwqOnki UCNeVmEgFwqToYZuodehek96WVh3uiVzF155XYnfb3p4CUOAHz2C4x0YWJDBG7xh8thVdwSQQIB8 CjgJnOscOSFCJSrIZAXkgSnldFSKF8yKNG24JaL0LSqhmGGUeQZqkaqTZh5czsImgQsAIIEEJlqi DXIBaBCAAM0pMuAAAqaqzKoAIOfpC3jy94YFwQLgJ4AEysMOof8VwWOdh6HaxOg1jlrJa1aTgicq bJY+i6l8LZC56aNn3geqT2KRKoAhAmSgjQW1anDRrBpYoM2Lf0ryroDuUmLnJb7mZ9wdFtSYgQUQ 1AhdFkgCVWKKiCbqbJTcdiXthdSOk2xpb2G7mbYRu4GxNpvmFS6G4xKz8cMvtKmBIg7wC29ykLQZ r7H87kutvzyWNNmgma7y7Rgoh6Xoth9be+20WI6m5bWVqlm0QSp96WTSaJ5srsoEtmxquvgycioK tdZL5Atcw6oBctrgrGOePCcMchhA5+Uw0R5HPTGHVWeIMYdOQ8wUtDblrXd95CbGb305B9EjOGwp zFR1ZQr9Ut3+gEs85eBxWdz40aZNjrVPUGfB95ZWilt4hjdiqTgQjHfjOMZyK/Q56JcbFjhQFGuu NOlNM9mx5UYz7RrJcu310OZ7KL06279ejGznGtP9t4i2R5t5arxHjxPtsT1t9+jb/0OyXsdLw+Ly +LH+g+vcwC6+L917Hzr4o8AfP+H39U683/QHfzfmkDaQ8hkFKcmzT/rus75M7Ox5PRue56bnv+oJ D2+lq5j2ICi9333vf+HTIPeMd7yjdAZ9CQzOAnvVwMaNwgWP80Lf+me9CFUwgALcHZb2F0EOUu92 M8zdlghYQJk0yD2DYcdTJHCIACxAAxIowAUwUYAmFiSFGmj/HzWGcUSU6HCDPMRdP8AIQ+yNJ4MW 5N8XxYibHzIlhr/zC4gEE0YkFuNJpGriIzzyp0tQoorN+9cdDAGBAAxgEAVQBAQQNolhvNAwbmzW BH1Yw0aREVJmtOEOIRnJOaixNb57ow6KqB0IneANhpMQZUj1ETllYFZBGoAibgSABQTgAPjSRgIw QIhIWMKPBDBH29ywgCayy16BmsQivdFF7klQjWECYQgxmMP7AUSGbCzlNR1JqUopI47vkWTCTpbK rAVqVm1q048CEMVJCEARBVhZH/VhRSxeIQO1msCQAFWNEz3wjJnUZDaf6U/P5a8+yxyfNSfJyWx6 Eo2/m1qK/kRXCnESQZUsE5A5hZScG12gEJWQ1wXg2Ut5/lFnAMtCqhKQNbIdq5+Y9CJAFUpDAFLy gjjcGzWdIL9R9XChMgXiJ2NKU5+Gsxjj1AAGDqGMALTyTwGAFUZRMCwNROIQyllqPNUxzxWOrhAp 0OUAIiGjlr4tpzlpZkA7+Z1KAuOAWTRreBI61Jl+cKDMlKv96jeEU54QJFs9qfugB834ofWnczTs GG06nhGyQVKD7cVOXwE8Cs4VqA4Val0p642+orCki+NqYF1a08t2qad0zatdo9lWxvLlfEqB62vS mFaGrjWopd3kaffKWeD81SSCTS1h8XqFSyFWm4oVBmv5/kIWxwIXssIdrkSh+9hqPpcI0VXebhvT W7cpE7bpqa4u9EoZ71oIucl9gGwgqiK1nmebmM0sOOF7Pdtep3KanWh2GbNdB5Z1ujotbGUPG+DE Xku9KSKiN81SBIG+9K7vRe19pdtc6sp2wGs8Un7169nWgTaLv20wQsELh+ta17//BWVERPkZCDF4 tP+sbwcjPF4Tx1XE8iAxEPiaYXJsmH0dpmh/J3ziCss3tzMW8pC7lJ0EW6PF8yVttmIcXwi7GKa3 daYpdzyV/bIwYY1s6IthjFskdIW8gajUbBL8IMNM6KDBfbCE2VxmGseWyFSeAz+0LBUuv84b/SBv ZCVr/9oTkMbMgU6UgaEkZ5UA2sY39kKh6fxdO8eZO9Nxq561wGc8ZAACMfLDHvAs6TM7OktISBah R01qOJfY0hRJNZJrzOpW//nVX64tlKO8SVLaJdMa/iVxAMlHlaJLkUBpdKmXBmtbu/nNV74cql0d 6zo/W2K8jvS0Jz3rEZ+ANVv0dWI2TZiPpOA500E2pccbbVFnW9tizs21mY3uan8s3q9uNoW3fWNv YxrceRC3hMg9UkH527cF5+7BQ5vwPi9cuz1mIGDH/a6BI7PhDLc4kDEuin4fnOMaLw4wnacHUylj Ah8/OcpTrvKVq+/hKow4y2Mu85nTvOa+DLmwba7znf7zvOc+v6LLgQ7zDAUIESaxRFZ9/QhlDKCp zgFCvGZedGU0MeopX/oi3IA4LVj958wD9n5ynkCSB4pOPzD7C9Bez2pQPO0wUHtfA2SJlkU9AwcT gtnRnoF1LnxlLQOC3QGf8JWZ6ulw14DZ+ZX3uyMeBlF/utcVGHR6ZsirAniBAAIg91cJ4EcHUMQE 3HWRi+hK7gEqJCVaFokACSBVbVfa5h1vrwQgkmYCCJDmCfGqAUAAAIkcvNaSk9EF5FERAliALos1 Vi2vDAaPuAAj8DUB5EB/APNahAQCRHwBPUICtcLAIyBgqum/PvJbnvyP8TMze1niAlGkhAUSEVV8 0f6fEvZH+tyjuvmkLy/2L4jXBajN2HzUzVyUAPDdwfmd8P3J9UWV2R1AO3kcRzTfC4yeC1jgxL3L ALCJ2lwfUjnHrHSU75lf4qDf0N0H1hHIvuBfJcyKOGDg/RFg/unR/pUf0aXeAgpLsQxgDI6eBPzI MYGbAo4N6LmgICiDSmWaAnogBo7eqvRSEQoIzFyUBSxAAWRAsZFgbwAcfugeO2kAyRFIJJTc1FlC VR2CDN4SDtJgq/Df8tSKMqjLrOjSIuRgGlqgMjCev01d1oUgVk2AOMAAu/iaPZnIBv4I1dHfRYxh uWnAIBACAEwAIsIfEppIEGrhU3Dhz+mK52Eiy9A5ACF5oigqjSb+nJyMIiqmoioSQymuoiu+IiyO YivGIi3Woi3S3Czeoi7uIi8eXC72IjAGozCmz1+BgjEeIzImozIuIzM2ozM+IzRGozROIzVWozVe IzZmozZuIzd2ozd+IziGoziOoxX9gDmcIzqmozquIzu2ozu+IzzGozzOIz3Woz3eIz7moz7uIz/2 oz/+I0AGpEAO5Dq+AUEeJEImpEIuJEM2pEM+JERGpERCpEFOpEVeJEZmpEZuJEd2pEda5DCGpEiO JEmWpEmeJEoCADs= ------=_NextPart_000_0000_01CAA5B5.DA51AD90 Content-Type: image/gif Content-Transfer-Encoding: base64 Content-Location: http://www.stsc.hill.af.mil/crosstalk/1999/04/subramaniam4.gif R0lGODlhQQLCAMQAAAAAAAwMDCAgICEhITAwMDIyMmkAaUBAQEJCQlVVVWBgYGVlZXV1dcZ17ICA gIGBgZCQkJWVlaqqqra2tr+/v8DAwMrKytLS0urq6vr6+v///wAAAAAAAAAAAAAAAAAAACH+FkdJ RiBTbWFydFNhdmVyIFZlcjEuMWEALAAAAABBAsIAQAX+oCaOZGmeaKqubOu+cCzPdG3feK7vfO// wGDNQBQaj8ikcslsOp/QKJJocFFPksNBoolsY1mubqIlZbWHRyqz+LIuB4Z0Tq/b7/g8ndroN1RX Jg8ADxIAAoQHABaDDxYABxqKFo9qjZKLAJBaF5Umj5GKIpolg5GgmARaXI2PAWiYFiSoXZpieri5 uru8vUp8foBFvsQ9DgAmDsVPmhDOz9DR0tPU0gAK2NjIyyfK3N86wH3CVeDmyccw3uc5zdXv8PDX 2fT12dvs6+z7VkR+fykC8TPiKZYGSAcjXTqoSZO+Gw8HomhIsaLFixgzarLHsaM9fOciwgCAoKRJ kyD+FalhQEjDgoYBMGAAEIAAgAgaDBGwyYWUITWHBgXQdKGEJkqaFB1ChFOoTUiaMGhApImAhgua qBpMAClAgAwHhh6wMHRoAg2NMiBUJLWEOIAoBErEBRKHyLkkAFTYy7ev37+AA/ud57Ewx7rg7rrQ CFKCIw0YHjzI0OXAgqIaJixIgFOEhAQLZKFV4/lB28+hBZmOzMWxzMeMLLDOaVpDBgZxKGdOMEEy bakXEjh6UDTyatwRKMcW8UjOibfk8OZCbEOxdL2Cs2vPTtiwd238rEvfBz3gsPF4qNcQPxf79vfw u383rH4Z+xwq0a5ddIKWoIqkkbDQCpOMEkku5cX+dR56dtQnAwD3jVTAhBRWaOGFGGZ4ISksuAff h4LJN59HDhIT4UQOpKiiiim1ZIpBpLQSY0sjvEiCKCJcImMkHBqUEGRQ3bKHP8GYVw6DDY60wokj aKLhk1BGCcAAVFZpZYmjgKhlYCKOeFh4SEqX4HMLhtnDIWjAcdNLneVlVHUTRSnnnBROaeWdeN6J 5UFb9tlXl17WsycvTJrJy5gmyGWoDobIMclQmdHo5g7ekHTSpZhmqummCNiZ56egUrmnh35qCWig 94BZA6l/3ggAS14swlUEiHTSEI8ALGATAS+dAslLC/g4Qn6PIDNJI1jVBMBZk6AiFk1QIOpWmYv+ OjGoC5Vyqu22mnoaqp6oClqqn6d6CQAF6Kar7rrstusuuw5d+0KhNHQiFQaUUHbBAxOIkIEFmGlA CWSyCYyZBckN3IluJGAQwQMR2BSsBBfgK1UGjgUswQT/YtZbv9ESOY6R1UohLwvZcqsyt95WWW64 4o675cv3vGvzzTg7tOLOKZ7cAr0l5yEtCYoGvYTPKqS88tLd8uz001BH3bPMM9eM89VYo6uz1FEj vSQS+d1gUZtN7kkLjrsMPULRRifhtQZdMy03phBybffdU1MN4ts01I33in7/zTXfJ9TXYoCocNjQ I4i0VbZoOUpKSuLIDCLa2dsMIsCBeKgtAtv+bR9RouAszm16SYGTrvoxeu/NROo8w7464ISH7ovn GoBu+6rIFJTX7HGfPrfswNvNaushvr518YDvvjvuujsvfQnFjoJMBk8FYKsmxA0FAAGUscR9sT3q t81RirzESEsYeB8TJqLIqIlZLvE3/f38QE8t/vyT/z1mShGAVF6EAaocSHwAEANXQDKIuhRIP2rI gAEpg7aFPCVSZ+GfBs2hvyNt8IMgDKEIhYA7EvzjhChMoQpXyMIWuvCFMIyhDGdIwxra8IY4zKEO d8jDHvrwh0AMohCHqMIRGvGISEyiEpfIxCY6cUgefKIUp0jF8XTwBIOwycMSsQgZuc97rKD/yfxu 0xCrME5SIlAEInBVvhehwn/7asgDlLKsq3gPABMgnxYa4pwq+vGPVhQZXMgURc984QwUE0ACcEOx OGgANxeAQ2sOEAGuyCE43+uXJPtjkwQAkHOGREObGOAV54ThKrvCDAMEEIAF6AY3W7DJAQImhXQA zXkAiIcud/mM8pnjloCEwRWDCZF0JO2DueSlMt/xstrRAZjEbMEwjfgTEVzgjv1CoAAYhgJj0gCa 50jmMscZDZoVxplzAKcL6HgIxz3IZ0cBAikeaAK0lQKNRvpHdI5YkAIhBCEDahIP1EmM4yEPMOYc ETpraQNLZepw+sGVZGqTk6H0q1kt4RBL/l6CiQVIBnLWQ0ooFuGJHcWCcROFIEMMosUCiQKlvrnE TBQhpLUJcp/RfBOlqmXQgw4GZh1ZaBQImlNC6pNkRZ2UDojai576lC8JnY9QocDUpNrUAEdVUCFz KtSq7sKpT61AVL8z1SdAk3je7KgFemUQRFDFERUBWeQiMRNjiVGljyDArBgCqUmo6RX7cacUpmnV t/kSnE6ik2Kf1LJRhZU7QP3SPs6qETPAphC/uQrEGmEcgT1mBIyoqWNA+1nN8CZHAXINc3zTnM7d FKlWPYgKfPm1Di32thjSxLeu1AKwPnWs3imrE7xqVcIm8REJwF4mekcji+x0FLiNbp12/kvdATj2 sVyK7EdUZQSl4ISeOgCvHSkiV3AYt6hdHYXwhlddUG3EXNjNrnbpIdwmnDV5wyJEBA6xXP2ItGwq RYghEoAK8RJrnl0kxCBCA7CTQiKvlAApE87L1YGqd71Ma1me3jtfbcQXoTA7V9ZGbDPZlZW4SaVw NNN7EAxn2LrAjaxvfXoqEZP4xutCK9ROHFseqJiY6NyZQ13MMuaRbsYHrW8KdPy35Rk5rX1bsqsQ 988AaMAmkUEIIoxyEwsgQjI0eczkqqyfAJgGc7WQgO/s8ONgygtvQyaytpj8ZCF/+C9KRhGc69y8 JbISF21WoiHmJ1jZlsDIcZbzpujM/+e83Rmqr4tdo4+R5x6/INBJHHOABGoO40impiXoDahRgNlZ TFQytFythFGwnNkYqtKWpgGmY62L6q20gUmJKCbk6B8D3dOB9rsERtVQwZa8MVf4pHVRZ63s2Gag 0M1e9mu1Gu1qW1vZJRwBEbfN7W57+9vgDre4x03ucpt73NdOt7rXLU0quPvd8I63vOdN73rb+974 zre+983vfvv73wAPuMAHTvCCG/zgCE+4whfu7n5sld0Qj/gRmS3xilvcdhRuoJUD2gKON7QuL9ov KAWUbFYj5OIoh3jGCSEAATTiWIQoMEld1JJJIPi/1NuP9e7pq5EixRIx1zlGe34Q7f5JYNQpT3px p01ILLbEe3NMMFx9XtJcXbm/wzYBViJhk52TXCEIUd8EaG4ahHSdK/8FxSOCBS07TPrtcI+73OdO 9xUpnWhMT9T+Kp5WFKNHnOQM/HthjQS/55TiyoYyCQzPD8AHfpyAIvwRGE9MxFtV8d3kn+Mfr0wO n5O7d7d8MDGPMlxyXvDhkrwRKB9M0bcNKqTxJ4/AvunFk14dQdv86XXp+RCD/p0ZmfIjS74YeNrP B7Stp3o8btQiUZufNJL9jzzO4lc/+k8dTtVkG7poV92KMjMCGE3s5WCgnO8AY4/V+mp/FJn/HILu l3mA7SqLTpCFPy/9Z+Va8pKZNv5/ZM9nRAUxdg3BBS8xRiZQfWaCZMgTY5+3favSffkFdFwHAGyQ YBGwdvUzAdXEIVhRc3gEK1z2X1lnUpOgFlaxZQNGgDCCDDYhC4pgK+ADLQuxQFqVVU3XYwoYJgzY Og5IIr+Xbg0BbbmTd9PycEBmYYvSg3rzg0EVhCnnek20g0jChFTjhJKVD3d3VTiod0joZkpofde3 F1i4XRC4hVLIRFTIIFYoM2UoKFD4AmgFUTaSAr32dVeRfH0zcq6FVc6Xg7G1hn83hmSYfeBxhjMw h66yOWvUgvA3I5uGa7n2QDoSdLiSX6IRdgDgf2xmhHj3hYBUO/hAVAAgXdKlW//W1VuEKFaGqACq JwRntWPC9yL+ZIFcAVf7F4knl181QnbItVJpZD+HQBk2eAdpuETGl3l9Y4rRpWGiooqE+Ib0FYdM oAg28T4mExV92IVHaGmjgi0Swoy35Yx4cl1jKI3ap4VbWIR+CICAWEUNVjiFgzTiUYriuFjkGCrm eH3oeIjqiIaeeFVNNBOFUBualoCjkAMPkVj3SCf5WF37+Gj96IrUaHHHOD02x1y1Z2hLdWGKNmft RV0ReWcT+YpBcF9c4n1WJl44YGAtMXYCwA4XaTtkkEm2kQABIABkw2kd2WIfCZIh6V4J1YbjUpIV 2SH4FYxe5nL9ZTn+AYk/whz/siQswagG1QNzYBYa+yFSENZ1TzCTSMRiifaTl/KQ3/KDRFkqJckY bNmWF+EDKIlQwuc9FoBlQwdglwBQuUKLxycCXEEAEjCVKnGLyGUIIwVmPCIA9xcy7ThIXuiNSjiW ZIkS7dV7IbaKWKgJOLaZ6RI4PLaOYHlEYjmZi6aPrXgPmBkoNsaZN+ZkfRaGoReQn7N3Sfhckkma 3mKZp+mKqXlOq8maI8ZoPQOX68iO3PiJkGmbpNktbtmcbtmb9AWcOCacTvOZANmYOFVYA9Uzy/lQ dWdn0aiZ0kli1Nk15Yk3DSVlE0gCGfAwXNAxtlEwFpMX/MFZspFHAiMa7Vlq/xYjARgAn3VJUpeT alB0nAKpg3YRO91JN98JOJgZacVznsaDToYjfMyhf7VAK4gwdn1kPR7FLyq1c5SjHyCDZhrwCixB hFEQmkZEj1Jzm5MpocyTluQCoYIjoy8qT+oZjN9HEyyRQS9hFYqQQfQJUgNCCmoRAD/qXxd6IGpU FJpQXp2InbClnShzowtalg3KIg96NBM6aSY5EBHwCgRaByw6QuoRoVl6EjgKPDTaJyZJPG3aNcUp a7LJjn50kDoFN3UGo2Q5p6vzpjNjo4D6NGGabme6O3qqVHUKDofKbolqOwi0iXhRLAijh/tFAGVa OCPXQBYwaBK2HCywHHfYqP8Td6fRY6pAYGuQiAp8uX7yF5W9WBFgkUCRQ2wzh6sgcwnyl5OqKkKR +qsqUD1qkYuuupUmeH6/VgKUCILvV2xWeXJPwU3Cyj/BWq1mskp9ia34c63c+q3gemmoSpvhWq7m 2gMdxHDquq7s2q7u+q7wGq/yOq/0Wq/vaqAmcG76uq/82q/++q8AG7ACq6/nWrAGe7AIm7AKu7AM m2L2+rAQG7ESO7EUW7EWe7H26nAYu7Ec27Ee+7EgG7IiW28O17Ame7KzCYoHirIsm7De2rIwu24v G7M0W20rV0fMpwI5OwPeQ3IUsZEisF/El3N8WLNGeyjjulW4FkfEYQHPJhv/HcMIFQMxtyq1pZEc 8Wl/JbCfEqAc8dSLRAcZknEvwCIVjnELxiEL7fk/XHu0bosgSet0P/pyUsc4VzcB+0WBBbRcF2AT 28MAIDUTVgEpwAi2AqOXD8ASQAcUigkpLLExdZQ4g1u0b1u5jImvePofDwAsXASrW1l1uFosYTFz HdqkmOB1hitzO+GLNDG6VIkKLBEASGe5tDthcau5H9i5yfp+wtZfC7NmposjvuRGW/kvA2SJVwGD iwAHcgAKkSF+VlG70usEK6cGC/QAXMFKyFuC80MU1ggtwCsCQ8ESMVm4kfOzObEUtnoJa3QItdBy N5GH31NKdVQHx8B64+Gc//rLGEcpcTP7R8MJN0rkDrvXeRSJiEr3v1QEZfjbeAW8TN3xqMZQnArs RAF8aGj6wJ0nIhIMmwl8uxJ3ewKMTLqnwdOgm/4YEhQMwusmwiPQwAVVwiZsDcHVvxFXwUbkwouH PzI8w9AQVR38XNeJuakaWzqcDNLTwz7sDCgMhAicdDi8DxUzAg5DHKSFtVfqNTCcBwS8xCeMljas ciwcQgdIGrCydScaEy8xu3BDOFtcB13sxTQsY2F8A2o0ukBbfO+0ajpwcyhgTySXx0dIxOSKP/8S ffzxT7TXDUFWMkq8xOgYxD3ZNxJYlSqVAXNEYCJABmlQI1rQGZNjxQ+DFv9aIKUMIRsFuQVkkFye VUmeNEdigAFtsABtAcs+URtscACa/ABfhgHBYRmUwVkgygiFFsXn0E+JPHshyqh2wVPQ2WGSrJDc 1y2LeI1FYRMX+HMEqQFo14HIgBUMgBVtIHUjmFexUE0yIqSLgIJTgQyNslIuRX+nO6Lsa2VeOch/ +JjQF3vJPH3JJoj5+8zzFc3NHIHUvJ4sIQcVIQsHKAZZxyEHwJRqIkfl/LmsS3UVMX8GMdH4x1wU oVKG8M5eSMgqez8F8RITgMbD+LgIKcQLKNDaRdAQMc0PNYsYSsxZcAFjFywPvQ2G8Aon9VmPE38X 7WCRMFp5Kc9+nH8Ksbj/4jsUZWrMnSbUpgUy7blIZQrQ1wHTdPzEI1HJmKAGImc+NPFsNIEIXACT QyELPRKl6TuEFY3Ri+t+u4YQuevHVBFPC+RlDREsCzHW/+eY3YigLs2DXA1UMg0nEEdGSCfVoaPV eCGoptKKib0ei80AC2DKXJjPgx2IHlyFh+17Xo1yjv16n82GoZ16dSyzY4xepz2I0UjZq61upW00 kN0eqY0qlf1NK0ylAWilk/zSsW2Iuz0DbzxFtZ17rx3Qw519xS0Dxy1Fye3Iy73VzQ3Ns81lwbee r9ofu3hPr6AV9yTIrpKJlGumrV1hhQ3a1z3Q2f07hmrTR10bcJDZjgAK/15QujZyyxNwFqLKyaSB avP9GkExzBRVoJyNnIQd3Ibd3jH93k0ii/lFET63UopbLBeoiy5ng+kjC5U4dQzxGA+kZfac2NPN UzsAIfu74ize4hqR26rp4jI+4xQRi/HN3Vt5a0GXiyRHAJQAQPjw4SOqlL4W0q9ijOm9Yu2ADEzF kA0pJ1Mykh9mlKO9GBJuyd3dqlBpuK7Si3ONoUQelV4BAJv6lUlem6uywzhgj08O5byFlA7e1f+Y iFce1rpWIFQxFLi4zLrW5beahzlpqz3yQD0L2FNK0sn5IN3QUG0+J6j45isg2Xsj21W+BGXA0lLA 2NuY4CsL3ItxTInY6P9u/ilSHl9UPudQkMuanOmYrdmDdeZg2FtYUihsLuoa8uikDo3nSOmoDsWw HoodMi/FZ+tQYpYuo+v8yOsqDJq//kcVGgPsUevEviFBWerYderLPsScnrJOtNDXVBXUypE8Cd1L Ju3TXifGvmHILpHK/ku9jehTuGmLypN6KOwJeO4Zku7ute4k2e6J8e7bnrlL5D9fMe8J2chNgu+5 pe/6yO9T7u/fgL/2py9O2wMLYwJVfOAc1OwfZLyxYPDiXtA0vuJB2fBwvuvEDeHWs908yizb2pLb yhVjIfMyyfEb1LvVE76juaaoU/ImH+kwbi4qzycpmV9+61eIoKRXEQf/bSAJcfA9Y4pHIiA+ByAV LhmtyABJp/QAYiEVkMS8meFWfJwEJ/4NdaUUGPCXa93SPemnP8nwIvmccY7YQ99TLTJglkMKLIEU X7FzaYEQlUDXwneVUvcTaoHRVW92Elz2rxaZPN/zPg8uChX0ClX3SYkJEcC51eOUuwiVCFGrYkGV dk746yeJiO8/i2/zVLTzPA/35Zh6lC9Vll/0VZm7mz9znOP5tCf46/nTsdAIwuYouA+7530EjL+A jv/4ri8qTijpHxLJvzmeOWOdiVhZ66kUl1oVh5v7PA74WZHjJPDtZYTJ3asGTyH1QjuVCFS6S3D8 PJj8rV/tAx37JCL9/sHJIsS5KBnAyhvv2++YVCAAaCNZmifqjADSui8cyzMMDDee54DS+z8wGARU isYjMqlcInnCJxBAmVKr1is2ewUAHF6vCCUem1TkMzqtXrPb7rfaIG/QG2S5Aa7f8/v9sJ+GGQtN oWGhjY6OE1QjFBFTpGQko+MPl1am5mbXlydgoGDoKGmp6Rpend0Y3qnrKyxo3+BhrW1LoiKXJa8Q 5CRwcKUj5qbx8VSn57KsnxksdLQ0Wmrdndx0trZa8x7tLThNLs5wr7nCb7A6U7kv8rum8vJ8N9/z Nn7+aDXddZ4+wG314HwLZ/BFrnbnzKVb5/CIwh5S4FHcIm8evVL3/gJy7EhtjjVW2DySPDXwTcGD B3ctbBnlIUyIviZWrMkFI85PGkvy7KmB3yoxrXwS3XPSTUqV4S7mbOoUY8OY6yoVq2nzadOjKKMx OCChz4GwYRdkOBE2zYUDDEZ0LVoCqD+3cttoZZNU6S2mWPfmjCpVmESrVvXy/VK3zUa6XBYzBnUA wIM/ACxoIAD5hAXKaDJYuDDi8dyfIPuJ/Bf69JjDagzjXVn4dd+/MBvTrm37Nm7YOAnrNqwaxUDH l0lYtryWiwYJABZoeKx5xWQNDCA/AGA5MvLK1gGsrS5gsgUABzQw/t4ZQIKecEujbg/cGbPWBnn3 xupXtqTffujz/+XfW78YwZXwWGQjhDcecgtM9lgA5BFgAm0SVLcWdBocSJ4I1RV4YXMiXMBddROo N1pQKAzlHorkebObfEvVV999+LGjjX9Z1fgagGQISAKBJFyIXHjTKajgVyVw8ZwGGpIApHgYJhmd hU12OEICARzQ4IgGqBJXiijmKEhWLYJz44v0yDgVjf+VCdWXOqY24HKZXfBjGJaJ8B2WRkI5gpIV 0vmkZhw+VlZ4w/G0nlAjddmefnsRImYtZK5p2JnqtAmHpA7cNKlOJuFWQldirSWBZQlgMIIEB0SW apEDHuAZCazyOF5ypZ6aKqxpUWhBWJ49dmqWW7K3KKNrFPYopP+GZDppjJU2kaajnH5yKbF6RPBd ej4haqKi1c6llW7IJiuOtP05Cwy1bvi3rH3eumvKtiec+G5RA70o7rgysFtms+cWkS5dfe2bFb0F 70Milwbnc8EDEQALYQr85qtsuY76mx+0zFQMhsId9xGvCfN6LI1lF0Rg6JIlSIvvxAht/FS//gLM hjybljvzyAaD/Fa3OUPDZIFGjlAxyy3j8rJTMZ+LMzdgbJydz1G3sTMJIkvtaZQF4rY11117/TXY YYs9Ntllm312bVerzQbVI1i99ihAn8A03HXvZzfe8iI8bN6hfJfByUG/W+hiFJ7woZRoJM5nY3mS 0OcZSnLYN+X+UrctmmmV9xFBAgvAWnChFRpoQVlP0soZkhhY8PDiT4ZxJAkXlE4C6oebMLnmuSt8 +du66x66kyertVwGChKAQQLcPSYBBgAIUChn1iFZnYdHMviqktNFMF1k1z/5AODWffggaL6b3yXv PZ+fO/DIVXeAguO93+ECBF445wq0Ps6FWCI6x3j3wHOZXwEwSvKbDHrWp0DUpC9zC6xc+zKEsvn9 z0BNklD+TEA9E1RQSc4JTwD3FxkObQdJDzzhofaWKAeiEG+FupLzkmOd5HWnSQpKgGUm8KEAVAdL XDCc6zgIJSVNKHkRmFIBw8ND6cSwhU4kSQOfqEAMfO5wD9PYAGdmF4gLXDENE0CZFMO4jSiKsYwk yQAXHmTGNUqDjGx8IxzjqDMVcouFcrwjHvNYFDfqsY9+/GM++AjIQRKykKFoIB4SqchFMrKRjnwk JCMpyUlSspKWvCQmM6nJTXKyk578JChDKcpRkrKUpqSj3k6pylWyspWufCUsYynLWdKylrAUFhpU octd8rKXvvwlMIMpzGESs5jGPCYyk6nMZTKzmc58JjSjKc1pUrOa1uylGq6pzW1ys5ve/CY4wynO cZKznOI0JDrTqc51srOd7nwnPOMpz9MAADs= ------=_NextPart_000_0000_01CAA5B5.DA51AD90 Content-Type: image/gif Content-Transfer-Encoding: base64 Content-Location: http://www.stsc.hill.af.mil/crosstalk/1999/04/subramaniam5.gif R0lGODlhCAITAdUAAAAAAAgICAwMDBgYGCAgICgoKDAwMDg4OGkAaUBAQEtLS0xMTFVVVWFhYWVl ZW1tbXV1dX19fcZ17ICAgIGBgYmJiZWVlZaWlqGhoaqqqrKysra2tr6+vsDAwMrKytLS0t7e3urq 6vb29vr6+v///wAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAACH+FkdJRiBTbWFydFNhdmVyIFZlcjEu MWEALAAAAAAIAhMBQAb+QJJwSCwaj8ikcslsOp/QqHRKrVqv2Kx2y+16v12EGEwum8/otHrNbrvf RjHCKYfb7+0JSY/v+/+ASACDgUJyEogSSnWFjY5DE3xJko+VlpeCAB2bAEqUf4eJi2OYpWafUaim q6xlmpuwnVSqeKGIo3Otuk60WAC9u8HCQoOwxh2yXsBwtopJjMOsy1XJR9N3ABfa29zd2oTRgcXH nHDXbc244YXnUuBS7a7e8/T19e/rWYOvsdXsjemekcqnJhIYf1riWclmr6HDhxAv4Fv3qiLCVgrT BEQCjeCWjE4mMhkUqeQEfxeFgFQCQIFLBYMiypxJ02FKOOOQ3fS40sz+xiMdPfIis1OQyaNIkyZN 1pPYy6dQXV6MWbOq1XlUuxWdkhOZ0FkAxSRyxnHgV5VNWbpTyrat27eRUraMSndrSIZXr+LN+42c MX52z6YKi2CsukYiEgAwQMIDAAdCHADwQGIfyS+BiZCEy7mz585zX2b2srdmVr5WRQr+SNiwwFxJ ACRoLJsEA9kAGIRQnCCEAAEGAFgYIYCAZAsZCACggHkhFdkJoktXPATl5+vYs8e90tWvd05/JaIe T1P16idpwfyMY1ZY0dFHoE9XbP659vtuN2Pv/t2vZcvulEbegA3Bx1NrorzmhAcUUJBBFR58AAUA BAxBQAHKNSCCEZ3+GMjhfNJ52AR8+uF31C8masdffyy2eIyABMYo4Iou1mjjjSzWB4pYCZYFmxKK UUYIBZMR6YFjFCCp5HKVzXaQO/Pp6FwX/4lWoklXpqglZzTiaONpMd4T5jZdemkmJ1IqgWRjFITQ JAkb5FYcnIsJMFsI+wgwgmO/MdnEekUEFcxO8ok4EmlglknOf2hu6Sh+KJ6k6JnhpTamjJN6l2Yw gBIhKB6ODVKhEAEUIEQBAXBIzKbuYKYVP5TG2s+jJkZKa1v73JgNmJf2ahOrHnU6xKe6GKrZQdzA Kuuyy9qaorO3RlsStNf9Z62v98D3gQD7PCiUsIa0N+gTxsbGkLL/zKarLo4lUitttO6++xmvBQLL RJCVdWIkkvu2aaSRl4BLArGF0DcIc0VUYy+5ma7r8MMQv5ilvCdRDGl/iS583hUCEwzICMGJumHC WvwS8ckop6yyxM7Ga/FbDfezsR8di7tLZiiurPPOPPfss6waH6FYnBAs92+RAHxw8G0z08HjLQpi wgG3ucWX8AQ/Z6311lynvE/TnD5NFlA2N1LBPgSMfKwf5VL5R9tusw12PjX/OPfdeOettx11txIh CUFakFtwtCXpJxiOOUkCkQ40GELigMONBJENLpYA5LcRORuRZwTnAeGLI82g5Cz1BgC3H1AXpwGS UWbAbRaAsaeb/oRQR93RlO1tSd+69+7778BPwXvwxBdvvO/DH6/8FdENEp2ES1BHhGIGKLeV9LEp vjzxyW/vfYBCELnBvqJHPsL0k6E/wu3JULek4RDgeXn53+/dff34569/IHXL4f//AAygAAdIwAIa 8IAITKACF8jABjrwgRCMoAQnSMEKWvCCGMwgAV2zvw568IMgDKEIR/hBj5HwhChM4f1SaA4jpIeF xysGPl7ohRXC8AsKoeENmwYrYOmQCzbcYSp0+EMh3gxdQbMGgqDmIyM2oYjVKQIU5fEqJ8YHXWhK wxSxEEQUGqQ5QxHHgJKouxWRLgtb5JjYDjPCLUoujSyBEbaw/nLGUmiih3U8Axyr0MXipTGPlSHC H+eILTK+YRx3BCQe9kiFPt7ti4iKgsuUchMahoYuViKkJiXChq4o8hKMFN4aoza3P0pyko8y0CUx KZUJyXGTYpoJFl3UQ+WFUgqObMMHAhCABPCSMhgAQAEKAAAMLMGNJELly+CySlZ+sjq+eiWBEPnB W0ahj9NB2BEYQABvKQEC2kPCA+YDASE8gJcPsFrJuAKi5i3TYs9MWJek+Q16wtIbhgSbNaHQx9pA jjofoAzl3hShNSXgQdiL212o0U7pjUaZ74RZvKz1ombViF731Evw9vmEXL4BBLkbkaoW0tBPTmRi 8oJodrqD/9Jb2TOjMlFpROH1nx0VpkdksxsS8BUdEvjmN4x5H58GEarahGRUSmPMLk1FMoZGSZFn bGl+ZuqZmKnrpbDE6K+6dlFtMSABDNhA+Lw5hRAkyWk3ZWJOn3Ck8w3hA24Swgjc6obAyCehXMjn Fc8llRzJlKr3sWqstMoXrPZKsOwi3ZoAVpuj+dQDG5gTwMI3mTX9aZRNDAdCoAPIzpbGWrME2l8B i6XRlhax/rEJTMfEn6/NzKN1HRWeGCCE28Q1ilTqbLJCy9W/mFZFv52pZTRl2NVCRK/hgK0lRARV MvVWZzKVKmmrOl1mFlcihsTXkB4jmQwwLjgjoFzrAobZtf5q1lyY2dVz10vLfVSXme/NT3mQ64TJ Jre87NFpIxxDAAoER5u4pVIk2Etgrro3sPFli2DpezzlVsJQJiuwhCfMM9BaJrCJ9RB9NmDZ/Dn4 EYTCGoVHTOISDzaPQcrAwQQHAdApxgGhui3y8BuosgWCSN5SDocGbOIe+/jHErNiGGjsKRsL+chI vtuH/xCkEFiPOY4JjgBC+gXIDYE+kOGvcsoJBs0JR2m8gRO3BCBWyoWzC/IDgFhDJxnGmNkVCSDA bOJMHZANAjLso3J6I8dnIgWHMUm+JpGHZWRHjMC7GXBrBjzwuLkyenZl2BP0hBCCDCSaBJL+wHDI 8Dg3Bf8U038bgQVC8OnHmaHToJYQkQ7tJlOb4UiPdROs4cRoRhMDMpy2AAXWPIINfKDUHhhBowPN z0GHS7/ETrayCbLkZTv72UscW36hnTdw4M4xEIgTZfBKvZDxmX1rdt9y3tckbGub2o9oNrrXzW41 qLvd8I73kNMq7SSM5d74zre+983vfvv73wAPuMAHTvCCG/zgCE+4whfO8IY7/OEQj7jA5U3xig9D gxjPuMY3zvGOe/zjIA+5yEduQLRa/OQoV4+xB1bolLv85Ux4t8UhyVGYF+KOKok2G22uRCTUnOeH TGTP+bdyE578HD8HuhlyYpekL0HmxIYkFJyudH0oy0D/VH8NB81b8RdmveoMWxQWvp7TrU8b3lAk O9jlqakuqL3G9N65s92Y81a49tmzvDsOdU7KQEt9C6p5+xLoyeDlhbbwT+R7ZoWMTCYIXjPXraII zRjPxBcC6tuj+2DYdtjKR6O1noeH4rlOwr8D/gpUjzx5EF8sZPi2FI8/ttnhLkJTsmYNqp8j62+u k1gEI/Ysj3vf8df43O7hDLk3LnaHgcgsrgP4mD+P6dd5yvqIZIr7UL6MlmuR3f8B+kVved5sPyGY OT6vmCSs9qsSenP95ZGjP7vvyH8XSLnQF6yESjLUv/7j1nPptaQ74Cd8i/dIxUcuKdUTzZR+I5F8 9/RS/5tlYTQSgBsVf7TXNJonSdNlFwtYF9TggDKSF7zlW76XeRZYZMj2fE+igQm2FFWSf1YigTI4 gyAoSzV4D3/RDzS4gzzYgz74g0AIWgM4eyhYCN51PiAAAAswW7UFADImSCtYfi04L/nXfpUxTTBl hQdyeeGXgnAAAgIQABpABKVyKqnSVICnSiEyhbWihfJ0R+XRf8miP0OIU/InCLMRJ7RlBB0GJFDl Cx5yV9Oxf8HFhtuBKCO4KFghhw/hhtJwgoTmhQRVG4pBATo2UBSiOYu1bX/YKs9RUsQweIY4L1hy F6jVD99xg8bliJZQh2p1h3tFVKHDLZAxUIrBLQ/yG//K8SCSoSekERIkBSKyYCyF+DLRFQsW1VWW wYim8TuuWG9FeF6CwB3t5H35UozP4iiJ6HxUwozloTfPKHd9YFbENCok4AAC0CAOMlL5Qo3VSBSD h40S9U7bGDH8541kcoo9Zo1c1IVNcBsZoDTmKElnxhLmGAIgIATKcT415Y5RchDEKI/yqI0/c4+r 2Ij1WGB6Vwm5hC9DcG3jFpKGwxxGtSBUxiCjpk5Q8pB7RiU6AVETuUz6eCaqKIKrl5EV5iGOARm3 ISQFyVaH83T+yAS3wQBFMxsqRgG3EVC5IRkjuSanQ1Z5NSEs+YuIKBFScXWREpMJNpPsQkgWeZM9 o1f/ffgmjhVlsEYkFtA65ANguECEkUgRIxEinZhXp4FY0jWKoIGTlFKT4uGNXulbdaQ0dsIttDMb guMABCAAJMBiBFAhq6MclUYhyuGWWmeHF/h5eEgfYJRbr0JcuaIrekldEhWYEhNTfplV28iPNDOU jRBnRHBOAZBO7NgknnWV+ViRozktNHWKvJKazAhaYBN9mDEqOxkZ6QN5neUqu6WRXLkfbBiaGAOc csiau+OaN2eOwTRMxVSbvsCczgVkOpGAu1maMoiP9cKKkHhscjkl3Rie4tkieYkrzylc7zWDpiEi 2lUNdOVT/YlppUCc2HBK4MmX8ZmKLlOf9rmbkYdc//sZOkdiNIuhHKNToeRFgKTnHofCnAZ6oDR5 YOXpguV5XRuZBXGyD+VEJFQDV1SzOfTDkdipod5ZMurloc45n/CkoFvCdPa0KyMkoK4QAWoRSIho o0aqgzTFoF5iLScEpGAAUm4lTDPKHSJ2pFaqKyCKKyOaYU7kpIdEpVV6pWIKMTJYXXy5MAFFGZCm P15aV2sxpnAapzKYWFbAGwaAJNviPCQAHCRQNB8QJ5apN23aBhwYpnF6qIiKpaRDHdUjoRbQVmlG AYrBZb8zqJ0UHzyWqJq6qQiqS0+oPJa6BXEipJUBaGsTYZyaqqqqEykXqlqgAQBQAaTCVGvTB+qp kv94cKu4OqAw56pr96s35KvAOqxNGqOr4JFvwCfSMWlCYGVlQCSDcD4h85NpwJl2sA/MAa262o5K J6xfECQMYmeFU5ZdUFQHIzr/9EmUI2c4VhuWJj8QagZGUjUQijtLlwDzKqmdwCAh0DqOsQEJYKpE 0Qm2oy8vCm/eCngE4AEUqhghQDjkRgbOKjhrWVkU4gHcQgaUgyfCATmLAa+cUwZJpa9lJhw9GbKu QLC1sV0e0JN08kmOEVa1kzSEY6/ylrDEOgUo2whFYwD/mbNbgLNAO7T4I7REe7TLY7RIu7TcY6xM +7RCprRQi17hYwAG4C2WSACxAwHcoj0POgKKmQD/EpIBCbABkAFOQgCwBoAwlig4m/OYsTO1ooSh sCi31yoLfvY8PhUcuMZtySkEvNFTKCtuT8kktfGwj2G3uOS0itsGDTIEZusA0BO5uRNeAOZdT2gB Rnk+DDIE3iUEH+AAYZW2MLaOlNu4xUa3mYm6rPujjNu6sOthXUhytFu7tnu7uJu7uru7vCtAcGkE Ehe8wju8xFu8xnu8yJu8yru8zOtvsfu80Bu90ju91Fu91nu9LNS72ru93Nu93vu94NtxaBW+5Fu+ 5nu+6Ju+IGdy2Nu+umN07Om+8os3Uju/9osOr0usIbq//Nu/87Kewbe0fAB82HthAAy/PPcJBEy9 /zIECQcsfi83fQsMvRVxf1youtH4q8AwwaxLTT73wJKIctfAwXYLGEURjsMXwe1AwkzrSWFEdBgc lzaXESxMtC48dSAcwVVQw8PKUh6CwgVIcdOHw/f7gUG2wzncdQlRxO7gH2iUxGjndkw8Ilh0RkCc odA2xKgXhFzcxV78xWAcxmI8xjwImmT8C1CcxU/CwxQ3T2RwxXUbddjHFKyQFdtaRnlXeXC8ulEX hQ5sR/eAZNL5ImuwxxmcZBlowQ+GVXd8FhPoiIYsw0emxdxxTA/Gfij0yHYQyfFrRYnsCY2gio1c LKDnB5wcwEZ0gJbHNtQ5ypcwyNxoymmcQpTsnv9EnKu6Z3it9QinjMB06Meid7cPGDywbJ1jN8sh pMpgcUja58qdNMjObMkX/LuonMyfbAU1R50iCDYyJDMYgcy/3JK3hwZhmYWC0c2x/M3TjJmHnD/0 B3iScsaghZ7ZIs/23IO995L3vM9CCM7LU8vU0JlwVM6AOShowqoE0csQPH8Q2Zl1Rxr0nJ92hIzR PM6AUL+Y8M4jUopTenxUIhURzX4VXRk6uDEKHcLgeM0joWAdshM/tIDavIqNjEf65M8CCI+uxCUu DXgwGNOrZczx8Xrwt86vyMc3DcwbvR8X8UIdGBUdEtLzpQY0DY42rU/KvNLPUg3p0dT6lzBQrVr/ c3gQQn3UMEzNvvwVGh2PLtUJTcHVT0EoX60V9qAxM+jNzljVgpHWokgxQePWokEuUB2WFIWldl2B RA2NkjwzAP0cJEJVwOLXreQO+Jh7C1bYfoTX+aDSam2mhAKDf80d60fQ9JCIJvw9Jy19DZ3T0blZ nh3ZNJpRwAl66WyChy2OfRC6CfAAk6YBC7AAY+h4ms0SIUoIkA0TA6t7hQQeCB3OZc3Oie0HD6Bm dAIZAzAAQlDdoOzQIeG/ceHWoSfaGKmaIHTaj5A5x3mOfwsJc3xK3I0Ug+CBzzwePm0T1YTZbvAB A5C4JJCEDSAEDQAACVmrU1l97c3Srh107MeI/yPNy/bNBpIxANJRTkXjAJJBqQGmD+w0iDrq2I7o wxEx36nhznh9OgnALXFbBPwllUJTl5W8ksKYLwVOSfGUKVoF4vJt2iM+G5AzAgwgAAzwa8FBAL9W 4g5wPmpb5FxbkgMe0Bn+4qEI4zHu3pIDy8qIF+CtfAveB+SdPbQxZyobOiQ5G7KBJCPAJHh1elzB XKDIrfIU5aelSjOJRFfef1neQrWdwgmj45QoH2BeGXaSABDQh2dOfRPiVE7OgRtOnqXFIYM9WPIZ 19lCPFsuCHdKJOXEqBWCiQlQ5jC2HErDHIrxsxgOjIau4WzOMPyrTD4sWl8J6VgBPJN+BOpIAf+T ZgH4yrlnNQLg5C0fAAEM4C2Wq91XFIwPCTc46tjV4pU+LIEc4up01DuxPi6x4ZCmjigtKJFnymA2 flh7E+03M+wfWI0mVR2kdex7iaDPNOehXed7d+dBDAgY8pGWcWYoQe2DiNOYKpM5SuUqs+2X0qHi CdSLG8Od/AepoxjmmAAHMKRtI4jNA1V29S6JztJa4+9jlJtwWqIXnb+QaxkWTpUTYo4AEAARcBv0 3vDijtTbvaMTr9MVGdc1DvDsJfBW0JHpAw69VrmPtvMh4Gu0YQDMOnijIgK2RgIHAABq8+SlHiIq zxWQMoqmuaT0XFz8TmE0348EX81+SGWCMz7+Flu42Wa41Bof5igZgxAAekakTc70Tb8Qf2Xu5S7z H1qdhSX3/b4wvoEbTvgmEBKUb+ncBZ8E+JKW6CqS5KbkCkWVVZn4gIcMogEzMW73Uv/TsS3565JE i3XzTrIBENAmQpABRhlXHgABEAA9awmoaGXW4vegIFm4I8n3wj54i2+XpIEMWNlXnODebl5aFZZV uhf1S1ouZdlYkyEZEGr84nM0G4D6l5X18PsB1jNlmBYykJH35i2SfYr4aD4i1W6VdslXMIHuLe/Y OwPiFv+Zdw83a8KvZok0lCUAqmYZFUquPqL6KI0JHHjvsc8dd9kwcO+/QAAAdIhF4xGZVBL+ARfn ExqVTp1C6hWb1V6ES++3IySNyWXzGe0BQEgOgIcESJAobwiAjk/g655+fwOAAg0NwVACUYJwzBBh 8REyUnLSDO9RLmGPcjPOkhPSCqoLLElo4hQ1VXWVtdX1FTYWVWiI1FYpdOsqV7fXF2v0ttTzEzKD AIAgYyyBbcRBwGCDZIQhGo4EgkCZmoHAImFZsjFRcbGxOF19krgyU3OdsD2ejLcKjJZ2yVS23//f Hz9V+YQJa/KrCkKFC3fVskWLXkSJhQ4leoRuYsZP8+Jk4khPjEZ7XIzkC1awiECAK1nGArBSH8p9 ug4ytHkzykkkITX2pEfO4jlDPomCcsf/M+PHeDWfxJRZUGVLqQBfTh2oEyXTnFpxdu0aDGlRsZyA Iro4dOynER7cECBzx4EbNpXqwRPpk6vDp3uZVLX6l1VUwIGd3mLK1Wtir2HTNo5U1hwhjI45JRsD AkCDMQ0AgKDLOCLopVr18jVdcvBfwan7EcQ3UnHshaIp1z4D+awj25Qsk1DjYIwbOCaVriuujvRp 5V5Ws3bl1zlg17gQy7be8Pju3biF6tYeqTeJAQPGjD+TvTLew8vZk4IePdV7+PMLp4R93Trt7/u5 S0a7/xEKLChDgwUW0AAN9CTRDzlRSmsPwn3kk27C+eBrbhbi8sHvJgbPWMuDEAAUqz+K/7wbEaQU e8oFqwhdZK5CmCycMT4aJ9BwQ9kg4sQNBywID8WMSrztvyCNQ07Bypp68MUm3YvxFShtpHBKqe4D Jskz9hiODzmSGWEEA4QYpA4xCRjByNzKyS3NpYrJkreDWnSSzoekPOXOKqfKU8/AEPIwki07ycMD NSj44w1EsWmTSATW7I5R5DaBk7cqmKwTU6gI6/NCPjmN75cd6RFUjD4MrcOBQ0NQNNJGH/XvxFZ5 W5DSBS3NFNflTPH004B6/ac6LDViQJBABMiDgjsOBSADYkdgVVYyhjRjsmjZMWrFJubMlVu+aPnV SnBbC5YKbYlalYJpCM0AmxAsECcED/9GiBdNa0mYtoxq7cX2PKK0vbTbgNv7VtznCnbpSilEdayP fSnBV9oiHV5knlrBq0/gjDEl+Fdee2XyyoUnbhViRiSOlgADDqVg0XpcznZbjWXONJ8pPeY0Zvu0 5WJkh0u+9+RWQwDgAAEEiCBBQDeaYOamnS6JY9UOhgVg6izuWbuf9Y32AxFIGNqAMUxakemnzT4b 6psxnLpq6rDGWuugI/3xARIwAKBuuiSqCu2+/T4i6oGmjhKqq98GMO5YrdWAggzq1Xupsv+enHKd Ax8cz8INPxzxisyClHPI35S88tJN/wLHy1nLeSdK9wAgXkFC/y7x2cuole/Tdd994NT+oWy7dXVe h8DQrzNQ14MPvoZjLRFtl6j25+OojHTerb++b5E/2aOOQJa1gNgQNNljg4al/8nzyEw8X8GXsH8f /qeVDhSAEbxclhYPAvmRAAcMCPt86HNUUGDFPlCcIn4JVCC35jeJLblBEH3IQAKUJwAASEMI4gjg OqL3PKW4b4EhFKGLtBeRLQ1NEGEa0xjqMCAAHGuD8eigwzzgmQSdp3oj1OEOZdLAGPosfWxy2Hg4 cIcBQS53PFTiEl/zQyfOMFo/EgcBYHi7OOSQiVnMYgmdKD0oyqoCzBoDAQIgNt+dEY1pVOMa2dhG N74RjnGU4xzpuMYuPjGIoLOXCAL+EDYNrKFfRtrcXQTJqEFO5JB3lNUXo+UHDDxOkZGU5CQpqSYC rq+SmdTkJjkZKUZ2EpShFOUoZZjHApISlalU5SrzZUpMHk5Q+1FDPlrmGzmkpQ4vHAOxbrmfO8wB QPmzpQ/Vw8qefdJhsQwBJKmRFjUAkwxgGsMzvyaWOlhwOBaEpvLMwMwVWdB50eymWLwEHDfYZZna 0Y83jdm5AX7ulLBMFLPEZEsG/GgQRVEDAVaWAQAwYA8f2Kc/E4mGOtQBAoH4JQkSIABq1oEAcyGK GpTFQkEQyw/JkCheEkCsOAAUD2oAHwA+UIcQuCEtYhhflyiA0XYGCZn7EtQxxGD5qOLp84KFyqgB kFGoW9qlJwe1hnBu+Sxt5oGdGXHDCKJh0YxmNKmITID3ALABeGwAGW+IAwUEAJyxqFQPXXrqS1EU U1n99A17GIEmbCo7nELTVBSI3RyAqpGDBmIN1LxgR/KQ0nwoT1F1SOkcLIiHUs0TDnfQqk+KB1aG itV8ZOWPKxv1ttft9UcvrF9b8znRXn6tsGFTQ2GhGVTZCcGncyCWBcMmWLF4jwT4zIOYBLCqglIM D3AZFEHFxLy9EoWptGBDLi1IqMhKNmuUpZbckNvcM7hWVmwBwBGdC0qzVhe72dUuZa67Xe9+F7wc VG4rFRde854XvY+h/ywQSEIhojg8IpPKJbPpfEKj0im1ar1is9ott+v9gsNiZREhOUua5TG77X7D 4/I5vW6/J8voNHON/wMGCg4SFhoeHumhqRkhOj5CRkpOUoopnjEiVG5ydnp+gsZd8i35hZ6iViWs sn48QSQgwbJmvMY2JUCk7k6OZvICBysBAEglFB8de5AQcwHcCkcL+vY1Sl+jNpN4ABAQAISEPB+H ZHxTJAMsN1MAGAgQkLQLxB6TjLi3Z4gLMDyLrwKHbSAcaqWsEUxYSVs7B8pInBtiL52BVRtIHKPw TZ5AjMW4UQAJklmsiApPhjFIBiHKlocYqgvhYUQICh/8uZooUV06Eu0flrXbuQ1ASKIj/9nE6XIp FpV5WDKNikfbBwACrNLsRkDACBIOsO5cNsSC1WcchZIQAE8dvrIJxHnjKnUuFKdITNHNq3cv35Vm FlXT1Hcw4cIt7SaCangx48afEA/B63gy5cqHIBNRbHkz585uMJOQ7Hk06dJNi+zJVGY169auX8OO LXs27dq2b+POrXs3796+fwMPLnw48eLGY6eGsmc58+bOn0OPLn069erWr2PPrn079+7ev4MPL348 +fLWo5hPr349+/bu38OPL38+feno6+PPr38///7+/wMInWkDEliggQcimKAgQQAAAAAAOw== ------=_NextPart_000_0000_01CAA5B5.DA51AD90 Content-Type: image/gif Content-Transfer-Encoding: base64 Content-Location: http://www.stsc.hill.af.mil/crosstalk/1999/04/end.gif R0lGODlhDwAPALMAAAAAAAkACRoAGisAK00ATW8Ab4AAgJERkaIiosRExOZm5vd39/+I//////// /////yH+FkdJRiBTbWFydFNhdmVyIFZlcjEuMWEALAAAAAAPAA8AQARA0MjJ6rwgX1lGoJUlhctm ZCgQEGYSvuaSmAazMFOAcqhOZx/a5BaiESsI2EQROlyUhqBxIkAVJoUUxpqSXnTADQA7 ------=_NextPart_000_0000_01CAA5B5.DA51AD90 Content-Type: image/gif Content-Transfer-Encoding: base64 Content-Location: http://www.stsc.hill.af.mil/crosstalk/1999/04/subramaniam.gif R0lGODlhiwCtAPcAAAAAAAEBAQICAgMDAwQEBAUFBQYGBgcHBwgICAkJCQoKCgsLCwwMDA0NDQ4O Dg8PDxAQEBERERISEhMTExQUFBUVFRYWFhcXFxgYGBkZGRoaGhsbGxwcHB0dHR4eHh8fHyAgICEh ISIiIiMjIyQkJCUlJSYmJicnJygoKCkpKSoqKisrKywsLC0tLS4uLi8vLzAwMDExMTIyMjMzMzQ0 NDU1NTY2Njc3Nzg4ODk5OTo6Ojs7Ozw8PD09PT4+Pj8/P2kAaUBAQEFBQUJCQkNDQ0REREVFRUZG RkdHR0hISElJSUpKSktLS0xMTE1NTU5OTk9PT1BQUFFRUVJSUlNTU1RUVFVVVVZWVldXV1hYWFlZ WVpaWltbW1xcXF1dXV5eXl9fX2BgYGFhYWJiYmNjY2RkZGVlZWZmZmdnZ2hoaGlpaWpqamtra2xs bG1tbW5ubm9vb3BwcHFxcXJycnNzc3R0dHV1dXZ2dnd3d3h4eHl5eXp6ent7e3x8fH19fX5+fn9/ f8Z17ICAgIGBgYKCgoODg4SEhIWFhYaGhoeHh4iIiImJiYqKiouLi4yMjI2NjY6Ojo+Pj5CQkJGR kZKSkpOTk5SUlJWVlZaWlpeXl5iYmJmZmZqampubm5ycnJ2dnZ6enp+fn6CgoKGhoaKioqOjo6Sk pKWlpaampqenp6ioqKmpqaqqqqurq6ysrK2tra6urq+vr7CwsLGxsbKysrOzs7S0tLW1tba2tre3 t7i4uLm5ubq6uru7u7y8vL29vb6+vr+/v8DAwMHBwcLCwsPDw8TExMXFxcbGxsfHx8jIyMnJycrK ysvLy8zMzM3Nzc7Ozs/Pz9DQ0NHR0dLS0tPT09TU1NXV1dbW1tfX19jY2NnZ2dra2tvb29zc3N3d 3d7e3t/f3+Dg4OHh4eLi4uPj4+Tk5OXl5ebm5ufn5+jo6Onp6erq6uvr6+zs7O3t7e7u7u/v7/Dw 8PHx8fLy8vPz8/T09PX19fb29vf39/j4+P///wAAAAAAAAAAAAAAACH+FkdJRiBTbWFydFNhdmVy IFZlcjEuMWEALAAAAACLAK0AQAj+APcJHEiwoMGDCBMqXMiwocOHEBECmcisosWLGDNq3Mixo8eP IEOGnAjEIUlnzpqpXLlSpEuMzZQNSwUMmjNotja1KnaJEjCVL4MKZUYykNFACklGW8oUmtOnUJ9J nUrV2TOUWLFC4yTHhwgPGRxAyCAhAYICDCwkwCDgQgICCRjA1TDBgoULGjZU0BDCwwgq0S42Gxq0 6NGkE6kpXsyY2rTHkCFLm0yZ8jJAMUZcyJBBBIgMEEBo0HABQ4YLHjyUqMGkCpYrYcKAmeIldpgv W64wCcJixIcPCxpcuNCAwgYRScyAKZMomDPCHg0bRQwEm3Xr17Bd2859u7Xv4MP+WyO2RQcKECxS gOiwwUSNHDuCLHlCZQuYMGzm4LlDJw+hRIgsssgjjzTyCCOKHFKIIYcI0scdcsihRhlbYNEEEzaE sMEGLRjhxRSSTEPViFmVmBJLKBI10VFIJUQSN9vEKOOM2tRo443Z5JjNM3MkoQMMMwxhwwxFNPHE FV6Y4UYcc8BxRx57DIKIgZZ0kskmoYRCCiqqsOKKK62scgonnmxyySSLJKiHHWycgQYZXnChBRU8 sBBDEbA4Nk1lfFLG1J9LQeWUdC1KNNE3iH7jzaKMNtqNN91EKqmk3pDSBRE76BDEE1iAIYYaa6zh BRNLNDEFFVqQkUaEdtghBx7/fhySCCSUgKKKK7HAsoomjjSShx5zuAGHG2p8kUUXXFTxhBNPaPEF GFXsco144VVj7bXYNtbYNIRSN4444IYbbjjklmtuOOCkmy43sCCSBqdijEFGHHkoIkkkh/Dhhhdn sEHHHHMAQggihBiSyCB42DHHHYM0Yskkf5jBhRhu6FGIIo78wUce+01Yxx6EEKJHH8rcWKOOKOd4 3cosd3dNty5OZM7MM5dj8804l0POzjzzPM444bRiiiB2jDFGGW98TAgjk0yCySaifCLKrbLcYrUt rRQihx165NEIJ67oEssnh7jRxhqLiOKKLLbIooonlTyyCCGA3LEwMJBOOik3/nz37feMgMeoDcyG ApHO4Ygnjs7ijDd+zuOQn2NOKIAUIgkcasiLBh1/FIIIg4lQcoknoXzySSioaGIJI4TwIcjFkViC CCWW8BEHG2iY0QYefCRi8CGEKIKIIHXEEccblpyLrrrMp5vo8883uijhB5HUzvXsZK/99uys4/33 4K+jzvjkq7OOLoTQ8cYbc+hBiKzCH2LIIpDw4e8aZYjxhRtyNHIlJ5CoRCUY8Qg/uEEMWKBCF65g hjrwQQ93cFMc2lAIbuisZxgkx882yEFxeXBc1DMISeBBwhKS8B3weIcKV8hCd7jwhTC8XjhEYTsx lKEMYVjDGdTgpjOIQQxg/kADG+rwqz8I7GKMiIQjJHEJSwiQEoswRB/sAIc2HPAKXGgDMNLRuC4u LnJgpJkYxZgzm4WwICShhxrpMY82uvGN8oijHOcoj3jEI47jMMc5FrcOcVDCd4LwQx/qoAb9hcEM aVDDG+AAhwjJoYoLU9gb+DeHOryBDGKIExa40AVetIN72wufKMtHSsUl7nCLOyNBSFIPe7iyHq10 pSxnWY95vONw7HiHO1CYQl2SMB7zoAcs6eFKYdIDHtnQBSgekQhACKwPcogDHcymwzK4IQ25+8IV tLCFLnQBDG4gxTnakY1roCOXLEynCmHIzna443rwjOcnuafKgbCyHu0A/5o52CGPYNJDHiSMYzCD WY9/Zg+FKpxHK2Npj1je46H4wMc98JGPilI0oviwhzyuUQpNxK0QG8vDHUZaQTu6Ax3dQMk42NEO c3hDUS+FKaS20Y1w2Awd7YBHHO3I03iY8Ke9rKdASMJOd9zxjm40JiyXWtA1NnWNUK3HRDEa0XxQ 1KoPhehFqVqPY+BBD2uYwhOeAAZs/GIZLwVHOr7nzuy1Ix3liFw5vgWundnMHOgwR03RIQ5ttMOf bHzjPIS6D5Kw9J2f3CU87qjGVj7UHlmNqCuzOst7NPSpwwxsPa6aD31UtKL68GxG/2mObATDDVNQ BCTikIZlpKMdKjzs9f7codM2ysMd2VvHOQ6njryOY4/oOEc5wCEOc+hMHT5Fx1rHYY13uJKwJHmH +d7RT6gKs6nziMc7GJvZpcryoQtl4xoBul2CGrON8KBHRGP5T2K8IQ9nuEMTNkEGN4CCEH74wh7C cQ51oBOp85AHL336jnWkY3ze6+05NEgO4JojHO7wJzyyB1foTsSfcUyh92C7S56u8bv2MKYak3pd WF6WqeP1aS93+Y7rrQMd1FgFFiZRBSIQYhOHMFAh3jAGKDQhCI0gh3/ZgY644nJ77mzHOh78DXGA 4xvgIMeTbUqOb4QDHe4AqJK5YWGSePnLYA6zmMdM5jKb+cxoTjOYWf4UkTa7+c1wjrOc50znOtv5 zm0mCXT2zOc+97nLI6qKibCCIpZ8pBnPaIUhoqALIWRgAxegwAQIsAADcMABD1BBDUqAgQVMYAMY sABpYGADQbihCLEYjJ9d0uXH9IlPgAKUoKDSDE3IAQcc6MBnKBCXCTQAAxCYwAQSEIAANOAAAfA1 AypwgQdUQAENWAABCICACBjgABI4zQZK4IMdkMAItlBGMFS96ox0mVrfwZa6q6EtxVRjF2TAgQhC TQELVEAC9v50C3JQBCQwwQlWiIIUtnDDNLiBf3SAIJT0UIc6zOENbGADHMyghScYQQc10MwEIKCB EuyACFYoAy2Scf+VQhu6MCs6TMyAYKOUocw62WDZdbaTjUIgYQcuMIEOhAADFwgBCUuYTxUmpqo2 bK0PiWAEJTBxiUw4XROa+AQoPpEJTWSCEveChNwA0Qc+3MENawiDF7wQhSXcIAQjcEERkiAKach6 1oIKtKC7rLdI+e3ufAscTSkxhR/sAAczCEISptCFMKgBDnGoAx3iYIc8+EERldgEJzpBilOoQlem SMUrZKELXuzCF7/wxS5uAYtSlOISijCQIwqRhzaQoQxpIEMYtFCEHbwBG9Rgd7unoZjIRObVk+ly uZqnLugZH6bfGIYcplCEHwTBCl8QAxvigIc8AIwOjd/DHvBAh4T+tWEMq6IiHOywNEygohSOAAQd 1LCcNMgraaLqghWyoAUsYOELakgDG1yhDZm77P/Tgm7pZi1dlkEcdIB09UHhog2i4AmZYAc+5AVq gAeCwAiSwERk0gmgAAqeoAlMZ3VRYwmKoDF+gAiScCWYwAh0sDmCoAiRAAmREAn3MgmSQAmUEAnN NFLGkHd6RyMmgyMulw1d9kVgBDljdITGdTPnUAuIoAZqQAZfEAZn8AYJ5weFkHSYwAmmQAuhxwu3 sAqoYAqX4AZ+oAdccARSAAeRkAmVAAhTMAVi0AVuMAmcIAqpQAusQAqYgC8PMkF9QA3G91LSIz11 tzd902WjhGD/pDQ+p3RKXQQOlIAHbKAGddCCjkAgvZIICpIgjVAIfaAHfpAIUdQqs3IJjLAHbmAG cVIGbXAHeNB1g/Aridd9i6cH07BBCiguyjN8xKcuXaZLRRVD8iRPoLQ9wbAHD4cGb7A1e/AHZXgH dkAH0oh9fUAIgmAwipAghfAHfQAIfjAIfZAHfFCNn/hwT2gGmQQGekANkoOENVNGN5NBPXOAXUZH ddRT+AhUJcRLGlZc2uAIfvB1xIIGYJAkQsQGbZAGaWAGRmMGb2AGdLAHFgM2HegJZGIJUjQHZHAG YLAFbLAMCbaIjNiIiuNFjFOEkdNlJ8ZUS2Vd1vVG8YAO36BH/+nAUii0U+3QC6BwCZCQCH+AB3BA BpgERF/QBVqwOY+kBujoBd7UBV6gBUnQA0IgBUywBZNADu4UQ4g1jPNUjLklSuBTPio5SycGD4sD Wz4FWE41TK9UYrHkUFMlURNVS+FAC5YgCZMQCcPjB3zAfcdjBq93BgZnB7qQU+wgDtNwOOPgDcVF DjPzOLmkj/uoTroEjMHYZeMjOUqIDusAW8FURwL1TzrVS/EAVQQ1UVpVVZ/1WZy1mpI1D+owDI5g CKEQDuwADwHmDuEQDcrQDNRgDmv1OItzYOognJAzV+RgM1LmPOKQDndkj/fIU/UYYP7kXQ31XZYl VZM1WXNZUP9PZV1StZquSVXZeQ/4BApskAheMAiGkAdUsASmIA79tVsGFlfnhD3Z047CmVfiEA7i UA54lVfg0A17NA7uUA7yaQ7vwJJdplOieUcptFgRuljVCVly+Vgolllr2ZIjJg/20FmeFVr4kF3t oA7jEAyHQAljcAROMAh2sAVJwArlsFbmsw42CUy2lEIupD3gw0W+ZVN7NFzsEEzusE8vVg7v0GUB JmAGllPyMGLCdJ2Q5UoD1UZrBFiWJUvYJZowpEI7tVgCpV3p0ApTYAhOgAWWcAVIEARYoAg01gVH sAkxtw3hoA6wtVgEFqG6BFd6JFz92Z+OuTjiMA45hV7uoA7/4qCkT8pG2nVC2uNO5UVMUzpMGrqW jbWWailHKdQ9vRVc5VAKZGAGmaAG3ZYHiOAIUDAIakAEPBAEFccI1bAOeApQJORC8yQ+NooO4/AN 3BApMYUo5KIN/LVL7JAO40BYBMEiyrqszNqszvqs0Bqt0jqt1AqteHat2Jqt2rqt3Nqtd6Zm4Bqu 4jqu5IpmJkER5Zau6rquXQYUzLAM60oY8Bqv5dZlg4YSzXAiJrcR8woSzeALfAAFafACRaBrLEAC uZYCLyAHRgACL8ADRnADN0AGoXALxoAV9PoRXQZ30CB3UnGv+GpyQOEMtoAMZFACG2AXGdABedEA EPAAFoAB/xvgAQ2gAR3wASeAAXgBAhxAARhQARlgAiGgAnfwHBm7EV3mJ7H2JxzrFB57E8QQDYKQ AhxQAQtQAAcgHAuQABMwAAYAABEgAAzwFgAwbQZgAAiAAckWAQ7QACHwARLwAR5wAjUQBbAADKTQ Bk4gCLlAbvHaZXrie5IBfJOxtNEwDa7ABkdQAhoAAQugABHQABPAAVw7FhBQHAlAARXwASTAAkNg BC8wAy5gAy8gAiegIRJwbKE2AQvgARsAAipgBZnQCa0wC8fgt+raZetmLbnXbr0nuI+RDFnwAp8R s6ORAR5QFxZgHCCQAjeABE0ABUgiBmawBm1gRVRIBxBSSf9ysAZocAWvIQQroAKawQAMELkfMAIn kAIqMASEgAspka5d9n8CCB67ay3YEAlC0AIcoAFfoQEySwIocAI0cANHgARRUAVWgAU51Aat8op7 AAidoyDguCZ/AI13oHhtUAZeQAVQMAQ3UAIegAEfEAI9EAVMEAZ+YBNZka8hK7LuqhFdpiMy1zIA yB3YsAlMcAMm8AElIAIvoAIvQAM4QARN4ARTkAVe8AXVuzV/gAiM4AiW8AiYgAkCdAlWHAkEwkyM wAgHE452AHFPeQVNgAQw0AEZMAI14AOCQAxw9wwdK3dWAbIq0WWC84MtF4QxB3PFcAZGUAM1YAM6 IAMyoAP/a8cEVqAFXcCK16QfIDMlklAJlmAJnBAKpYAKYXgKpCAKp2AKosAJD9MIiFAIe2AHcaCU YMAFVLAsP7ACL4ADSWAJe1K4bme4ThENHEsVXYZ3PNiD26ANv2wj2AAIUOADOFAEOCADQuAEUNAF XzAGjMQ+c5AHgkAIjhA7nPAJpXAKpUAKnkwKrbA2tGALs/AKrJAKpgAKnywJjGAIf7AHcbAGiYRJ WKAENbADUBAMiwG8rka4tQwoXTaIeVOIdod33RAJUzAEOyADPFAEzfLMaNBIc3AGz5wGdCAI8gMI wJMHifCCqTAKrDALvEAMyLAMyOALtSALDljKbXAGYyAG/2ggBlOQBELABIrsBVlwBKGADfd7Lb77 u8DbZcV3fNEj0IxCDGrQBH4XBFHQBYi0jHFgBobnBnPQB4bwCJmwgZtwl4zQIHywB4YACZlgyZ5g ig9iN3LAB4bQCI7ACIjgB4QUBk7pBWBgBodQTjdcv9bQ07rXZeCyi7zYi88DDs4ACXCwBUugBIVn BnDQjIBQB2ZwBVBwBEYgBVdQAzrAAz3QCZGQBnQ9CHMwjlPCCH4gBx35BlRABo2QBDvwAyiABoqw BDQgA0owBg4kTccQhDU8czccgODRZQj4M7k4Lru4DIwgCGjwfYfkQPMjP13M1jX4NJygzqCQCZPg CIvwB/8jYwiRgCWUEDFuYAdIVyCLgCCHMAiVkwgL8tV9wAw+iMcno8d7zDJdBo/yuDPBjYvgAAzJ oAt7gE1oEN5/kAhUnAmcrAqpcMmkYHqvYAuwYAqSEI56sAeL4AiawAmawAh8gAZjsAeNIHmgMAqj 8ESYMAmMsI11YAd4w8t/04PwbSNdZoTuCKDwGI87Yw6w0AdjMAc4VAZroB9+EDwWWCXgnAvAEAy4 YAuvoAqdYAd/sAjD4wik4AvLIAyt8AjRZDyXMAqvcAu8gAti2AjojQfGYwfIIFOESNDdwOJ9EyNd hkomGVwoGTlidA6tQAdr8AZr8CysSER90AeFkERbzgr/tVALD64KqrAJk3AIi9c77ZwHb1AGuNMF Y+AGfIAxTSN1noAmhuAHD9cGhLANxEfUiWLUeTPQmCmS6kCSJelF2gALtSAMngBWiGQHGs0IjZAI GIPrkOAwmvAIilAI16gHg/AfpogIs1sJfTAHreKKeVAIDaIHeLAfeWAHxRMHtJCAgE0u4LA8o/5k xtdloASWYanqjagOufAH0yQHYE1A9OMIisAgh8AIcpA/WiAFUwAGZ7AGt2MGkzgvVZQGcZgESsAF b8B1+zEsAMMGpXBBGpTfw/3X295lL8SV8eSVPCpK7KALgBDGbkAHcIAHxV7sfMAHUrKN0FgHsBLB fvAH/4wQNaMgCYKABxJSRWrA4WlgvZgDmGmQB9DwjjV+3+Pw8BAPLr9ImSsUjLZq8djTDryAim+C SKHiBtfOBmbA2MzOnoigCFEcyVjnCJBwgU3Tk+pdyhPkBVaAJJNwZXMuOe2IhDWunBjUZfhoR5I5 mZS5Sy6UDuBgCnDd0mZwBmFAvWswIWGASbGhBlS9ByrPO4Ug1pdAJqDQCYsO6cRSSFtgBp9QDqvO 6qjko44j522PhEoqWNApR3YUnT21jzeDDtzwCIKwB9AoBxy+yGigf7cDKmggmGqQeHiACEoEPB0t PHqQikypBWDQCZ9E7iFp7p4f54zTZRzqklcqWLYlR/8B9p/oYD7Z4w7rMAuUAOyD4AfwjENwYkM+ Lgd1kOVmgztjINdMyQVbQH/2twRcEAxM35XjTu7mo+rkczgAAURgIIKB9h1EmFAgEHsN7dVz+LDe RIoV6cl7p07dunLo2H109w6ePHol6c2Dp63WKEqJBum5E6dMGTRmyJThUqXIjRUmRoggkSJGjyVW uHAZE0dZO3Xw3j19507qVKru2l3FmvXjVq7rvH5dt7CgwYRlF0aEGLFhvXnvmK57J2/evHh1S7Kd O+8uxYb37uFbG29brUyMDC1SdMgQIEGA8NS5Q6eNTS9bwIg5EydTunfnnmEbh04dOnPo2tWF5xTq aqj+VatmhS22YFmzAh1OnNfOHLl18eTKi6fX5F157J4Gp8cX4sS+f/E9f54P3z178radwhTJkSJE hfrQkSOHDZs0cf48cwev3Tlr0NLBUzeOG7dt38id85ru3Dl17uqiTi1AAVVj7Sl3ZCOINoUEEscc 4+r6zbe55DLJt4tECnBCvSB6zi+/osMnHxFHJFE6fOp5h5lLFFnMjzzqKIWdkuJZRxxplnnGnHY4 EiecccIJ5xtwhhTyG2+++SYccswZp5ze5AHuPyn/CxBBshTcZyH+2gnpqXbY+Uojr9jpUkKMkBvu LuegYxPENkcMcToU9UCCCCnsIKW0cJhRBhtrxln+R52PuJxqx9HQOaccRRcthxxFwfFmnG/MyWtC KC+Vx0osD1oouLnS5MuhD6mzxy+0QOUrxBLhnK7D6f6i7pxQ1BCjiSJ+IAKKT5RpsKPR0hHUrXai OmfJcs4hjRxywgHH0Y7SMccbcEYTB5xx2IknzZI03XQhYY0bqaS1Sh3VQ1GhI9Uik4QrSdVVSXyO rXeaMUQNOEpRw4w2CCknnWHd+Wi0dYyLCx538kMnYXUS9fEcYNEpB1Jxzhln2HPMSYcccOaRiFss vYVKwJAtpGctD01FizmHNlR3HuDuyUcfmWfWJ5973vkI2GbiQGQRPkYBpQ1cPNpxUC7lykvYLzf+ UieddBRFR79yxDkWHWXhoQcej9hhtp16PFZwoXXg+XRC3y6NJ0AKkyOVurQkouikk6BkeaKSXLYn ZpkhugclcwSBA5A/ZCHEEkvMYGIIOr5Blsz04Enb5bRTc2tMr5pGtNH9EA3nnGxzM22ddMYBm7aF UHIHnbHZ3dDteu6GUtu9UpZ7LqtColAvuki+251eUOGEj0k8uQQPLj4hIg05kPCDmm2cbAdykuSh PMAdzXE4nczJEWcc7EkDBx2y5zI4vtJrA6Ik4HBOZ51hJZy91LVeX3e4tCp6/XXJVfvyy5D8P5g0 6EALTjABCZJYwxWWUIQ21GAOP4iCGJjRKEj+OalLBsKQW9DBrHIwSUjgsI85mNQNHQUnHu9YBzfO tyAgCId8aStaVERCob6sJTn0Q9UNlWM32KVNaZc7RzFAUQpY9GELXmDEDvwwiB+gQQc9MEIaxmAK dWRQKmAaGBbTwY4dnQMc3ThSksSBJB99ET+pcYc6yrFChCyEJCahnlOcshWAxWVD8mvZSfInO7nB 8R0wZEfTkGW1oNUBEnqAwhC8oIgpSEILUEADGvTQBzzMoRPo8I8L2xIVeHDxI15JFDi4cQ1uhBFJ X8zGOT45OjZyaiGvhGUsZTlLWtbSlrfEZS51KcuxbKosYwFmMIU5TGIW05jHRGYylWlMXzb/05nP hGY0pTlNalbTmtfEZja1uU1udtOb3wRn2HY5TnKW05znxCU00blOdrbTnbBUp0CYMU961tOe98Rn PvW5T37205///OdC4gkEgBbUoAdFaEIDOpDZdEueCoVoRCU6UXq2MksCaUZGM0rRgzbDGcxYxjKY 4VFnlPSjHEXpPC26EJOWVKMvhWlKl6ELSjhCC10YBCX+oIQlfEETXUgCFBZBi2LQ4hYlnadIU2rQ lQrkGU91xjNa2lKPwtSqBW3GMmCxDCWMAAUi0AAHIlABB2BAAiaYgQowcAELfGADIPBACFTwAzDQ gQ2RSMZSFwqEXjoUCNAA7FMFO1ipTpWq/lbV6D2XsYtOQEEGIWDCCzgAAg5oIAMP4MAFRFABDXQg BB7wgAYaoIENVOABEsgAB2iQghlg4qR63WdTgRANwNbWtoEl7GANe1iNRrURS5hBBhgAgQ9owAJr hcAAGiABBWRgAg5wwAcwYIEHVGAEHbCABTLQgQ6YAAUjsAI0YMtP2UbDvOdF721t+wxo5Fa3JZVq L7hggg5UoAENuMAFJrABDRygAAL4AAEQEAADFIAABAhABBpgAQZY4AIceMAJSBCBEXggBUqgBDA8 ugxa5KIZ460oQxPkV2mU2MQlRm+Kzave2g4WGsCwxSSCgAINYEC0BYBAaiNwAQosIAAF/jgAAgqw gAMooAEDWIAFHICAAzCgAAiQAAUcIIEOlGAERugCE2Jwgz40QxkfBrFspzFmMk/jxGdG8XmhkeLa RmMYXmABB0xrgbBSoAEU+EAHOLAAAjCAAAd4gAgmMIITfIACIRBBB4z7gARIgMkCYEAJIPAADGwg BjFIwR2cgSMQq1TEVxInEKgx6lGPmRplLjOaT3xeaSBDDDSQbghAwAAE5DgDGWjAARKgAAhUAAMc OEENcHCDHzxhCkFdQhSc4AQdBMEFH/jAdilQAQMsoAEgKIEMgrCEKnyhELTYNGxlWw1yk5vU5yY1 qtVdYmo4wgcnMG4GLsBZDUyAAhEw/0ADKpCBELiAB0VQAhWyUAY1sKENZegCGdrwBjSswQxlCMMW nBCEI6hgA/ptQGozIIIcVIEJTYjCInTxWpTK1hont0a5VV5udKO7zJQYwgk8UIELbGCtGYhAASig gRGkoAZFWIIUrrCFI45hDWpwgxvigIc8MN0OlYyDHB5ehi9MgQg6eHYGLCCBBEBgAyywwRS4EIlL QAOxEZXtNdSOcrajfOVvH7U1UrGEF3QgAxswgQcuS+cSpGAHQQD5E6ighSuIgQ1xoMMc5EAHPOzB D334gx/2oIc65OEObTh8G8aQBSg04Qc0kDkE9nsCHdBgBlCQBTQMi9iNgpm8n/blQv7UPnvaz77t bF/5NN7AAxRsgAQ2mMEF1loCGNRgB0NIAhOocEQyKK8OT++DIQpxiEEUIhGLYASL/gAIPexh8nZw AxvI4AUpPIEJOihBjTXQAyo4QQ+ImEZUpVrY3TqD9Rutp2yxsX/+7/8a2Kg92rOGa2C7a0CFJ7gB EggBEhCBD+AAEViBHdCBH2CCKMiCLuCCMAgDNpCDPegDQTgER3AESKiESrgESXgE7GOE6zMERjgE QsADmJADWumCRyKCGwCB1HoBJ9gCSpiGFnMvwao/l4Ip2cqGI0TCI+y/JfS/2rOGRkiCHCgBuFoB npABHUCCJajAK/iCLxgDNGiDOv/Yg+lrhEmIBEnIBE2wBErABE3YhEuYhEpYhEZwBEZ4BEQgBD54 Ojg4gy/IgipQgh4YgQfEASegglGQBtpiMdwKwt1qBtnShkiUxEhMwkpUwiW0hjpAght4ARwYghtQ gRUQgiw0RCsgAzRQAzWYAzzgA0JIBEaghEqwhEr4BFIohVJAhVQYhU/ghEz4BF6shEdwhEIghD2w gzdYAzEYAyroPCV4gRRoAR04gkqgBhVjs0VsL/eSrW3gRm7Uhm2YxHC0RCQEBjQwAhygARzgARfY gR9YgiWAgi3QgjBQnjl4gzvIAz1oQRLEhEvghE4oBVLIxVRYBVQYhVIYBVH4BE3/uIQ6XARAuIM5 cIMzKAMw6IIqeAIliIEZmAEmkAVqULU0UzFsrC3Zmo+T7MaU9EZwDEdJvIY7cIIfqIEcMIIYwIEj aIIq4AIwWAPEa4OlywNBQARFgIRL6IROCAVO2ASETAVWYIVXiIVZcAVVKIVQ+IRMwIRJcME/0AM5 AIMxGIM0+AIukAIicIEaIIJQqAZ1M7OQFElrhAbZ6oa5PMm6rEuVTMlv5IZHmAIfuAEdKL4jeAIs SLg0cAM6SDw7yIM9QARHqIRMAAVQIIVVWAVWWIVUaEpYoAVb0AVbmIVXeAVYSIVTyIRJkARFKAQ+ oIM3YAM0KAMxAIMo4IEdEINq/2i5U2PLMXPLRDQv2fKGuQTO4KRLuyTObeiGSpgCIdiBGrABH+C2 LggDpJsD8ICDNrADQnCES8gEStiEStgEUdCEyzSFVoiFXRiGYlCGYwAGXtCFVTAFS6g+Q+ADQ9gD NwCDKugCZQyDLmiCHQAEbLBNlbvNdMvNtiwx2TISb1DQBV1Q4RTOuvSGS7gCIsiBGiAKsQODMjgD N3iDMFCCJpACnkRMNqgDDs0DOBiEQdAETSAFp8QFXugFXVCFVSCFRviOmjADLxjLKoACJBgCIViC LBiDMqACVyBA3Hs72xzQcyszBEWSJ0USBpXSBg1Ob6iFMkiCHcABIXiCnaRHN/9oAzD4AjDggjR4 AznAA0CYQ0VoBNQsBEM4BEFQBEvQBFBQBVUIBUggBEHIg/DoAz8ABEDYgzuwAzaIuC3ogkQlAzVg BgG8PSRN0tuUrSGhVCKB0kudUgX9BmRYgyYYAh/wASlIOHshVDdYgzZYRT8oBEZAw0yohEY4hD/I gzkQw0FYBEvwhFAojD+gAzhQgzWgAz1YIj/QAz2MgzQgU2X0AjUIBW1owgC0vUd1u7eTLSCxViCp 1Gy11EsFB2DYAzBYgiAIAizwAjFoAzzogz2ogzhYgzQ4g4jLAimggic4Ai/AQMOzgzjgA0F4BE2Q BGNMgy44gzHggrFEgh6ggSb/cII62QEimIIxqAM4gINdcNZsYMIlhNYjlVbZEoeO7dhrBVls1VZK rQZRqIMxMAIm8IIwGIMz4IM/YBE7SIMoIIIkcINEEAJ1FINY2IQ/iAMsqIM6gAlCUIRIMIQ5CAMt 0IIteIRXKAMj2IEdcDwyoAIj8II5CFQ7uINl0AYkxAZLvNj+y9iTk61xMFuz9di09diQtVYi4YVE 2IMuyJcvOIM66ANFcIQURE3H6w4RvIRKiEVKoAREcBE7uM5HyIRMaIQ6yFA5WAQSvIRIsMNFKASt FdY91YM/WAaWbMmu7VqwtdiLnT3ZUpZxIIezRV20/Vi1DYceaV1peIRHEITJ/4BNN9CDRFAEYYwE SrgE7cSETdiETwiFpMSE0mSEQdiDP0AE7uwESyAE1swDV+wOOD2ERVgER4iESXiEQyBWPUgGvPTG zqXEcUzC/ZOtRlGW9FXf003ds1VbcfAGXvAFWzCENCCDM2ADOwAEorSETiAFUSCFXQSFT/CEAEYF VBCFSCiESfKDRYgES7gES3AEP2i4PWgE313R3m1D7WWRPMiDYegGuwTfvBTf8T3C82WUFHaU9VXf 9gUHX5gFSLiDMzhFN6gDVb3gTADgUzgFUiCFO22FWriFWkgFR8ADxsMD7BQFVCCFSMiDNDDXSRiF VIgFpySFTaAESFiEQ0hNw//thd8MTuK8yxEOX0mULewRoTRO4xRmFBYu3VNYhDaQg9ckgzWoAzwQ hERIhEOIxU0IhVSQBVzYBV3QBVmAhVPAAzsohD8QnlPoBWswhlV4hDmIAzfwA01gBVrghV2ghU+A hBUUBD2wg8UbhgRVUAcNYzEmY9naj1Z2ZTVWYzZmFHOgBUd4gzgQA+jMDDpI3kIgykrQhFGYBV4I BmDABVlghVSwhDvw5UQA1EcQhdDkhG8tgzRQA1pMhQ6bBWV2hGLUg0rWg2aI0kxlUFQeTjGWrYRR Z0Rx5XaG5Xc+h1mYg5YtAy8Y0nvcAz5I0UWghEwgBVjghVyYhctMBVKoBLv/sgM+zQM0yF82iIIr qMEy+ANHsAQ/RgVTCAVLWARC8IPVZANAuIZtvVQjMeUpNee5lC2ncZp1ZumEaWd3vhiTtQM9WIOZ aD464ANA6AOifQTvNIWCNAVQOAVdnYRDMMboS4Q2dYREmMExIAMwsORXbARJoIRJaNNC6NM2aINH GFltHelxJufflC2N0AhgUemzXumWdunNieHE7co0qGY7+INC6JlDSGpImARLwARO8ARNoIRGQARD +ANCqN5H+GtKCAVEoOQ4gAM5UMxDQIRByOkU9YM7qAPF+wTXZduuFumvVlDZCpNAIevRbhq0Put1 1h7RYIZDmEE2uINBIMo6/2yERsDdnmkERiBGxjCEnkmESvAESHgES/iESYDBGBzUPtBnQeCDEw1a O6CDOKiDYhiH9/1YthVZzn5S2foksOBu0iZt0z5tUvhmNuBlvKVDSGhT6dvjRfCDytsDOlCDMBCD N+iD99uDxlYDL/jDJqCCL1iDO+C+fI2DpVtNRbCW9p1u6q5u685W2foSrugKLOJu0fbusm6fXxgE Oki6PPCDV2SEFYTsniGENjgDMcCCgm2DlyWEQgAEDyYEB16EQWCDK3iCJ/CCOFhMmFgDNlC8ORgG 9G1hBFddBV9wIHFw2LgKCFfyCQcLplkHXBAEDYeD6RwEQ+AOQxgEQZhDP//Agzdog8ZGg8Psg0XQ BFoQ6lCIhChH1q8EAzA4AznIRzo4g19dg1MwBxUGcjcW8u6hbtlyDalAcqzwJCXf7jHxhT6oA62O OqGNvD6wPHRtujxAbj0oBOtb70lIBDiFQcTLA8SUgzZYg1Bng2QMS1B45zuX5UVx49IVctkqkKj4 c0AP9CR/cC4aBj5gTTUgAzeAAzp4OjyIA9bkQDzAgz+w6xAcQRaxPu4YSut9U0Ht1fjOgizYAjSQ hZe+mFNH9VTPcxaWrQF59daIdatA8nWIBT14g4EluHwpAzYwVTOQbzVYPEEgRkSw6919BEgwnBK0 BO1dBGe+bDpAA8LDgj3/oIbUXmts3w9tX2NuXyPY8ysqmZwBoZxwN5CqWAdt+IQ+mAM2OAM0WNRU 9NWHGwMNbDjF5PDkjb7HJcFO6EdLcGY5iINRz28tSINXAG+1VmeFb2WGR3XZgpIpEXqKB3fWcAdF gQVZDfYzWAM0+IIwYPpUrIkyIAMxMMw6IFHvKwRMMAVR6IRRUAVQyARDONE2iCR7dQRwoHALB++0 1nmez3YRki2XwRRMEfqhJ3p0oJp0sAVB4PheXYMzuAwzsGZTRUWQD4NIggNJV4SjpITHzYRLaITU tE973YJEwI8mr3Cybnu3f3tXnvtKqZTfqPug9w2hl4eMESHVwQVDIATH/jBGNhiDlZ0Je3mDNwg/ MFU6dJVVNNUDxvgDOehDLziKNJCFWpdwJg+Tze98tZYtPjoJ0c8L0q97eliHcXCYjeAiYKBDPPyD P6gDiGPZalaDN5iD86+DxIs6OBgPNIBrkP8C4t8CLsACKlgFckf+JVf+y/FuYGH7swYIIAIDEQy0 7yDChAKB1GvosB69iBInzqto0aI8c+LMpVvHzp07ePDW9coUaZGhP3rkmEkDBgwZmGbi1JGTBo2Y l168aNGShUvPJlCiNJkjDqS7dkqXMmXn9CnUdVKnUl2n7ipWrAsLGkzodaG9sGEfkm04MeK8d+rS jRuXLp26diDnSZzn/q6Zqk+QEvXRUyeNGTFjwoTh8mRIEipdvnTJUuXwjRUufEyR0mUNr7jv3iHt DFJpUqZKoZJmV/X01oJevwoUO9ZeQ9ewHdKL1y6u09vs5L6DF4+uWbPz0h1LpakRoT1+9syB48ZN mzZr2qBBQ4ZMmCyVoVz5suVLGjzU3Ikjt/n8ec/qRYsuTXpdaoKrFbaWDdt+PXnvbsdFCo9uWu/E Iw9wr0F0zz0QqYNMKpYocoghEPZhxxx13HGHHXW0oUYZY3jBRRdgjMGGHdG4o8401ZyDDjrquBNP POjJqN567C31VHxdzbcPWGLVM9uP9cDTjoC9yVMXWrFBJJFs99iD/iA++DipoDKmnATJI40YsmUf d9RBBx1yoOEFGGGoIccf2XBmDjTlmBNOOiySU8456/gmEp7wyDgjjUgtleOOB4H1o4/ypEPOOS8S eOSSETU0D4FBBhmWk5RKGSWmUSbYDjKeUDIJI4gcIsged8QRBxtnVLcGHZqYI5I63UyzTjvogMPN NuKcA9dbHfkWT57B7omeeoAG2qM987QzDqLuHEkPpHRN5FA7H4l05FixUQplpvjk862m9azziyaZ WOKIIYLgYYcccbwBRx2kqDNPPOuQI40z5cDjDjneeLMNN+CAE44445RHZzrsDAgjjMEKu6exOy4U zznhnMMOPNCi/gXsRRpHxFmR8gxIz7ZO3nOptyl/m4+3K0tJTzm3WEIJI4Pw0cceeOAhzDuQslMO Ns48Q45p5ngDzjdII/3NN914g6s33egK1zsiM3w1sA7DE/F8C+1qlVKchSQtRYwCu5k7PUNbET1Q ut0yy+CuPDe498jjzSmNKKKuJ/pGBA862zTDTDTmmHbOnOaU4+ac5YxTjjjhEDwnOeEMjA488mi+ udVYx8P1agspzpE6Hpm2m1xEgvSfnpnrOXJdBNoj5cmYxg1ulHTLjU+y69RCySWxXPNRrdgc88w0 3qDzVulSOWXVWyyec4451ZdDTnnkgOPNN+H8BynnnH8+kGrH/gq022bV8oaeSALCeNH7aElEID31 1G57t/m3nDI+9ZSzBhSS4AQ1gCEW4aiGMKzBDW2gw3lE2o872HGVOJ3DTYurnjnmtCKjjaNi7LgI CCEFOtYAQSQDoh9wHAUksVQKSI3yGLTqob/caepJKMPUPeaxDEJwIQk92AETtJCMWmwDHBspXW7e 4ZETFax66FCc5b5RsMdt0BvieOI2vqG2ulhkhPQBAv3O4hD7mOxJCPLRC88yDxnqDnc4nFJY5vEL RrABDFBAwhKYsARKbMMc6CBHR9JRPReFph1scZMf0YE4y5XHjxVsmpzSAQ6OxMNREvEiQhbiuu8t 6UckQxAo/l1jskvB8SFnaRvc6NatJ4lLFXyARBq6MIlMlCEJmHCLaTyiDkWiA30gcYogp7fLc4hD HG6SXjnA0Q06HYUduiJmPB6CSUEJRGygERCj6mdG3tmPdm4roylPWRF76E4fuuMdtNyRjEQMAhKb qMQTxhAOdKRtP/txSkh+Mw94nC5OLXpieaanSHIkbSPiYIc82LGrwJ2DUNPk0fnwCbLzZO5ZrATl lOwHTte8MFloqYi39CHSfIh0pPhQaDiuIY1j9CENYaDJISgBJ87QqlofIRCAkqLLjqwlg45EnBEX t5HfKJQd6dDGOOZhj4cuREAOk1HmUliyJ83GkrSRSKMs/gKs2a2spPnQj2nScQ5IoOEQN0sEHqCg B3EkbCqla4dvNKend0SFVxXcYOXIYQ5ihsMdkFrHxdBBp3kwVSBjE1nVYnSn9n1PhVGaXYKCUz+I sO1ZaVwbOUuqD3xExDbkgIQc+nCGROzhDYAIwxCoQIy9wsUjRNKTPPCUvqe0wyOAfWLl6PTEb6jj SPF4iyH3WlgglK5qXCwbnthWP21utKoe2+d/sGoWGLFRHySzBz3kwQ063EEOdGBEKMigiD1AQQlY 2IUf0debAUEKRr3Z1zraKhXpNW5FfyRHxmrTz3EMVx4nUodITunJzr4OSdPSFmU1Ri/epBBa+snY C+HB/wo1EEIMf/iCIA6xiF3kgQ1LcMIvJqkOdrjvSGlpnZ5qtTwWKVKoG1Hkm9ahzxjdhr/kk4/5 gNAwwK7DuBMJy2Sxm11IqdFRhILN2iqSYhK/jy4iW5SD2dEKQTzCC3RoBBiSEIQ6cOEPU0CDFIJB jWy4JW2/Odtm9BTBu7Iog9gL6BPBQY4XKdkd6RDHcCsCLHcs77WaK1ALkfzRj9KFo2bZ2FxHsz41 p48bl1jEIpbQhEh0QQlBeMIb3qAIKjRhDNnwFzfYCtfWUXQ/64AcncwxMMvNCbfeSIez9vmOdHgj z9A6W1EfyN7JKgk2APrN2oxs1UNbRM1JIR5ojCqNTf7gIRJRMEIj6KAFJUChDlaA1xaoUIoVVdCY CUPdbkCyH3VorzzjEFg4kra4cHRjHOqo2r7UoQ1b18Y3nLlKPQtMlzLexyFkM3JwwunguYIsglYx BzAckYdRAEIPoWBFX46QBDHoQAZJgIIanpEOudh5eTzNJbgF6y90j+Mb3CNHyb1xsRgxMc/Sim3m 6CoVe/ZsjVQV8imlm8ZD1yVrdJUgi9RxjliEdhRq+IERHLGEQyinDT3QgRHIgIdN9FK2nLGpaa5i W2JCjXtLk2I4kAprmdsYCFwJFETBqNw80TTcZgaOpC4rbIqI02r7CisyR1EdSMCBCDtogyNE4Rcr MM7hB0ToAhHi0AsXNRlPEfx56a6SzG5oQxvcYNrIwQGwiyXlzi53FKTWK5Jr9qbQrJSsGOm+XKzW O7Yy32WcyDEKONAhEo3gghXSwAk19OEPhLhDGb6ABTFgQQ/l8OtyIcX2CNp0HRr5xja24XWwe+Ma FwsrOYa7kO1zv/ve/z74wy/+8ZO//Obn/tnRjhCusL/97n8//OMv//nTv/72l7/613///fO///7/ P/3l30EAIAEWoAEeIPsJoAIuIAM2oAM+IARGoAROIAVWoAWiHQA7 ------=_NextPart_000_0000_01CAA5B5.DA51AD90--