User Community Issues SASI Final Report 2.0 User Community Issues From its inception, development of NoteCards has been driven by the needs and feedback of real users utilizing the system for real tasks. In the PSA project, we monitored a single surrogate analyst and shaped the NoteCards system to his task and interaction style. In the SASI project we monitored a much larger community of approximately 25 users performing a wide variety of NoteCards-based tasks. Feedback from this user community in the form of informal observations as well as users' group meetings provided guidance in extending NoteCards to suit a broader range of users and specific applications. This section documents some of the NoteCards applications and experiences that members of the user community have had in using the system in a variety of contexts. A taxonomy of NoteCards applications has begun to emerge. In particular, NoteCards applications can be grouped into the following general categories: 1) argumentation projects, including both scientific and legal argumentation; 2) information structuring and analysis tasks; 3) authoring tasks; 4) personal and project management information bases; and 5) more traditional database applications. Specific NoteCards applications will be described for each of these categories. Many of the applications within the user community have required users to tailor the basic NoteCards system to their particular needs. By using the programmer's interface and by defining new card and substance types, the NoteCards user community has greatly extended the functionality available in NoteCards. Many of the new card types and functional enhancements for specific applications are described as library packages in Appendix E. 2.1 Argumentation Research Argumentation research defines a significant application area within the NoteCards user community. Since competitive argumentation, evidence-based reasoning, and hypothesis assessment have been important concerns of this project and its predecessor, these applications have been a research focus. Our aim has been to develop the technical and theoretical foundations for computer-based tools that will aid authors and readers in analysing and generating arguments. The SASI project has helped to develop a technology with the capabilities needed to carry out this research. Argumentation has been regarded as an important research vector for several reasons. First, argumentation is a key element in many kinds of authoring, including the creation of an analytic product; yet it is relatively unsupported by current tools and technologies. Systems that support the creation of linear documents tend to shift the author's attention away from the complex web of underlying arguments that comprise his subject. As a result, important assumptions may go unnoticed and connections between ideas, including contradictions, may be overlooked. Particularly during the early phases of developing an argument, it is important to be able to keep in view these complex interrelations among ideas and to separate out such elements of argumentation as premises, evidence, and conclusions. Second, there is evidence that not only the construction, but also the comprehension of complex arguments may be improved when they are presented in a structured manner. For example, Schum and Martin's studies [1982] show that jurists' and assessment of a body of trial evidence is substantially affected by breaking down the assessment task and explicitly representing the inference structures that relate pieces of evidence. Their findings suggest that making explicit the often implicit structure of an argument, including its evidence, may aid people in assessing and reasoning about it more effectively. NoteCards provides an ideal computer-based environment for argument research for two reasons: First, through its link-typing mechanism, it supports the development of epistemologies or categorizations of relationships among ideas. Such a capacity is necessary to represent the structures of argument. Second, NoteCards is able to handle multiple representations, whether these be different levels within a single structural scheme, or wholly different types of structures. This capability enables researchers to examine arguments at various degrees of granularity, to compare opposing arguments on a single topic, or to look for mappings between interacting structures within a single document. Two principal types of argumentation have been captured so far in notefile structures; one is the comparison of scientific hypotheses to determine their relative strength and the other is the representation of legal arguments for analysis of their structure. Both types of argumentation are discussed in this section. 2.1.1 Legal Argumentation Legal case analysis provides an ideal domain for studying structured arguments. An extensive storehouse of generic arguments is available in the form of statutes and on a long history of use of these arguments by the legal profession. Because of the constrained style of argumentation (largely appeal to precedent), the focus is limited and prototyping may be done quickly. The groundwork for legal case analysis research was laid during the SASI project by Art Farley, a computer scientist who studied law during his sabbatical from the University of Oregon. Farley designed a representation for contracts case briefs, including the arguments used to reach decisions in such cases. In order to instantiate these representations, a new structural entity was developed called a Cluster, a small network of pre-linked cards. Farley's work is mainly concerned with devising argument epistemologies and designing representations of these in NoteCards. An eventual goal of this work is to perform automated inferencing over the resulting structures. Automated inferencing may force a reconsideration of the effectiveness of the existing representations. However, the current structures demonstrate the usefulness of NoteCards for argument research, as well as convey the potential practical value of a tailored NoteCards system. Farley's framework partitions a contracts case into issues, facts, decision, and rationale. Issues consist of summarizing text with particular emphasis on the case's import for other cases. Facts are divided into legal facts and real facts. The former include facts about the legal proceedings of the case, while the latter cover relevant real-world data. Decision text, in this instance, describes the judge's decision on the case, but alternatively might contain the current status of pending cases. Finally, the rationale text consists of arguments used to reach the decision. These arguments, in turn, have their own structure. Each argument is viewed as an instance of an application of a generic rule. Some rules are well known from jurisprudence, some exist as statutes in the law books, and others are synthesized from statutes, precedents and an interpretation of the case at hand. Instantiating or applying a rule requires the user to analyze and identify relevant features of the given case in order to bind the rule's free variables. Because rules appeal to other rules for support, an argument is represented in this scheme as a network of rule applications. This rule-based scheme for representing legal arguments is captured in NoteCards using the Cluster facility. A cluster is a card type that, when selected by the user, creates a template for a mini-network of text cards whose linkages are predefined by the user for the purpose at hand. The ClusterCard capturing Art's representation is the CaseCluster. (See Figure 2.1-1.) At the top level, a CaseCluster consists of a head cluster card linked to four text cards representing the issues, facts, decision and rationale for the given case. The rationale card then contains links to subnetworks representing the arguments used in the case. These are captured using a specialized browser card called an Argument, which displays a network of rule applications. Figure 2.1-2 shows an argument card. Each application is captured using a RuleInstance card. The RuleInstance card contains a link to a Rule card describing the generic rule and to a Bindings card containing a list of paired variable names and their values, where the values are drawn from the case at hand. Finally, RuleInstance cards contain links to other rule instances providing support for this rule. In this way, a network of warrants and backings forming the argument is built up. |<9s33 ?3 ?$fx$fx0s0<8|<9p9ssg3?3'O?3 ??8g0p'dyϓ ?3 ??ɓ''&'dyϐ9fs9$fx'#'$3ϓ8$fx8g 0'$3O @` d@A@`??@A````@`b@ @"w ?@`b@ $ D @`b\9\ 8pDp.D.xKg8p$q.z8 @`bDH"b DD1"D̑1 LADĉ$1"DD??LH`B|OB|(!(! HA9!"D8 H@`B@H2B@(!2(! HAd!"D HH@`BDH"BD1"! HAD!"DDB HH`B8hGB8pp.!HA:pu!D8B ?G@` @` ````8@?` @!`.<ÞpxDq< @#=.`& 1#HLHH"$$H! FDB H@!"$$f$đH!1`$ !!HH"#p! D(BO@!>$$$$!!`$ !!HHH" H! D2(BH @! $$$$H!!`$ !#HHHH"$$H! D"BH@ !!"Ą$dd$H!!` F Æp 4q<G@ `ã# !`  `@```   @@`  @  @@` @:8x:9.x@ `xpx8py\8t\K8"FD FD1 @ D `$&!Б"D"bDbMDB B|! @ D `r$$!8"B|BH|2Bd B@! @ D `$$!2"B@BH@"FD FD! HD `$$!"D"bDBD:::803`xp"x8p\8tBtgH8B`< `x`@`@``,G_K8tp`5$$H"j"LA2D`!">BHAr"`!" B2HA "d`!$!"B"HA"D` AH@":tp``x```H $`H $`Háp#g!/8t=g0`"I2$`A$A!$DL`I AqG!$|H`2I A D!$@H`"I2$`A A#$DH`Hàg胁#8t`````p@8`@@ `G=Ŝ8qpx8,`#"DLHȘ&D! D5 `!>GHC$>! ! `! DHL$ d! d! `#"DHHȐ$"D! D! `CGGP:q :`````@`@`<\98p#<`"BbFD D MLH$H` BBD@ |8HHOGÐ` BBD2@ @HHHD `"BBD"D D HHHH`HH!"D"C! | `B$$H$ HH!"D"B! @| `@`B đ$$H$"HH!"D"B$!"D3f``B#HG!4pAĎ88`08fsg9`lgm|mmm`Ͷ|ǰmfo }`Ͷ`3mfٍl a`<8fg=` " ` "  `0Ã8!!!Ȉrr$9 >!>@` !!&HAE$ ! @`"D!!dH2$AB$"!"B`8h΍!ÀuqB`````Figure 2.1-1 Sample Case Cluster N??9???y>8808?8& G399̜ܓ2rs?󓓓 99Ϝ̜2rs?󓓓 y9 2s?󓓓y9?ɜ̟2~s?󓓓999ɜ̜2rs?883??N& $3s?  @ @ @ @ >fٞ1m6mm`c|l0@ 3fͽ 16ao``f|0@ 3fͽ͛ 1mal6`cf`0@ >>@ ! 33@ !͜39q3m !mm mm !}om ! l3ml3m$ !Ǿ}= ` `  @ @0 @ 0   0    8      0   LLL``0Ls|<` L ٳs`3fX 8 L ٳ{``0fCgxٶ0   0͛766l30p a 0͛7fl39s3qw獛la 07f{o> aٶmفm P͛6<{l3|oنmفm P͛6<6l3lm mفm PwϘ3|vso͍mፘ        2`` @ 8϶>6yٶ  2 `f6cma6  c6cٶ &cf6c ٶ8 ;9{ @                 @ & @> (@@> 0 @? >@ >NI >NC >NC @>NI @>NL @@NNa  @   @ @  ~ @  A @  A!#@ @ A!$@@ $~!'@ A!$@ *A#$@@ HA#@  @   @  @ p @ P~ $@ PA @ PA%ät@ (~&$dČ@ A$$$@ A$$$@ A$$d@ @~$#t@  @ x@ @ @ > @ Ǐ9|A @ lٹ``8@yB8@ lٽ03`ٜ98yͶq<"$BD8@ ٽ{33xcmflmlٰ@ B8@ Dlٷy3`cmlmlͶ0#$B@ Hlٷ`ٶcmmlͶA"$FD@ PǏ3|پ;9lpx>:8@ ``@  |`@ 30  30  gٶ  fٍ6 ٶ ٶ ٞ     @     @     @      s|<>0    ٳs`3f30p a ٳ{``0f9s3qw獛la {xgf eٶmفm ! ٳo`ǰf3|kنmفma A ٳo`3fli mفma g|<|vso͍mፙ        3@ !3@ !8ٶ @ !ml66 !o} !l a g=           @ @ @          >fٞ3m3fmf7Ͷ̈́ 3fͽmfmfm $3fͽ͛ͳm7mf6m D>m  6cl0 m  9g     Figure 2.1-2 Sample Argument Card This application has also helped motivate the development of an interface to Mead Data Central's Lexis database to provide legal case source material. Extending the case analysis application to include Lexis access is an interesting example of the use of the NoteCards system with a large remote database in a specialized applications area; it may serve as a model for later database interfaces. The extension also involves a new text-based card type to act as a holder for Lexis derived sources. 2.1.2 Scientific Argumentation Like legal reasoning, scientific argumentation is guided by a body of specialized, and largely implicit, constraints that can be usefully represented in NoteCards. These include a specialized terminology (often created through prior definitional arguments), prescribed standards for the assessment and validation of evidence in relation to hypotheses, and often, a structured format for the presentation of data, analyses and conclusions. Scientific argumentation may also include developing and assessing new methodologies for constructing and justifying hypotheses. Current work on scientific argumentation in NoteCards concerns just such a development and has shown its usefulness as a system for representing and analyzing arguments constrained by a particular methodology. The principal scientific argumentation research using NoteCards was performed by Kurt VanLehn. This application made the assumption that hypothesis development has already been completed and the system is primarily used for browsing. Kurt VanLehn is a cognitive scientist at PARC's Intelligent Systems Laboratory studying learning in elementary school children. For his Ph.D. dissertation, he developed a methodology called Competitive Argumentation as a means of validating inferences from observable data on complex underlying cognitive processes [VanLehn et al 82]. VanLehn's methodology is captured in a large notefile containing experimental data, a set of explanatory hypotheses derived from alternative cognitive theories, and arguments for and against each hypothesis. The benefits of using NoteCards for representing argument structures in scientific methodologies are discussed in detail in [VanLehn 85], included in this report as Appendix F. Briefly, his methodology calls for dividing a research endeavor into problems or issues. These are explained by competing hypotheses that in turn are compared through argumentation. Arguments use experimental data as well as various premises and other antecedents to argue for or against a hypothesis. The methodology also employs moot arguments that are directed at the hypothesis, but are not in themselves decisive. This strategy as encapsulated in NoteCards makes extensive use of browsers and a new NoteCards matrix structure especially created for this purpose. One important result of using NoteCards as a browsing mechanism was that VanLehn discovered flaws in his argumentation while constructing the NoteCards-based organization; missing relationships were uncovered and critical assumptions were identified. As a result of creating and reorganizing the NoteCards represention of the argumentation, a formerly winning hypothesis was reassessed as weaker than its competition. 2.2 Analysis Applications Applications involving analysis tasks reflect potential NoteCards uses within the intelligence community. Some of the tasks approached using NoteCards include technology assessment and analysis, an implementation of system analysis techniques, research in the design process, and a competitive projective analysis of spacecraft. 2.2.1 Competitive/Projective Analysis An in-house analyst is using NoteCards to characterize the attributes of various spacecraft for use in competitive projective analysis tasks. Her application requires parallelism in the lower level FileBox structures (i.e. each vehicle must be characterized by the same types of attributes). The analyses performed depend on the ability to make documents from like elements in the parallel structures. For example, one analysis may require an investigation into the electrical power requirements of each vehicle. This application is well-suited to investigating uses for structural templates, especially as a means for discovering missing information. The programmer's interface functionality is provides the mechanism to construct this type of structural template in the form of a specialized card type, thereby propagating parallel structures within a NoteFile or set of NoteFiles. This application uses a variety of notecard types to maintain source information for analysis. Some of the specialized Spacecraft NoteFile's cards contain AIS images as bitmaps, and hardcopy source material text rendered electronic by a Kurzweil scanner. These cards illustrate how current technology may be used in conjunction with the Notecards information organizing environment. Another application in this area takes the form of a competitive analysis of major commercially available expert systems products. This notefile contains a broad survey of expert systems as well as a comparison of the products along several dimensions. 2.2.2 Systems Analysis Techniques The FAST/FMEA NoteFile demonstrates the NoteCards system's role in implementing systems analysis techniques. FAST (Functional Analysis System Technique) and FMEA (Failure Modes Effects Analysis) are two standard analysis techniques used in designing large systems, e.g., copiers. Currently FAST and FMEA analyses for copiers are done using paper and pencil supplemented by a variety of different and incompatible information systems. The FAST/FMEA NoteFile is an attempt to integrate all of the information necessary for doing FAST and FMEA analyses into a single database. Further, the FAST/FMEA NoteFile is a demonstration of how the tools available in NoteCards can be used to help in the task of developing maintenance documentation from this integrated database. 2.2.3 Technology Assessment and Analysis SAM NoteFile -- The SAM (Systems Analysis Module) system is a collection of tools that operate over a Notecards database. The database contains information on various xerography subsystems, configurations and technologies. SAM allows a user to quickly construct and analyze alternative copier configurations for system characteristics such as power consumption and system throughput. 2.2.4 Analysis of the Deployment of Nato Missiles A surrogate intelligence analyst has used NoteCards to perform a major analysis task - the deployment of missiles in Europe. This task entailed the collection and structuring of information from diverse sources. He has also used NoteCards to restructure and reorganize his information. His task has been the subject of extensive empirical studies. Section 3 discusses many of the issues that have been investigated using data from this work. NoteCards is also used to browse and analyze the growth and changes in the Nato Missiles notefile structure over time. 2.2.5 Research in the Design Process A development site NoteCards user analyzes videotapes of design meetings, using NoteCards as a workbench and an events database. His notefiles represent temporal aspects of the design process by incorporating browser-based time lines of videotaped events. 2.3 Personal Databases and Information Management Personal database and informal information management tasks define a large portion of user community applications. Most of these applications use the basic NoteCards functionality to organize a combination of structured and unstructured information. Several personal database applications center around organizing electronic mail; others use NoteCards to manage research projects or development efforts; a number of notefiles are caches for information normally stored on a physical medium, such as a Rolodex or calendar. 2.3.1 Personal Databases as Information Caches A range of NoteCards applications can be categorized as personal information caches; they are informally maintained notefiles that may make use of supplemental programmer's interface-based functionality. For example, a member of the NoteCards development team has implemented his Rolodex as a notefile. Using multiple filing, he has created a scheme in which there is a phones filebox and a people filebox; the people are further classified by whether they are friends, Xerox contacts, colleagues, or people seldom contacted. Alphabetizing (functionality implemented by the NCHACKS library package described in Appendix E) was implemented for this application to make the phones filebox more manageable. The NCKEYS library package (also described in Appendix E) was also developed for this application to make the information accessible via keystrokes, eliminating dependence on mouse-driven interaction. Another application that has driven the development of specialized programmer's interface code is a notefile to keep track of how its user spends his time. This user has written a NoteCards library package called "MakeTR" to automatically create and title his cards. This application also includes the idea of embedding Programmer's Interface functions in cards - first the functions are pretty-printed, then copied into cards. Several applications notefiles are devoted to customer support type activities. These notefiles generally maintain records of all transactions between software distributors and customers. Their organizational strategy is analogous to that of the Rolodex notefile described earlier; for example, a NoteFile tracking NoteCards distribution to internal and customer sites has three main types of organization. Cards are filed by source, by sales, and by customer. Similarly, NoteCards is being used to track received bug reports from its user community. A NoteCards Support notefile contains bug reports and status, suggestions, and records of mail discussions. Another example of a personal information cache application is a notefile documenting various NoteCards applications. The NoteCards Applications notefile contains descriptions of the user community applications in terms of cards describing the projects and the people with links to the individual responsible for designing, maintaining, and demonstrating the applications notefile. An analogous combination of formal, highly structured information and loose unstructured information is found in an annotated organization chart notefile. This application illustrates the difference between the NoteCards concept of a database and a traditional relational database. Multiple organizations of the same information are facilitated. 2.3.2 Managing Research Projects and Development Efforts One of the larger NoteCards applications notefile is used to organize and track projects and ideas for Vista's AI Applications group. Bibliographical references are maintained in this notefile, as are agenda items. The Sketch and TEdit functionality are used extensively in this NoteFile, since report preparation is one of its primary functions. Maintaining project information this way allows the developers to share resources and information gathered in performing tasks related to individual projects. Temporal relationships between events in various contexts are also captured this way. Informal templates for structuring project organization may be used to create parallel structures in analogous situations in different projects. Templates are useful for standardizing elements of basic project organization (such as schedule, design, report generation activities, and maintaining references to outside sources). The projects motivating the development of this application share a lot of underlying concepts; the assumption is that using a single notefile to organize and articulate ideas and general system structures will result in a greater level of coherence and synergy. A smaller scale notefile has been used to manage the development of a single project, CALLE, Computer Assisted Language Learning Environment. One of the interesting features of this NoteFile is the use of NoteCards to investigate data structures in LFG (the Grammar Writer's Workbench based on Lexical Functional Grammar Theory developed by Ron Kaplan and Joan Bresnan). The NoteFile looks at the individual data structures and their interrelationships. 2.3.3 Maintaining an Electronic Mail Database A common personal database NoteCards application is based on the need to organize a large volume of electronic mail traffic. Applications described earlier in this section are driven by the flow of electronic mail related to a specific topic; there are many other application notefiles that are used as a more general mechanism for archiving and managing this type of information. Because NoteCards is implemented in an integrated environment, it is easy to have concurrent access to both NoteCards and the mail system. The shift-select metaphor for copying text allows a user to save selected fragments of messages as well as whole messages in a notefile for this purpose. Furthermore, chains of correspondence may be linked together to represent the continuity of the messages. An independent mail sorting project has provided some additional functionality in this applications area. It has been integrated with NoteCards to provide some automated capabilities for organizing, storing, and maintaining records of electronic message traffic. 2.3.4 NoteCards-Based Authoring Efforts Many of the personal database NoteCards applications have used the system as an authoring environment. For example, next year's SASI proposal was written in NoteCards, using a single notefile as the focal point for the authoring process. The NCFileCard library function (see Appendix E) was used to facilitate TEdit file-based collaboration. Other efforts do not have as concrete an objective; many NoteCards users maintain general notefiles containing their work with eventual hopes of authoring papers based on these notes. 2.3.5 NoteCards in a Self-Referential Demonstration A self-referential notefile has been created to demonstrate the system and provide an informal basis for training new users. The motivation for this application was an increase in the frequency of NoteCards demonstrations and a similar increase in the diversity of the audience. The growing complexity of the system itself also motivated the preparation of this specialized notefile; an objective of the application was to explain the underlying model for the system. This notefile has portions extracted and compiled from extant notefiles to demonstrate the range of applications and types of information; scanned images and maps are included, as well as Kurzweil text, and a variety of graphs and sketches. Notecard text and sketches are used as a self-illustrating system description mechanism. The notefile guides a inexpert NoteCards user through the demonstration material including a step-by-step description of using basic functionality. It also contains an interactive demonstration script and may be initiated from a background menu driven interface implemented with the programmer's interface. Cards also contain programmer's interface function calls which may be used as another aspect of the demonstration. 2.4 Instructional Design Several applications involving instructional design for computer-based training systems have incorporated NoteCards in a design and presentation role. The principal example of an application in this area is the ARI project's Instructional Design Environment (IDE). IDE is an environment for authoring and presenting training courses directed at copier technicians. The application includes tools for representing decisions and rationale used during course design. It also includes two new card types, animation cards and video cards as well as full screen "slide presentations." 2.5 Database Applications Several applications have used NoteCards to create database resources or interface to external databases. In cases where NoteCards has been used to build a large notefile with a defined type of information (such as the World Factbook application), Release 1.3 (already running at Xerox PARC) will allow the database resource notefile to be open concurrently with the analyst's working notefile. Programmers interface code has been developed to automatically generate a notefile from The World Factbook. The World Factbook contains data about nations, including facts about the nation's physical features, the people, the government, and the economy. The basic hierarchical structure was built using the programmer's interface; relational links were added by hand as a proof-of-concept illustration of the types of relationships which could strengthen the structure. Another type of database resource application consists of a programmer's interface-based window on the Lexis legal database. For this application a new card type has been created which is a text card obtained from source Lexis. For parallel applications, work is being done on an interface to videodisk and other sources. A second application in this area is the Dialog Chunker. The Dialog Chunker takes typescript input from a Dialog interaction session and parses it, filing the bibliographic references found into a NoteCards network. (LIST ((PAGE NIL NIL (0 0 612 792) ((FOLIO NIL (PARALOOKS (QUAD CENTERED) CHARLOOKS (FAMILY NIL OVERLINE OFF STRIKEOUT OFF UNDERLINE OFF SLOPE REGULAR WEIGHT MEDIUM)) (270 36 72 36) NIL) (HEADING NIL (HEADINGTYPE DOCUMENT) (72 756 468 36) NIL) (HEADING NIL (HEADINGTYPE CHAPTER) (72 756 468 36) NIL) (TEXT NIL NIL (72 72 468 648) NIL))) (PAGE NIL NIL (0 0 612 792) ((FOLIO NIL (PARALOOKS (QUAD CENTERED) CHARLOOKS (FAMILY NIL OVERLINE OFF STRIKEOUT OFF UNDERLINE OFF SLOPE REGULAR WEIGHT MEDIUM)) (270 36 72 36) NIL) (HEADING NIL (HEADINGTYPE DOCUMENT) (72 756 468 36) NIL) (HEADING NIL (HEADINGTYPE CHAPTER) (72 756 468 36) NIL) (TEXT NIL NIL (72 72 468 648) NIL))) (PAGE NIL NIL (0 0 612 792) ((FOLIO NIL (PARALOOKS (QUAD CENTERED) CHARLOOKS (FAMILY NIL OVERLINE OFF STRIKEOUT OFF UNDERLINE OFF SLOPE REGULAR WEIGHT MEDIUM)) (270 36 72 36) NIL) (HEADING NIL (HEADINGTYPE DOCUMENT) (72 756 468 36) NIL) (HEADING NIL (HEADINGTYPE CHAPTER) (72 756 468 36) NIL) (TEXT NIL NIL (72 72 468 648) NIL)))))T T PAGEHEADINGDOCUMENT PAGEHEADINGCHAPTER  TIMESROMAN TIMESROMAN 1(DEFAULTFONT 1 (GACHA 10) (GACHA 8) (TERMINAL 8)) TIMESROMAN TIMESROMAN A%c?xIU; BMOBJ.GETFN21(DEFAULTFONT 1 (GACHA 10) (GACHA 8) (TERMINAL 8)) "NR BMOBJ.GETFN21(DEFAULTFONT 1 (GACHA 10) (GACHA 8) (TERMINAL 8)) #T<J&v")25%3 /:/(5L"z