ࡱ>  bjbj Hhhg   $PEL$x$L#>L#L#şNT.T.T.L#.T.L#T.T.@^Pȧz)nPvzl5+5^^5ޚ, T. !-L#L#L#L#5 : GrayWulf: Scalable Clustered Architecture for Data Intensive Computing Alexander S. Szalay1, Gordon Bell2, Jan Vandenberg1, Alainna Wonders1, Randal Burns1, Dan Fay2, Jim Heasley3, Tony Hey2, Maria Nieto-SantiSteban1, Ani Thakar1, Catharine van Ingen2, Richard Wilton1 The Johns Hopkins University, 2. Microsoft Research, 3. The University of Hawaii  HYPERLINK "mailto:szalay@jhu.edu" szalay@jhu.edu,  HYPERLINK "mailto:gbell@microsoft.com" gbell@microsoft.com,  HYPERLINK "mailto:jvv@jhu.edu" jvv@jhu.edu,  HYPERLINK "mailto:alainna@pha.jhu.edu" alainna@pha.jhu.edu,  HYPERLINK "mailto:randal@cs.jhu.edu" randal@cs.jhu.edu,  HYPERLINK "mailto:dan.fay@microsoft.com" dan.fay@microsoft.com,  HYPERLINK "mailto:heasley@ifa.hawaii.edu" heasley@ifa.hawaii.edu,  HYPERLINK "mailto:tony.hey@microsoft.com" tony.hey@microsoft.com,  HYPERLINK "mailto:nieto@pha.jhu.edu" nieto@pha.jhu.edu,  HYPERLINK "mailto:thakar@jhu.edu" thakar@jhu.edu,  HYPERLINK "mailto:vaningen@windows.microsoft.com" vaningen@windows.microsoft.com,  HYPERLINK "mailto:rwilton@pha.jhu.edu" rwilton@pha.jhu.edu Abstract Data intensive computing presents a significant challenge for traditional supercomputing architectures that maximize FLOPS since CPU speed has surpassed IO capabilities of HPC systems and BeoWulf clusters. We present the architecture for a three tier commodity component cluster designed for a range of data intensive computations operating on petascale data sets named GrayWulf. The design goal is a balanced system in terms of IO performance and memory size, according to Amdahls Laws. The hardware currently installed at JHU exceeds one petabyte of storage and has 0.5 bytes/sec of I/O and 1 byte of memory for each CPU cycle. The GrayWulf provides almost an order of magnitude better balance than existing systems. The paper covers its architecture and reference applications. The software design is presented in a companion paper. Trends of Scientific Computing The nature of high performance computing is changing. While a few years ago much of high-end computing involved maximizing CPU cycles per second allocated for a given problem; today it revolves around performing computations over large data sets. This means that efficient data access from disks and data movement across servers is an essential part of the computation. Data sets are doubling every year, growing slightly faster than Moores Law[1]. This is not an accident. It reflects the fact that scientists are spending an approximately constant budget on more capable computational facilities and disks whose sizes that have doubled annually for over a decade. The doubling of storage and associated data is changing the scientific process itself, leading to the emergence of eScience as stated by Grays Fourth Paradigm of Science based on Data Analytics[2]. Much data is observational, due to the rapid emergence of successive generations of inexpensive electronic sensors. At the same time large numerical simulations are also generating data sets with increasing resolutions, both in the spatial and temporal sense. These data sets are typically tens to hundreds of terabytes[3,4]. As a result, scientists are in a dire need of a scalable solution for data-intensive computing. The broader scientific community has traditionally preferred to use inexpensive computers to solve their computational problems, rather than remotely located high-end supercomputers. First they used VAXes in the 80s followed by low-cost workstations. About 10 years ago it became clear that the computational needs of many scientists exceeded that of a single workstation and many users wanted to avoid the large, centralized supercomputer centers. This was when laboratories started to build computational clusters from commodity components. The idea and phenomenal success of the BeoWulf cluster[5] shows that scientists (i) prefer to have a solution that is under their direct control, (ii) are quite willing to use existing proven and successful templates, and, (iii) generally want a do-it-yourself inexpensive solution. As an alternative to building your own cluster, bringing computations to the free computer resource became a successful paradigm, Grid Computing[6]. This self-organizing model, where groups of scientists pool computing resources irrespective of their physical location suits applications that require lots of CPU time with relatively little data movement. For data intensive applications the concept of cloud computing is emerging where data and computing are co-located at a large centralized facility, and accessed as well-defined services. This offers many advantages over the grid-based model and is especially applicable where many users access shared common datasets. It is still not clear how willing scientists will be to use such remote clouds[7]. Recently Google and IBM have made such a facility available for the academic community. Due to these data intensive scientific problems a new challenge is emerging, as many groups in science (but also beyond) are facing analyses of data sets in tens of terabytes, eventually extending to a petabyte since disk access and data-rates have not grown with their size. There is no magic way to manage and analyze such data sets today. The problem exists both on the hardware and the software levels. The requirements for the data analysis environment are (i) scalability, including the ability to evolve over a long period, (ii) performance, (iii) ease of use, (iv) some fault tolerance and (v) most importantlow entry cost. Database-Centric Computing Bring analysis to the data, not vice-versa Many of the typical data access patterns in science require a first, rapid pass through the data, with relatively few CPU cycles carried out on each byte. These involve filtering by a simple search pattern, or computing a statistical aggregate, very much in the spirit of a simple mapping step of MapReduce[8]. Such operations are also quite naturally performed within a relational database, and expressed in SQL. So a traditional relational database fits these patterns extremely well. The picture gets a little more complicated when one needs to run a more complex algorithm on the data, not necessarily easily expressed in a declarative language. Examples of such applications can include complex geospatial queries, processing time series data, or running the BLAST algorithm for gene sequence matching. The traditional approach of bringing the data to where there is an analysis facility is inherently not scalable, once the data sizes exceed a terabyte due to network bandwidth, latency, and cost. It has been suggested [2] that the best approach is to bring the analysis to the data. If the data are stored in a relational database, nothing is closer to the data than the CPU of the database server. It is quite easy today with most relational database systems to import procedural (even object oriented) code and expose their methods as user defined functions within the query. This approach has proved to be very successful in many of our reference applications, and while writing class libraries linked against SQL was not always the easiest coding paradigm, its excellent performance made the coding effort worthwhile. Typical scientific workloads Over the last few years we have implemented several eScience applications, in experimental data-intensive physical sciences applications such as astronomy, oceanography and water resources. We have been monitoring the usage and the typical workloads corresponding to different types of users. When analyzing the workload on the publicly available multi-terabyte Sloan Digital Survey SkyServer database[9], it was found that most user metrics have a 1/f power law distribution[10]. Of the several hundred million data accesses most queries were very simple, single row lookups in the data set, which heavily used indices such as on position over the celestial sphere (nearest object queries). These made up the high frequency, low volume part of the power law distribution. On the other end there were analyses which did not map very well on any of the precomputed indices, thus the system had to perform a sequential scan, often combined with a merge join. These often took over an hour to scan through the multi-terabyte database. In order to submit a long query, users had to register with an email address, while the short accesses were anonymous. Advanced user patterns We have noticed a pattern in-between these two types of accesses. Long, sequential accesses to the data were broken up into small, templated queries, typically implemented by a simple client-side Python script, submitted once in every 10 seconds. These crawlers had the advantage from the users perspective of returning data quickly, and in small buckets. If the inspection of the first few buckets hinted at an incorrect request (in the science sense), the users could terminate the queries without having to wait too long. The power users have adopted a different pattern. Their analyses typically involve a complex, multi-step workflow, where the correct end result is approached in a multi-step hit-and-miss fashion. Once they zoom in on a final workflow, they execute it over the whole data set, by submitting a large job into a batch queue. In order to support this, we have built MyDB, a server-side workbench environment[11], where users get their own database with enough disk space to store all the intermediate results. Since this is server-side, the bandwidth is very high, even though the user databases reside on a separate server. Users have full control over their own databases, and they are able to perform SQL joins with all the data tables in the main archive. The workbench also supports easy upload of user data into the system, and a collaborative environment, where users can share tables with one another. This environment has proved itself to be incredibly successful. Today 1,600 astronomers, approximately 10 percent of the worlds professional astronomy population, are daily users of this facility. In summary, most scientific analyses are done in a exploratory fashion, where everything goes, and few predefined patterns apply. Users typically want to experiment, try many innovative things that often do not fit preconceived notions, and would like to get very rapid feedback on the momentary approach. In the next sections we will discuss how we expand this framework and environment substantially beyond the terabyte scale of today. Building Balanced Systems Amdahls laws Amdahl has established several laws for building a balanced computer system [12]. These were reviewed recently[13] in the context of the explosion of data. The paper pointed out that contemporary computer systems IO subsystems are lagging CPU cycles. In the discussion below we will be concerned with two of Amdahls Laws: A balanced system needs one bit of IO for each CPU cycle has 1 byte of memory for each CPU cycle These laws enumerate a rather obvious statement in order to perform continued generic computations, we need to be able to deliver data to the CPU, through the memory. Amdahl observed that these ratios need to be close to unity and this need has stayed relatively constant. The emergence of multi-level caching led to several papers pointing out that a much lower IO to MIPS ratio coupled with a large enough memory can still provide a satisfactory performance[14]. While this is true for problems that mostly fit in memory, it fails to extend to computations that need to process so much data (PB) that they must reside on external disk storage. At that point having a fast memory cache is not much help, since the bottleneck is disk IO. Raw sequential IO For very large data sets the only way we can even hope to accomplish the analysis if we follow a maximally sequential read pattern. Over the last 10 years while disk sizes have increased by a factor of 1,000, the rotation speed of large disks used in disk arrays has only changed a factor of 2 from 5,400 rpm to 10,000 rpm. Thus random access times of disks have only improved about 7% per year. The sequential IO rate has grown somewhat faster as the density of the disks has increased by the square root of disk capacity. For commodity SATA drives the sequential IO performance is typically 60MB/sec, compared with 20MB/sec 10 years ago. Nevertheless, compared to the increase of the data volumes and the CPU speedups, this increase is not fast enough to conduct business as usual. Just loading a terabyte at this rate takes 4.5 hours. Given this sequential bottleneck, the only way to increase the disk throughput of the system is to add more and more disk drives and to eliminate obvious bottlenecks in the rest of the system. Scale-up or scale-out? A 20-30TB data set is too large to fit on a single, inexpensive server. One can scale-up, buying an expensive multiprocessor box with many fiber channel (FC) Host Channel Adapters (HCA) and a FC disk array, easily exceeding the $1M price tag. The performance of such systems is still low, especially for sequential IO. To build a system with over one GB/sec sequential IO speed one needs at least 8 FC adapters. While this may be attractive for management, the entry cost is not low! Scaling out using a cluster of disks attached to each computing node provides a much more cost effective and high throughput solution, very much along the lines of BeoWulf designs. The sequential read speed of a properly balanced mid-range server with many local disks can easily exceed a GB/sec before saturation[15]. The cost of such a server can be kept close to the $10,000 range. On the other hand managing an array of such systems, and manually partitioning the data can be quite a challenge. Instead of mid-range servers the scale-out can be done on lower-end machines, and deployed to a very large number (~100,000), as done by Google. Given the success of the BeoWulf concept for academic research, we believe that the dominant solution in this environment will be deployed locally. Given the scarcity of space at universities it also needs to have a high packing density. The GrayWulf System 4.1. Overall Design Principles We are building a combined hardware and software platform from commodity components to perform large-scale database-centric computations. The system should scale to petabyte-size data sets provide very high sequential bandwidth to data support most eScience access patterns provide simple tools for database design provide tools for fast data ingest This paper describes the system hardware and the hardware monitor tools. A second paper describes the software tools that provide functionality for (c) (e). 4.2. Modular, layered architecture Our cluster consists of modular building blocks, in three tiers. Having multiple tiers provides a system with a certain amount of hierarchical spread of memory and disk storage. The low level data can be spread evenly among server nodes on the lowest tier, all running in parallel, while query aggregations are done on more powerful servers in the higher tiers. The lowest, tier 1 building block is a single 2U sized Dell 2950 server, with two quad core 2.66GHz CPUs. Each server has 16 GB of memory, two PCIe PERC6/E dual-channel RAID controllers and a 20 Gbit/sec QLogic SilverStorm Infiniband HCA, with a PCIe interface. Each server is connected to two 3-U MD1000 SAS disk boxes that contain a total of 30- 750 GB, 7,200 rpm SATA disks. Each disk box is connected to its dedicated dual-channel controller (see section 4.3). There are two mirrored 73 GB, 15,000 rpm disks residing in internal bays, connected to a controller on the motherboard. These disks contain the operating system and the rest of the installed software. Thus, each of these modules takes up 8 rack units, and contains a total of 22.5TB of data storage. Four of these units with UPS power is put in a rack. The whole lower tier consists of 10 such racks, with a total of 900TB of data space, and 640 GB of memory. Tier 2 consists of four Dell R900 servers with 16 cores each and 64 GB of memory, connected to three of the MD1000 disk boxes, each populated as above. There is one dual channel PERC6/E controller for each disk box. The system disks are two mirrored 73GB SAS drives at 10,000 rpm and a 20Gbit/sec SilverStorm Infiniband HCA. This layer has a total of 135TB of data storage and 256GB of memory. We also expect that data sets that need to be sorted and/or rearranged will be moved to these servers, utilizing the larger memory. Finally, tier 3 consists of two Dell R900 servers with 16 cores, 128 GB of memory, each connected to a single MD1000 disk box with 15 disks, and a SilverStorm IB card. The total storage is 22.5TB and the memory is 256 GB. These servers can also run some of the resource intensive applications, complex data intensive web services (still inside the SQL Server engine using CLR integration) which require more physical memory than available on the lower tiers. servercoremem[GB]disk[TB]CountTier 1295081622.5040Tier 2R900166433.754Tier 3R9001612811.252total41611521057.5046Table 1. Tabular description of the three tiers of the cluster with aggregates for cores, memory and disk space within our GrayWulf system. The Infiniband interconnect is through a Qlogic SilverStorm 9240 288 port switch, with across-sectional aggregate bandwidth of 11.52 Tbit/s. The switch also contains a 10 Gbit/sec Ethernet module that connects any server to our dedicated single lambda National Lambda Rail connection over the Infiniband fabric, without the need for dedicated 10 Gbit Ethernet adapters for the servers. Initial Infiniband testing suggests that we should be able to utilize at least the Infiniband Sockets Direct Protocol[16] for communication between SQL Server instances, and that the SDP links should sustain at least 800-850 MB/sec. Of course, we hope to achieve the ideal near-wirespeed throughput of the 20 Gbit/sec fabric. This seems feasible, as we will have ample opportunity to tune the interconnect, and the Windows Infiniband stack itself is evolving rapidly these days. The cluster is running Windows Enterprise Server 2008 and the database engine is SQL Server 2008 that is automatically deployed across the cluster. 4.3. Balanced IO bandwidth The most important consideration when we designed the system (besides staying within our budget) was to avoid the obvious choke points in terms of streaming data from disk to CPU, then across the interconnect layer. These bottlenecks can exist all over the system: the storage bus (FC, SATA, SAS, SCSI), the storage controllers, the PCI buses, system memory itself, and in the way that software chooses to access the storage. It can be tricky to create a system that dodges all of them. The disks: A single 7,200 rpm 750 GB SATA drive can sustain about 75 MB/sec sequential reads at the outer cylinders, and somewhat less on the inner parts. The storage interconnect: We are using Serial Attached SCSI (SAS) to connect our SATA drives to our systems. SAS is built on full-duplex 3 Gbit/sec lanes, which can be either point-to-point (i.e. dedicated to a single drive), or can be shared by multiple drives via SAS expanders, which behave much like network switches. Prior parallel SCSI standards like Ultra320 accommodated only expensive native SCSI drives, which are great for IOPS-driven applications, but are not as compelling for petascale, sequentially-accessed data sets. In addition to supporting native SAS/SCSI devices, SAS also supports SATA drives, by adopting a physical layer compatible with SATA, and by including a Serial ATA Tunneling Protocol within the SAS protocol. For large, fast, potentially low-budget storage applications, SATA over SAS is a terrific compromise between enterprise-class FC and SCSI storage, and the inexpensive but fragile SATA bricks which are particularly ubiquitous in research circles. The SCSI protocol itself operates with a 25% bus overhead. So for a 3 Gbit/sec SAS lane, the real-world sustainable throughput is about 225 MB/sec. The Serial ATA Tunneling Protocol introduces an additional 20% overhead, so the real-world SAS-lane throughput is about 180 MB/s when using SATA drives. The disk enclosures: Each Dell MD1000 15-disk enclosure uses a single SAS 4x connection. 4x is a bundle of four 3 Gbit/sec lanes, carried externally over a standard Infiniband-like cable with Infiniband-like connectors. This 12 Gbit/sec connection to the controller is very nice relative to common 4 Gbit/sec FC interconnects. But with SATA drives, the actual sustainable throughput over this 12 Gbit/sec connection is 720 MB/sec. Thus we have already introduced a moderate bottleneck relative to the ideal ~1100 MB/sec throughput of our 15 750 MB/sec drives. For throughput purposes, only about 10 drives are needed to saturate an MD1000 enclosures SAS backplane. The disk controllers: The LSI Logic based Dell PERC6/E controller has dual 4x SAS channels, and has a feature set which is common among contemporary RAID controllers. Why do we go to the trouble and the expense of using one controller per disk enclosure when we could easily attach one dedicated 4x channel to each enclosure using a single controller? Our tests show that the PERC6 controllers themselves saturate at about 800 MB/sec, so to gain additional throughput as we add more drives, we need to add more controllers. It is convenient that a single controller is so closely matched to a SATA-populated enclosure. The PCI and memory busses: The Dell 2950 servers have two x8 PCI Express connections and one x4 connection, rated at 2000 MB/sec and 1000 MB/s half-duplex speeds respectively. We can safely use the x4 connection for one of our PERC6 controllers since we expect no more than 720 MB/s from these. The 2000 MB/sec-each x8 connections are plenty for one of the PERC6 controllers, and just enough for our 20 Gbit/sec DDR Infiniband HCAs. Our basic tests suggest that the 2950 servers can read from memory at 5700 MB/sec, write at 4100 MB/sec, and copy at 2300 MB/sec. This is a pretty good match to our 1440 MB/sec of disk bandwidth and 2000 MB/sec Infiniband bandwidth, though in the ideal case with every component performing flat-out, the system backplane itself could potentially slow us down a bit. Test methodology: We use a combination of Jim Grays MemSpeed tool, and SQLIO [17]. MemSpeed measures system memory performance itself, along with basic buffered and unbuffered sequential disk performance. SQLIO can perform various IO performance tests using IO operations that resemble what SQL Servers. Using SQLIO, we typically test sequential reads and writes, and random IOPS, but were most concerned with sequential read performance. Performance measurements presented here are typically based on SQLIOs sequential read test, using 128 KB requests, one thread per system processor, and 32-deep requests per thread. We believe that this resembles the typical table scan behavior of SQL Servers Enterprise Edition. We find that the IO speeds that we measure with SQLIO are very good predictors for SQL Servers real-world IO performance. In Figure 3, we present our measurements of the saturation points of various components of the GrayWulfs IO system. The labels on the plots designate the number of controllers, the number of disk boxes, and the number of SAS lanes for each experiment. The 1C-1B-2S plot shows a pair of 3 Gbit/sec SAS lanes saturating near the expected 360 MB/sec mark. 1C-1B-4S shows the full 4x SAS connection of one of the MD1000 disk boxes saturating at the expected 720 MB/sec. 1C-2B-8S demonstrates that the PERC6 controller saturates at just under 1 GB/sec. 2C-2B-8S shows the performance of the actual Tier 1 GrayWulf nodes, right at twice the 1C-1B-4S performance. The full cluster contains 96 of the 720 MB/sec PERC6/MD1000 building blocks. This translates to an aggregate low-level throughput of about 70 GB/sec. Even though the bandwidth of the interconnect is slightly below that of the disk subsystem, we do not regard this as a major bottleneck, since in our typical applications the data is first filtered and/or aggregated, before it is sent across the network for further stream aggregation. This filtering operation will result in a reduction of the data volume to be sent across the network (for most scenarios) thus a factor of 2 lower network throughput compared to the disk IO is quite tolerable. The other factor to note is that for our science application the relevant calculations take place at the backplanes of the individual servers, and the higher level aggregation requires a much lower bandwidth at the upper tiers. 4.4. Monitoring Tools The full-scale GrayWulf system is rather complex, with many components performing tasks in parallel. We need a detailed performance monitoring subsystem that can track and quantitatively measure the behavior of the hardware. We need the performance data in several different contexts: track and monitor the status of computer and network hardware in the traditional sense as a tool to help design and tune individual SQL queries, monitor level of parallelism track the status of long-running queries, particularly those that are heavy consumers of CPU, disk, or network resources in one or more of the GrayWulf machines The performance data are acquired both from the well-known PerfMon (Windows Performance Data Helper) counters and from selected SQL Server Dynamic Management Views (DMVs). To understand the resource utilization of different long-running GrayWulf queries, it is useful to be able to relate DMV performance observations of SQL Server objects such as filegroups with PerfMon observations of per-processor CPU utilization and logical disk volume IO. Performance data for SQL queries are gathered by a C# program that monitors SQL Trace events and samples performance counters on one or more SQL Servers. The data is aggregated in a SQL database, where performance data is associated with individual SQL queries. This part of the monitoring represented a particular challenge in a parallel environment, since SQL Server does not provide an easy mechanism to follow process identifiers for remote subqueries. Data gathering is limited to interesting SQL queries, which are annotated by specially-formatted SQL comments whose contents are also recorded in the database. Reference Applications We have several reference applications, each corresponding to a different kind of data layout, and thus a different access pattern. These range from computational fluid dynamics to astronomy, each consisting of datasets close to or exceeding 100TB. 5.1. Immersive Turbulence The first application is in computational fluid dynamic, CFD, to analyze large hydrodynamic simulations of turbulent flow. The state-of-the-art simulations have spatial resolutions of 40963 and consist of hundreds if not thousands of timesteps. While current supercomputers can easily run these simulations it is becoming increasingly difficult to perform subsequent analyses of the results. Each timestep over such a spatial resolution can be close to a terabyte. Storing the data from all timesteps requires a storage facility reaching hundreds of terabytes. Any analysis of the data requires the users to analyze these data sets, which requires accessing the same compute/storage facility. As the cutting edge simulations become ever larger, fewer and fewer scientists can participate in the subsequent analysis. A new paradigm is needed, where a much broader class of users can perform analyses of such data sets. A typical scenario is that scientists want to inject a number of particles (5,000-50,000) into the simulation and follow their trajectories. Since many of the CFD simulations are performed in Fourier space, over a regular grid, no labeled particles exist in the output data. At JHU we have developed a new paradigm to interact with such data sets using a web-services interface [18]. A large number of timesteps are stored in the database, organized along a convenient three-dimensional spatial index based on a space-filling curve (Peano-Hilbert, or z-transform). The disk layout closely preserves the spatial proximity of grid cells, making disk access of a coherent region more sequential. The data for each timestep is simply sliced across N servers, shown as scenario (a) on Figure 4. The slicing is done along a partitioning key derived from the space filling curve. Spatial and temporal interpolation functions are implemented inside the database that can compute the velocity field at an arbitrary spatial and temporal coordinate. A scientist with a laptop can insert thousands of particles into the simulation by requesting the velocity filed at those locations. Given the velocity values, the laptop can then integrate the particles forward, and again request the velocities at the updated location and so on. The resulting trajectories of the particles have been integrated on the laptop, but they correspond to the velocity field inside the simulation spanning hundreds of terabytes. This is digital equivalent of launching sensors into a vortex of a tornado, like the scientists in the movie Twister. This computing model has been proven extremely successful; we have so far ingested a 10243 simulation into a prototype SQL Server cluster, and created the above mentioned interpolating functions configured as a TVF (table valued function) inside the database[19]. The data has been made publicly available. We also created a Fortran(!) harness to call the web service, since most of the CFD community is still using that language. 5.2. SkyQuery The SkyQuery[20] service has been originally created as part of the National Virtual Observatory. It is a universal web services based federation tool, performing cross-matches (fuzzy geospatial joins) over large astronomy data sets. It has been very successful, but has a major limitation. It is very good in handling small areas of the sky, or small user-defined data sets. But as soon as a user requests a cross-match over the whole sky, involving the largest data sets, generating hundreds of millions of rows, its efficiency rapidly deteriorates, due to the slow wide area connections. Co-locating the data from the largest few sky surveys on the same server farm will give a dramatic performance improvement. In this case the cross-match queries are running on the backplane of the database. We have created a zone-based parallel algorithm that can perform such spatial cross-matches in the database[21] extremely fast. This algorithm has also been shown to run efficiently over a cluster of databases. We can perform a match between two datasets (2MASS, 400M objects and USNOB, 1B objects) in less than 2 hours on a single server. Our reference application for the GrayWulf is running parallel queries, and merging the result set, using a paradigm similar to the MapReduce algorithm[8]. Making use of threads and multiple servers we believe that on the JHU cluster can achieve a 20-fold speedup, yielding a result in a few minutes instead of a few hours. We use our spatial algorithms to compute the common sky area of the intersecting survey footprints then split this area equally among the participating servers, and include this additional spatial clause in each instance of the parallel queries for an optimal load balancing. The data layout in this case is a simple N-way replication of the data, as shown as part (b) on Figure 4. The relevant database that contains all the catalogs is about 5TB, thus a 20-way replication is still manageable. The different query streams will be aggregated on one of the Tier 3 nodes. 5.3. Pan-STARRS The Pan-STARRS project[4] is a large astronomical survey, that will use a special telescope in Hawaii with a 1.4 gigapixel camera to sample the sky over a period of 4 years. The large field of view and the relatively short exposures will enable the telescope to cover three quarters of the sky 4 times per year, in 5 optical colors. This will result in more than a petabyte of images per year. The images will then be processed through an image segmentation pipeline that will identify individual detections, at the rate of 100 million detections per night. These detections will be associated with physical objects on the sky and loaded into the projects database for further analysis and processing. The database will contain over 5 billion objects and well over 100 billion detections. The projected size of the database is 30 terabytes by the end of the first year, growing to 80 terabytes by the end of year 4. Expecting that most of the user queries will be ran against the physical object, it is natural to consider a hierarchical data layout, shown of section (c) on Figure 4. The star schema of the database naturally provides a framework for such an organization. The top level of the hierarchy contains the objects, which are logically partitioned into N segments, but they physically stored on one of the Tier 2 servers. The corresponding detections (much larger in cardinality) are then sliced among the N servers in the lowest Tier (A,B, etc). Comparisons to Other Architectures In this section we would like to consider several well-studied architectures for scientific High Performance Computing and calculate their Amdahl numbers for comparison. The Amdahl RAM number is calculated by dividing the total memory in Gbytes with the aggregate instruction cycles in units of GIPS (1000 MIPS). The Amdahl IO number is computed by dividing the aggregate sequential IO speed of the system in Gbits/sec by the GIPS value. A ratio close to 1 indicates a balanced system in the Amdahl sense. We consider first a typical University BeoWulf cluster, consisting of 50 3GHz dual-core machines, each with 4GB of memory and one SATA disk drive with 60MB/sec. Next, we consider a typical desktop used by the average scientist, doing his/her own data analysis. Today such a machine has 2 3GHz CPUs, 4GB of memory and 3 SATA disk drives, which provide an aggregate sequential IO of about 150MB/sec, since they all run off the motherboard controller. A virtual machine in a commercial cloud would have a single CPU, say at 3GHz, 4GB RAM, but a lower IO speed of about 30MB/sec per VM instance[7]. Let us consider two hypothetical machines used in todays scientific supercomputing environments. An approximate configuration SC1 for a typical BlueGene-like machine was obtained from the LLNL web pages[22]. The sequential IO performance of an IO-optimized BlueGene/L configuration with 256 IO nodes has been measured to reach 2.6 GB/sec peak[23]. A simple minded scaling this result to the 1664 IO nodes in the LLNL system gives us the hypothetical 16.9 GB/sec figure used in the table for SC1. The other hypothetical supercomputer, SC2, has been modeled on the Cray XT-3 at the Pittsburgh Supercomputer Center. The XT-3 IO bandwidth is currently limited by the PSC Infiniband fabric[24]. We have also attempted to get accurate numbers from several of the large cloud computing companies our efforts have not been successful, unfortunately. Summary The Graywulf IO numbers have been estimated from our single-node measurements of sequential IO performance and our typical reference workloads. Table 2 shows that our GrayWulf architecture excels in aggregate IO performance as well as in the Amdahl IO metric, in some cases well over a factor of 50. It is interesting that the desktop of a data intensive user comes closest to the GrayWulf IO number of 0.5. In this paper we wanted to make a few simple points: Data-intensive scientific computations today require a large sequential IO speed more than anything else. As we consider higher and higher end systems, their IO rate does not keep up with the CPUs. It is possible to build balanced IO intensive systems using commodity components The total cost of the system (exluding the Infiniband) is well under $800K Our system satisfies the criteria in todays data-intensive environment similar to those that made the original BeoWulf idea so successful. Acknowledgements The authors would like to thank Jim Gray for many years of intense collaboration and friendship. Financial support for the GrayWulf cluster hardware was provided by the Gordon and Betty Moore Foundation, Microsoft Research and the Pan-STARRS project. Microsofts SQL Server group, in particular Lubor Kollar and Jose Blakeley has given us enormous help in optimizing the throughput of the database engine. References [1] G. Moore, Cramming more components onto integrated circuits, Electronics Magazine, 38, No 8, 1965. [2] A.S. Szalay, J. Gray, Science in an Exponential World, Nature, 440, pp 23-24, 2006. [3] J.Becla and D.Wang, Lessons Learned from Managing a Petabyte, CIDR 2005 Conference, Asilomar, 2005. [4] Pan-STARRS: Panoramic Survey Telescope and Rapid Response System, http://pan-starrs.ifa.hawaii.edu/ [5] T. Sterling, J. Salmon, D.J. Becker and D.F. Savarese, How to Build a Beowulf: A Guide to the Implementation and Application of PC Clusters, MIT Press, Cambridge, MA, 1999, also http://beowulf.org/ [6] I. Foster and K. Kesselman, The Grid: Blueprint for a New Computing Infrastructure, Morgan Kaufmann, San Francisco, 2004. [7] M. Palankar, A. Iamnitchi, M. Ripeanu and S. Garfinkel,Amazon S3 for Science Grids: a Viable Solution? DADC08 Conf., Boston, MA, June 24 2008. [8] J. Dean and S. Ghemawat, MapReduce: Simplified DataProcessingonLargeClusters, 6th Symposium on Operating System Design and Implementation, San Francisco, 2004. [9] A.R. Thakar, A.S. Szalay, P.Z. Kunszt, J. Gray, The Sloan Digital Sky Survey Science Archive: Migrating a Multi-Terabyte Astronomical Archive from Object to Relational DBMS, Computing in Science and Engineering, V5.5,Sept 2003, IEEE Press. pp. 16-29, 2003. [10] V. Singh, J. Gray, A.R. Thakar, A.S. Szalay, J. Raddick, W. Boroski, S. Lebedeva and B. Yanny, SkyServer Traffic Report The First Five Years, Microsoft Technical Report, MSR-TR-2006-190, 2006. [11] W. OMullane, N. Li, M.A. Nieto-Santisteban, A. Thakar, A.S. Szalay, J. Gray , Batch is back: CasJobs, serving multi-TB data on the Web,, Microsoft Technical Report, MSR-TR-2005-19, 2005. [12] http://en.wikipedia.org/wiki/Amdahl's_law [13] G. Bell, J. Gray, and A.S. Szalay, Petascale Computational Systems: Balanced Cyber-Infrastructure in a Data-Centric World, IEEE Computer, 39, pp 110-113, 2006. [14] Hsu, W.W. and Smith, A.J., Characteristics of IO traffic in personal computer and server workloads, IBM Systems Journal, 42, pp. 347-358, 2003. [15] T. Barclay, W. Chong, J. Gray, TerraServer Bricks A High Availability Cluster Alternative, Microsoft Technical Report, MSR-TR-2004-107, 2004. [16] M. Hiroko, W. Yoshihito, K. Motoyoshi, H. Ryutaro, Performance Evaluation of Socket Direct Protocol on a Large Scale Cluster, EIIC Technical Report, 105 (225), pp. 43-48, 2005. [17] J. Gray, B. Bouma, A. Wonders, Performance of Sun X4500 under Windows and SQL Server 2005, http://research.microsoft.com/~gray/papers/JHU_thumper.doc [18] Y. Li, E. Perlman, M. Wan, Y. Yang, C. Meneveau, R. Burns, S. Chen, G. Eyink and A. Szalay, A public turbulence database and applications to study Lagrangian evolution of velocity increments in turbulence, submitted to J. Comp. Phys, 2008. [19] E. Perlman, R. Burns, Y. Li and C. Meneveau, Data exploration of turbulence simulations using a database cluster, In Proceedings of the Supercomputing Conference (SC07), 2007. [20] T. Budavari,T., T. Malik, A.S. Szalay, A. Thakar, J. Gray, SkyQuery a Prototype Distributed Query Web Service for the Virtual Observatory, Proc. ADASS XII, ASP Conference Series, eds: H.Payne, R.I. Jedrzejewski and R.N.Hook, 295, 31, 2003. [21] J. Gray, M.A. Nieto-Santisteban, A.S. Szalay, The Zones Algorithm for Finding Points-Near-a-Point or Cross-Matching Spatial Datasets, Microsoft Technical Report, MSR-TR-2006-52, 2006. [22]https://computing.llnl.gov/?set=resources&page=SCF_resources#bluegenel [23] H. Yu, R. K. Sahoo, C. Howson, G. Almasi, J. G. Castanos, M. Gupta J. E. Moreira, J. J. Parker, T. E. Engelsiepen, R. Ross, R. Thakur, R. Latham, and W. D. Gropp, "High Performance File I/O for the BlueGene/L Supercomputer," in Proc. of the 12th International Symposium on High-Performance Computer Architecture (HPCA-12), February 2006. [24] Ralph Roskies, private communication The GrayWulf name pays tribute to Jim Gray who has been actively involved in the design principles.      Figure 2. Behavior of SAS lanes showing the effects of the various protocol overheads relative to the idealized bandwidth.  Figure 1. Schematic diagram of three tiers of the GrayWulf architecture. All servers are interconnected through a QLogic Infiniband switch. The aggregate resource numbers are provided for the bottom and the top two tiers, respectively.  Figure 4. Data layouts over the GrayWulf cluster, corresponding to our reference applications. The three scenarios show (a) sliced, (b) replicated and (c) hierarchical data distributions.  Figure 3. Throughput measurements corresponding to different controller, bus, and disk configurations. SystemCPU countGIPS [GHz]RAM [GB]diskIO [MB/s]AmdahlRAMIOBeoWulf10030020030000.670.080Desktop2641500.670.200Cloud VM134301.330.080SC121299215000018600169000.120.001SC220905000826047001.650.008GrayWulf41611071152700001.040.506 Table 2. The two Amdahl numbers characterizing a balanced system are shown for a variety of systems commonly used in scientific computing today. Amdahl numbers close to 1 indicate a balanced architecture. FGH[\ijz{}   _ ` a b ̵̙̙̃{sgjhVUCJUaJhh%CJaJh*,~CJaJh*,~h*,~CJaJh$hpECJaJh:;CJaJhuCJaJh$hh%CJH*aJh$hh%CJaJh$h*,~CJH*aJh$h*,~CJaJh*,~hpE5CJaJhhpE5CJ\^JaJh:r5CJ\^JaJ(GH  ` a L g h i j s  $ & F a$gdc $a$gdpE $`a$gdTnI$a$gd:h^hgdVU $h^ha$gdh%gdh% $ & F%a$gdVU$a$gd*,~gdpE$a$gdpEb    1 3 4 5 H I L M s ڡpchkdhh%0JCJaJ#jhkdhh%CJUaJhh%hh%CJaJhh%CJaJjhh%CJUaJh: :CJaJ#jhkdhVUCJUaJ#jhkdhVUCJUaJhkdhVU0JCJaJjhVUCJUaJ#jhkdhVUCJUaJhVUCJaJ s t u     ) ? A B C Y Z \ ] p |j#jEhkdhh%CJUaJ#jNhkdhh%CJUaJhkdhh%0JCJaJ#j[hkdhh%CJUaJhh%hh%CJaJhh%CJaJjhh%CJUaJhVUCJaJhkdhVU0JCJaJjhVUCJUaJ#jxhkdhVUCJUaJ$    $ % ' ( P Q R e f g h i j r s ̯̝̋xqiaXh:6OJQJhYch=5hYch*,~5 h*,~h*,~h*,~hh%CJaJhVUCJaJ#j hkdhh%CJUaJ#jhkdhh%CJUaJ#jhkdhh%CJUaJhh%hh%CJaJhh%CJaJhkdhh%0JCJaJjhh%CJUaJ#j<hkdhh%CJUaJ!s MNOP𲧣wmcwTH@h^CJaJh^CJOJQJaJh^h^CJOJQJaJhthCJ^JaJhjKCJ^JaJh^h^CJ^JaJhYchYc5 hpE5hYchpE5 h^A5 hHW=5hpEhpEhpEOJQJ#h:hTnI6CJOJQJ_HaJh:;6CJaJh:;0J6CJaJh2h:;6CJaJh:;h:;6CJaJhTnI6CJOJQJ_HaJOA$v:;# e!#$$ $ ` a$gdl $ ` a$gdM$a$gdM $ & Fa$gdM & FgdM $ ` a$gdMza $ ` a$gdth $ ` a$gd%P3<CJOvw{|$<=>@")=CE AH>?$lذȐȠȈhJCJaJhxCJaJh$CJaJh: :CJaJhCJaJh%CJaJhMzaCJaJh2ZCJaJhthCJaJhsCJaJh^CJaJhKCJaJh9ACJaJhraCJaJh\+>CJaJ5ls~  3dhtz.1H}uv|$&1H𮡮hthCJ^JaJhCJ^JaJh )CJ^JaJhMzahMzaCJ^JaJhMzaCJ^JaJh1CJaJhCJaJhsCJaJh: :CJaJh}CJaJhu>CJaJhthCJaJhJCJaJhHW=CJaJ3 FJU&)9;᧙Ꮗypgp_W_WhthCJaJhL;nCJaJhth5CJaJhM5CJaJhb5CJaJ hM5hMCJaJhMCJ^JaJhMzahth6CJ^JaJh}hth6CJ^JaJhth6CJ^JaJhMzahthCJ^JaJhCJ^JaJh )CJ^JaJhthCJ^JaJhMzahMzaCJ^JaJhMzaCJaJ"mnop" # ,!0!d!e!!("@"A"""##b$l$o$x$|$}$$$$$$$$$$$$%%%J%S%b%v%%&$&ص𛋃h"CJaJhW$CJaJh:;CJaJhQCJaJhb5CJaJhc 5CJaJhlhCJaJhVCJaJhCJaJh|+CJaJhMCJaJh: :CJaJh )CJaJhthCJaJhL;nCJaJ3$$$&;)<)S)T)d+,]./r1s1t1111 $ & F a$gdM$a$gd^A & FgdM $ ` a$gdTnI$a$gdTnI $ & Fa$gd1 $`a$gd1 $`a$gdc $a$gdc $ & Fa$gdM$&B&K&L&M&]&g&l&{&~&&&&&&&&&&'/'m':)<)K)S)T)%*R*Z*[*****c+d+*,,,,,,𬣚}uuumumumh"CJaJhsCJaJh6CJaJhh6CJaJhCJaJh15CJaJhb5CJaJh1CJaJh2ZCJaJhVCJaJhfCJaJh|+CJaJhfhQ6CJaJhQCJaJh: :CJaJh"CJaJhW$CJaJ),,,,-\.]./1 1X1Y1a1k1q1r1s1t1u1|1111111111111 2軲}rjbjZbhj-CJaJh&!CJaJhwCJaJhhpECJaJh^ACJaJhb5CJaJh^A5CJaJ h%5 h^A5 hM5hh$5CJaJhc 5CJaJhQh"CJaJhXCJaJhTnICJ_HaJhTnICJaJhsCJaJh"CJaJh: :CJaJh)*CJaJ 112223E3F3Y4*6+6>6?67F:G:^:_:C<>$a$gdc $ & F a$gdc $ ` a$gd: $^a$gdTJ6 $ & Fa$gdtk$a$gdTJ6 $`a$gd: $`a$gd  2222292;222233E3F3G3c3l3u3v3333Y4Z4~44555G5u55555)6*6+6ȼȇȇЇwȇoȇgh^ACJaJhraCJaJh ^CJaJh}CJaJhtkCJaJhtkhTJ66CJaJhTJ66CJaJhtkhtk6CJaJhtkhw6CJaJhTJ6hTJ66CJaJh$CJaJh:CJaJh&!CJaJhwCJaJh: :CJaJhj-CJaJhfCJaJ%+6/60616>6?677777%8&8^88888888888O999999 ::F:G:_::::;;,;ͽͭ~tgthMzah:CJ^JaJh:CJ^JaJh CJ^JaJhMzahMzaCJ^JaJhMzaCJaJhbCJaJhSCJaJh^ACJaJh5CJaJhtkCJaJh ^CJaJh:CJaJhc hc 5CJaJhhc 5CJaJhb5CJaJhc 5CJaJ(,;p;q;{;~;;;;;;;h=k=}=~==>>??$?0?5???????????????鴪ՠՠՠՠzqh`h}LOCJaJh$5CJaJhTJ65CJaJhb5CJaJhl5CJaJ hl5 hb5h5CJ^JaJh&!CJ^JaJh: :CJ^JaJh ^CJ^JaJhTnICJ^JaJhMzah CJ^JaJh CJ^JaJhCJ^JaJh:CJ^JaJhMzahMzaCJ^JaJ$>???????@@@@A+ANAOAAABB $`a$gdU$a$gd\  & Fgd: $`a$gd}LOgdl & Fgdl$a$gdc $a$gd5 $ ` a$gdMza??&@'@+@A@L@X@{@@@@@@@AUAhAoAwAAAAAAAAAAAAABBBB7BGBMBSBTB\Bͽմ͂zrh1DCJaJh{YCJaJh<CJaJhbCJaJhlCJaJhlhb5h5CJaJhb5CJaJh}LO5CJaJhECJaJhsCJaJhDbCJaJh\ CJaJh CJaJh$CJaJh}LOCJaJhUh}LOCJaJ)\BcBqBrB|BBBBBBBB C C}CCCCCCCCCCCCCCCCCC DDDD?DKDRDDDDDDDDDDDDEPE^EiEjEnEpEF GGGGGGиаhE CJaJhTCJaJhJCJaJh1DCJaJhZ CJaJhbCJaJhwCJaJhsCJaJhDbCJaJh<CJaJh{YCJaJhMQCJaJ?B~CG+IJJJKK KKK $$Ifa$gda=w $`a$gdU $`a$gdDb G"G$G/G5G:G>GRG`GaGbGcGGGGGGGGG0H1HHHHHHH*I+I5IFIKIPITIoIpIIIJJJ0JCJJJJJظȸ𸰠y,ha=wha=w56CJOJQJ\]^JaJha=wCJaJhUCJaJh9ACJaJhMQCJaJh'CJaJh{YCJaJhZ CJaJhJCJaJhTnICJaJhE CJaJhmCJaJh1DCJaJhwCJaJhDbCJaJ/JKK$K7K8K?KCKRKSKZK^KnKoKuKKKKKKKKKKLLLɸɪɪɖ݄rcrcrcrRIh?k5CJaJ h J.hUCJOJQJ^JaJh:5CJOJQJ^JaJ#hT[Jh?k5CJOJQJ^JaJ#h J.h?k5CJOJQJ^JaJ&ha=wha=w5CJOJQJ\^JaJhmCJOJQJ^JaJ ha=wha=wCJOJQJ^JaJ&ha=wha=w6CJOJQJ]^JaJha=wha=wCJ^JaJ)ha=wha=w6CJOJQJ\]^JaJKKkd $$Iflֈj |L'''.'|'h' <V644 labp<K$K)K+K.K4K7K $$Ifa$gda=w $Ifgda=w7K8K?KDKGKA8,, $$Ifa$gda=w $Ifgda=wkd $$Iflֈj |L'''.'|'h'V644 labp<GKJKPKRKSK5kd $$Iflֈj |L'''.'|'h'V644 labp< $$Ifa$gda=wSKZK_KbKfKlKnK $$Ifa$gda=w $Ifgda=wnKoKuKwK{KA88, $$Ifa$gda=w $Ifgda=wkd$$Iflֈj |L'''.'|'h'V644 labp<{KKKKKL5-$a$gdc kdH$$Iflֈj |L'''.'|'h'V644 labp< $$Ifa$gda=wLLM{OPP.P/PRRVW\Z\vvv $`a$gd7M$Eƀ`jƦ`a$gd7 $`a$gdU$a$gdgs$a$gdmh $`a$gdmh$a$gdc LBLULYLZLqLLLLLLLLLLMM4MOMMMMMNNNNNNNNN!O1O2OzO{O|OOOOOOOOOPݵݕ{{{h|'CJaJ"jh J.CJUaJmHnHuh(- CJaJh7CJaJh: :CJaJhECJaJhNCJaJhv,CJaJh CJaJh9ACJaJh)'CJaJhTnICJaJh~Dh~DCJaJh~DCJaJhmhCJaJ.PPPPPP P$P-P.P@P^PPPQQQQQRR R!R_RaRmRRRRRRRRBSDSSwogo_w_hMCJaJhXCJaJh1DCJaJhTnICJaJhaiCJaJhdCJaJhd5CJaJhF CJaJhYOCJaJhUCJaJh CJaJhgsCJaJhlCJaJha=w5CJaJh5CJaJhc 5CJaJhgs5CJaJhl5CJaJhmhhmhCJaJ#S(T)TTUUUVVVVVVV#W%WWWW:Xh(CJaJjh1%UmHnHuh7CJaJhDCJaJh1CJaJh1DCJaJh|gCJaJ+rZsZ \\h\\\\\\\J]L]^^e^g^|^^^^^^^^+_-_M_O_____ `(`3`8`9`:`;`<`ƾ辶趮ƥyqh: :CJaJhE.CJaJhhCJaJhg6CJaJhgCJaJhw?CJaJhw?5CJaJhl?CJaJh||CJaJhu?CJaJhCJaJh n5CJaJh5CJaJh CJaJhTnICJaJh CJaJhQ\aCJaJ(\_9ce]hAiBiXiYivjj'kk $ & F a$gdX $ & F a$gd|n $ ` a$gdm $`a$gdlgd5 $`a$gdU $`a$gd? $`a$gd7 <`=`````aazbbbb)c+c8c9ccccccccbddddddeeebedeeeeeff$fDfNfź岪zzrjhYCJaJh CJaJh_|CJaJh wbCJaJh(!CJaJh4CJaJhTnICJaJhK_CJaJh{CJaJhngCJaJhghw?CJaJh'^CJaJhRnCJaJheCJaJhNCJaJhmCJaJhgCJaJhhgCJaJ)Nfbfdfefkfmfnfff6g9gcg}g1hIhLhRh]huh~hhAiBiEiWiXiYiGjWjvjjjjjʩrh[h[hph|nCJ^JaJh|nCJ^JaJhmhmCJ^JaJhmCJ^JaJh2=CJaJhh55CJaJhhl5CJaJh5CJaJhYCJaJhXCJaJhk,CJaJhmhCJaJh_|CJaJh^Bh CJaJhaih^BCJaJhaiCJaJh CJaJ!jk&kkkkkkk lll-lllXmwmmmmmmm"nonnn1o9oQooooo pp_ppqq"qqһ߱󢝘}th 5CJaJ h CJ h5CJ hCJhhCJ h5 hl5 h5h J.CJ^JaJh<CJ^JaJhph J.CJ^JaJhmCJ^JaJhphwGVCJ^JaJhwGVCJ^JaJh|nCJ^JaJhphmCJ^JaJ)kmooo ppqq"q$qt?pyȾȳ~vnah6hlCJ^JaJh \CJaJhf-CJaJhhCJaJhraCJaJhlCJaJ hq5 hqCJ hraCJ hTnICJ h!CJ h^^5 hCJh^^CJ^JaJ h^^CJh$CJ^JaJh1%CJ^JaJhoP.CJ^JaJhCJ^JaJhXCJ^JaJ#ȗӗԗ=i3GÝ\Bߡ֢F$a$gdEV$a$gd2=$a$gdth$a$gdl & Fgdl & Fgdqؘ٘ژۘGhjkl245*׼ׯyyynnfhthCJaJheVahthCJaJh )hlCJaJh )hlCJ^JaJh )hl6CJ]^JaJh )hlCJ]^JaJhQhlCJ^JaJh: :CJaJhj/hlCJaJhhCJaJhlCJaJhs-CJ^JaJh6hlCJ^JaJhlCJ^JaJ'*+GHIJNWZbdeo{͛ۛܛɜ^zreWh )hlCJ\^JaJh )hlCJ^JaJh|+CJaJhlCJaJh )hl6CJ]aJh: :CJaJh )hlCJaJh )hA_CJaJhA_CJaJhA_CJ\^JaJh )hA_CJ\^JaJh )hA_CJ^JaJh: :CJ^JaJhA_CJ^JaJhthCJaJh*,~CJaJ"ĝŝƝ̞ОԞ՞7D]^_ǟڟAȽ{s{kch yCJaJhQCJaJhECJaJhh:CJaJh2=CJaJhX1NCJaJh ^CJaJh )hl6CJaJh)CJaJhoP.hlCJaJhoP.hoP.CJaJhoP.CJaJhj-CJaJh: :CJaJh)*CJaJh )hlCJaJh )hlCJ^JaJ&ABDEPSg}ޡߡ4@Aբ֢ע٢ۢޢշ𪠪ynh2=CJ]^JaJh2=B*CJaJph!hh2=B*CJ^JaJphh2=CJ^JaJh: :CJ^JaJhh2=CJ^JaJh4hE.CJaJh*,~CJaJh4hE.CJaJhE.CJaJh#hE.CJaJh: :CJaJhE.CJaJh:CJaJ' LMOR~ţȣ̣ͣΣϣأ٣ !"vx{@DEFûðððððððððäÍûzzzzhdCJaJhdhdCJaJhdhlCJaJh!hEVCJ\aJh!hEVCJ]aJh!hEVCJaJh: :CJaJhEVCJaJhh2=6CJ^JaJhh2=CJ^JaJh2=CJ]^JaJhh2=CJ]^JaJ/FGHIJ!xyz{}~ǽǦѽћsogogogogo\jhIc3hTnIUjhrSUhrShn\h:;6CJaJhn\h:;CJaJh:; h:;0J h-h:rheVaheVaCJaJh]7h]7CJ^JaJh: :CJ^JaJh^^CJ^JaJh]7CJ^JaJhCJ^JaJhhCJ^JaJh: :CJaJhEVCJaJhCJaJ!Fyz|}$  ] ^ a$gd J.$a$gd J.$a$gdIc3$a$gdTnIgd9Agd:;$a$gd v$a$gdeVa$a$gdųq_TB0#hT[Jhq5CJOJQJ^JaJ#hqhq5CJOJQJ^JaJj;hjghuU#hT[JhX5CJOJQJ^JaJ#hh5CJOJQJ^JaJ#hhX5CJOJQJ^JaJhhXCJaJj~hhuCJUaJhM#hT[Jh J.5CJOJQJ^JaJh J.h J.CJaJj.h J.h J.CJUaJh##hIc3hIc35CJOJQJ^JaJhIc3hIc3CJaJ$.9CQXM$$EƀN{Ifa$gdqp$a$gdq$a$gdX$a$gdX $XYZ^efnƪǪ˪=>?E««žyyyyyyjX#hJ phJ p5CJOJQJ^JaJhJ p5CJOJQJ^JaJ hqphqpCJOJQJ^JaJ&hqphqp6CJOJQJ]^JaJhqphqpCJ^JaJ,hqphqp56CJOJQJ\]^JaJ,hqphqp56CJOJQJ\]^JaJhhqh5CJOJQJ^JaJhq5CJOJQJ^JaJXYkd$$IfTl4ֈ <` @`&``$``H' <644 lagp<TYZ[\]^befnffaFf`M$$EƀN{Ifa$gdqpJ$EƀN{Ifgdqp nrvzM$$EƀN{Ifa$gdqp%kdp$$IfTl֞ <` @'''$''H''644 lagpFTffffffM$$EƀN{Ifa$gdqpJ$EƀN{Ifgdqp%kd $$IfTl֞ <` @'''$''H''644 lagpFTƪffffffM$$EƀN{Ifa$gdqpJ$EƀN{IfgdqpƪǪ%kd2 $$IfTl֞ <` @'''$''H''644 lagpFTǪ˪Ҫ٪ߪffffffM$$EƀN{Ifa$gdqpJ$EƀN{Ifgdqp%kd $$IfTl֞ <` @'''$''H''644 lagpFT ffffffM$$EƀN{Ifa$gdqpJ$EƀN{Ifgdqp%kd $$IfTl֞ <` @'''$''H''644 lagpFT"',27=ffffffM$$EƀN{Ifa$gdqpJ$EƀN{Ifgdqp=>?%$a$gdqpkdU$$IfTl֞ <` @'''$''H''644 lagpFTEFG  h-h:rhmZhJ phJ p5CJOJQJ^JaJ#hJ phJ p5CJOJQJ^JaJh5CJOJQJ^JaJ? $a$gd v$]^a$gdJ p6&P1h:pO/ =!"#$V% O 0&P1h:pO/ =!"#$V% P0 DyK szalay@jhu.eduyK Dmailto:szalay@jhu.eduyX;H,]ą'cDyK gbell@microsoft.comyK Nmailto:gbell@microsoft.comyX;H,]ą'cDyK  jvv@jhu.eduyK >mailto:jvv@jhu.eduyX;H,]ą'cDyK alainna@pha.jhu.eduyK Nmailto:alainna@pha.jhu.eduyX;H,]ą'cDyK randal@cs.jhu.eduyK Jmailto:randal@cs.jhu.eduyX;H,]ą'cDyK dan.fay@microsoft.comyK Rmailto:dan.fay@microsoft.comyX;H,]ą'cDyK heasley@ifa.hawaii.eduyK Tmailto:heasley@ifa.hawaii.eduyX;H,]ą'cDyK tony.hey@microsoft.comyK Tmailto:tony.hey@microsoft.comyX;H,]ą'cDyK nieto@pha.jhu.eduyK Jmailto:nieto@pha.jhu.eduyX;H,]ą'cDyK thakar@jhu.eduyK Dmailto:thakar@jhu.eduyX;H,]ą'cDyK vaningen@windows.microsoft.comyK dmailto:vaningen@windows.microsoft.comyX;H,]ą'cDyK rwilton@pha.jhu.eduyK Nmailto:rwilton@pha.jhu.eduyX;H,]ą'c$$Ifb!vh555.5|5h5#v#v#v.#v|#vh#v:V l <V6,555.5|5h59/  /  / / abp<;$$Ifb!vh555.5|5h5#v#v#v.#v|#vh#v:V lV6,555.5|5h59/ / / /  abp<;$$Ifb!vh555.5|5h5#v#v#v.#v|#vh#v:V lV6,555.5|5h59/ / / /  abp<I$$Ifb!vh555.5|5h5#v#v#v.#v|#vh#v:V lV6,555.5|5h59/  / / / / abp<;$$Ifb!vh555.5|5h5#v#v#v.#v|#vh#v:V lV6,555.5|5h59/ / / /  abp<Dd a5~$"6  3 A2XJv*1Fip) `!hXJv*1FirfZ6x ՙoUϣgif S4bԀEXN6;&ƠAMv]{Q׳FVԮD5RztuOhU_W}uV5("fŢ""c JŨ̓(IH7;D*PW\V 1Er! hkǔG q3'Fj&J  -G?D۪m{)Xn;!O( E=>2HY(֋XtbrAe{ebC撷ٯ"mƊRȣU.s]?#Qv;{5=D5OJ<>ۅ[\BerѶAq^ O[zqܶ4gg~:vma7&0޵XODw/|ǭ 0PWR`+X\m9.೧:-xBeB$9,?PFP,DXav-tu6'EOmBuyGr62ԍW(9W\r O:_RO:\~z>`ּ|Pz>`>i09KxS(K,R^p,gOqfafx>3u*,C3~C{8W`Y =t0|Qz>`%"TTP3x?E=eكh?=e ɼN G}gu"۹T yݙ.Q  RЗ|X8V#J01}]a_D,Mm-L/y^ުV@u`Yz_ LךX_vr@Qaw]7; \0yS^.97\ )/iY#N,FTĆ1#|gr߅~}zy|eç=oNiw^^.z+:3O//is:_ ǸVv>K\[z%ϼ^l%&O7w;,? }L&v ޳9+?dz!R,jJF~}8 qr1}n37i\V9?+ۓ]3'W}p#SGV#_#ViCo"R mOOm7г&J']z~\ {ljs:^vIWpݳuT8H$Ϯč5]M;(F%3yƳ7W?dݦiy9&\1g،4.z_WŒ`[k2کf<] Y-ljG}sbߡϟ%PyBư_9Ƥpx?nl^%?wCCHp'*jO8X6;W@q8rxe&7H!-)˄E_(tà) ? X;?зt=@|enn-$'hx4snٽuI7Lw,񦀷>wO oXNJ5m7mз-nQ[닩LEuڸxԪ%[[``z}s`toߙ]w< 3|//Le0dƅ@7f`ƱY)x!2o9^ <EO &wL1įd.1$ k /573tgPγd{_hg:KMv;c#4X56޻>vY8{ SM*ۮn-ZҘzҌwhly޻E>Sxn4/2nQ9_(3Un~ }}Tڥח1K}x?4.G:Qᥡs/F+>uwJ+FNxmhcu([r~\˘_ ά1Ra.}쎞\$s@EhDX1 pc' %-%\DaR]LՇKYZb)qKKK9J# a),%QJhKqUr2"7rKq"(a)+,u0W7WRYQ&17f)!a)sZJ0"JX!"aln) ᄥTZGL\-a,i)FqIX*K33K9"MXʚl"KXJ8Ky5R5w',Eh)#O$,! ՄtLe1"ړ K1}WKSK\g)D\\GX뉸z`*a)t4" K8W?FX閲\"nNX軙D\?'"3 Kd̛CKf!"~8q.W6R5R&"V#,e?1KiY`)qea)@qt5q[wS{n 2^A<Fu^_&O }/^?΁wl2޻; +k|N_; CS蝠yimM_>0p5}uhuu8-[P'd_>6g_g6_A%nG.}o6lWLϜ\?tĄχOj>o|qF@PǙp|;wYR^Eo۞y-ߙO,paDDań`$+`}E0`;$R€m"40`UF hm0`C?;0`7W$ LKD2lX"z€m;4F և:0`7 OD/A %G`mLD" Uф`}1`fl"l€m wOD50`j€m~scR"l2a6Xol)Z€m\Ʀl zF C' -l軙`Y`}Dt€mPa6XomLD7% yD `ݲ 0`߁6# f"€m|) kvܵypU`٬!;oPC(>kAykA޶}Y +p >S 0ajC.&"S[ZbjqĖvEĎ(u0D)u LM4ĴajA*#"\njQ"k.'L0Hp0:6WVQ&+cLmonBDb{cjjÈHlPA?uFچv"00JSJS3FڶQD$f$LmfMDb3S;";805cMmD$f,ajL]&gL D$vx6􉄩9D$B5ajLmd"D f ٓ S1}WKDb;6bSHlraj'"S S[*0_S>ԶKDbۧ6~n&Af3 Sd͛CDb&LCC D$Զ6ƹD$0y4XƟ1x[o!fo_ݴ?kC]MMO4ԏo.=+pMuK]ooeX %tg0QDh ^]+qmw-*F3W Ʊ!=gľ;'Yw;<,}Mrd񓩏wRn'?xwмט-^>'`ԥSwS0ߔ=v0KQoepIε\Vj OJ0>6?4xhTcxqoSW.Z(EsbMK4^zcbS{eKcKIrY׬} fd%] @!-ŕta|H/b{׼_qWW$9D ZZjNqҼmnzmFR7.#M~P8sn4?GL䞙{)$uJ┺@'25g7dmK\7inẗc\)/_2[[d{}M WmhhTrљk/IEGR]{_}m!]^f 0II !gQLvD3IIֳ"geLvuIN ndPmv2?G2IIǙ$$ 3hpA=}pIN Lv`PŠ s"& ;ɺ\$a'Yw [Lv`'d{EqnDžlka[̗7ۅlbwG"?zsoKJNv[z0;#s7]~joG]6^پv}D+Z{k.\߂3pd^-ޫ1\)zذmַ fcl6c f xa۬7= cöY1\)ͺdl6 a۬gfƆm\ذm`cöYo] f ذmֻ gl6eƆm%8ذm gl6`cöYO:cöYԟ9ذmZƆm ndPmv2?s Wԟa۬36lpA8 1\PƆm ÕYà @3h3?s Wԟa۬[ f3?s Wԟa۬o;i*w {7.I]jcAYv0:lna1:l3%F>`t:'[g-KuF#zgt:٠a/1:lu;Xu֛AYow2:lnp/3u҃1F>^ft:_#[g8ʠLQF F1:lD3:llA:"pYwkz3AɠLu8Ƞ aPNzet:PTuփA3h3?S'κ\u`=3up#Y v2:lu_B_| E)49s>?3R&L8؝ؾ~Lk7m]{u} ?EQݿJooW>Mv^}T~"vW~X߁XsQ.Jҧ1r/$ 8{FW%6&^Y{m͘J٘1|+}hUݷkK.A-׊3~(I\<I,ۉDd 4  0  # A"WS2A"MHNvj:3.) @=+S2A"MHNvj:,,Qw6HxeXU[0iAa "lJIiK@i))i.Iix:s}5k9#X?c>M@qA &eJ@4ҏe@  46#CP HGr@px-اՅt _QoS񯵐zGw𖕒D3G@;)kp^p>-McZ~˾P|xpd3§ٿi`oLxV_-?Mt.S?d?rrǘ4h}ss&7M{5~)/}m3_ ܮOz[X {_ \QH⟗OjԿneE>E~NdE˲)-O6xw?0[?cG/ ̹K'lp>~G6vc_s]+św@B-{bs:vk'`Y?l[C# 3oe4qK#р_ÏH?<; wnǩ? Y߹A 3mh-?.Ÿ-@ #կ=7WKW |+ ;98J P;|&qI >6g}J@:=:C8,T<@8 9-#a y*0 㖗Ps>ZxrDUQQ;3 c2-AwB  | G99čGblx.'M[˘VE"0;DJmxI~Iߤzuh8\<vD300+o.S+ZLpSGfmk_2gߙVjpIhl_'%)Ns yqc^w*f4s%Tr:o? ۈ%Iz>9i9jgؖTC(.ӭ_Xڶ^҆IzW~[=_uuBC."ܛ/Z :2o`Aq;Wb9 _(+g oim~0KT~ڒiQb\k~X c+=w }HApGƓJ ^Ūx6GKQLOx>~aza򤮁˞ּ7_D̎nM'֍6BKk?}:Hha')UcaQPblJI+399:vtt` 4#z'?6΄8-dzo#A6nnP_NYO] cD` )"ǂ)][{_+.\\ĸEZ8:kY{wd O/D!67jO}4`!iajmrǿNQnTtef嗔U-< v:LA6 hT!!aT#" =LuZ)Cn5_Pӆ*vr_%~2AYLW';tl'p~~'d`9KSLxf+^e|(Tä?=MaoehWhOYj$vևs+ÐW?b}69b}TX6Q+haEcTE+%QLRyj |Y[ٙlΐܙOWw% ׵xkZmOO[56r7'nֵ. 銛/oK>hVWOW6AIK#+p OG;5f\uu GH4Kw gbt.h9g@1xe"Uz{^R$=2asVz2d22ϱL?3gR'ӂw#,kl*Ͱ [:[{MyžltuzX HPS;摔,N)Qz@RȾo'<ՓI O% Pi,cS <՟3Qi)neb`[2],wzu5O#q # -OBIK=\|厕:9er'ԾuMg>4{sOҽ1ԩT`vXgI -[-ܢ6NmKJeb8a%!B#fCEoǾh4N=%.)?_Z婭1S!&\g̈lG[V4u>i/v/:d=t'g *:<5ۘjɟ$ %S-9I- z5 &q)ٮ,2' 뤙LJ3 D>>2"KRg+6|jo ʂFdIfw><ګّ/i^.֖֫iCJ-hz!##Jsk4g]fX!b[~gLÔNz~[q2n[hv12?uXhh<"&{IK+9)8;<ٙv97CM~:^8Ew{G #_.ok0괃PyY!L':rҳF{A+: R,#'u<_|x;hX&٭5bL0vİ{%E}#FQWrUIBmw *-6X0IK&-p*g>Z^[ĝ%,,LQJΩ kQőUTd ?b͖c:KnI0ɡgܨD566IPmܼܽʠ0N`nmk#99ip\>/\xۍwFgg?y<)>s7~Ĉ%o-mnR}\jo~Nh~U@rSZz0~yT!|3=ۻ/:b؝#XW9 <E ="!+!~B蓩)jU=%X`:657- (.@BFK\[vvŗre*;yLVCpv, V7`v{#3Wafӳ%uRi'SĽDnBçҹ%wQz֫MۜÎCt#Gm 7n .[E&%&Nm\}* tF4R7oy#Pj:-elٞʳ̴S/i3QGb䃸y7?iNIv+=kKyUqBZ)'Ɇ4o#-*t/::O$: 6*)U#Zu4L_i~R0ZS쬞uN7Klvh}Lگ]=g$z^kaCݛvAE ,z,%ҀP?*g"a_O j_~f89Ֆ}g١#/juhԧ!666dyL@87ξQKH{;.F/kӺ4O28P=(`-ĩ56q=o`s<;U<+־+g|alj&Z=0^>cSc"ܛ_h޻RISC3#p.<,[ õ+dUCQ!6Ns]/Wz)R+je?%`Y3>  =j`pg3鳿@e9j̐`GHQ1.o{>.%zХ_6]Yg5)PI]SaPgbܡ4jW`[߻)cN$ewZQ~ct~e+emEƐ2قyUdŀoB'N0kl68޿׈7޹ȹT_95O׃Ht elŀk/9fV@8ƹYTB| 6 Y]L mܹS?gb]?^ 9o9N>8yyC잘^A-zi̴astTz:55uC\- uhFml?C?Ӭѹ:3pѰ}rq5߂kdDq~^^SzH([vt[TE|fMh嗱d.#ak:=UadjgᎴ{9ey(8bi17&܊jơ+]1  VnC=Kֻ?6 >B Njq5ݓR>~*gMMιΏ62/VpѴ4v(mp Nj|enV/:j\Z+,1k'|ɶj?Q|W2|E\gB}C>pqeOE3A& ZCw}9HS ~㒉["5w v ,:wb^,Fb+Z51G(TلHToNAK8>KQMHOFN[֘ *J#,!1]_ f]KXu:Mó0 kqj򽼌?9s^O8YopP D͹$˵uturi!~F|i(j_(%*X?{1` = u5*u{0cǤF KC̑ѣvCݹJvCeW5t!X !f҃Bk=[ Ec@|$|'C~SR4z0 uO#u^m]lsI/<^<[f⩊Cԑka^0',W,iN!쁇\7u[i$\SpZ'_`RS7ƯׄEPedabxJ u4 b.7pJ2""uMLڏs)))cR}k:^K!?EUݘ|h_,f]==4vrE(jF姐t+;< ]5zE)JGwё,{ݶτHJL-󶹼+ .' |bz_YXlMz9dB,ހSMY:yJK)l{%'ŭu`dLLt}NWن˶ 4;'3S=N4EؓbL;6_O)8?y4Qj:mF,4<:idqW]lmi}(?LWhԳP'JE?=~y'kuޥg[;om)r+pmypٺhEq)37{ӝk輽O|XvV LjNY?"bJ.b7+ڌM(8ˤhe)>q8ikqۚy.v~B{9#WZ2J􉃱wE]z y?>ֈ}}sWz RVScuGFO'pʄS@Ɨ )'FpRӞ|ȈrY N'3UxVnq5gvPz.,geE Z9DMR?=KHN)*YPTyW@i1@ yQ^HtɡN*1IdD9mk~Qݍ${Dɼyڠyyxr8ʶRO qC* $5޾Wɣcc}Kni?~'JP%KwueeInq2hD黫2nU|=d$2͂2՜IzZ2;jFMJ;%ɛÚ̗%MS(n3]Y Pbָyؗ}B'#YSE>#UwK@59!k9d=BA,1ډYi=@TFY>UK-7wu<3R >9?_>T#aN ZQc(K4Zc1˽js}ϗ}k?sr# ^e~EV$ޚOCM_{"W_^شt/STٛ#fBϴj 0pzv*-UbmJ`'f״7Gb)#޼, 6Z@K`\*QU{h懃/UknJI{9;VV9"UlSp(@BÓS=47MDףS@e?.VOڌIDtG5geAE*=cdDP=M4r?D=2@$v딁g?yU)NU\bkD]Z]޿h8W+~~zu/P=e-wM 2"=Cnt<my83}v ޳s -{_~V*abLEG$(-ysGIqH-I'Oe26PB.]]mWx1}~ll,cs*?]yZCXD=x&XrF ĿcBoP6l0+5Z^.1#g-`6u4,VE9*99cPךYY וRݦLa֓n$Ęt !XĈ~(L} vdxn&"fx X`<~R=daKiA<>ƃ'D%z ZG%c%_Hl" im!7}̍&#"= #Z,W&ۛ;JIR`lqjKI^%$Iyu<*R]"̺N 1dz9+rgG5n>gt3JeѺ{eEs,䃬fN-KNf"I*?:٭0N35؏ƾpt {( n9-ڦ&*5(u#Www#Od&];GlAj.Uwmk[SfRA²|+;jsy;HiSG5y[rq?yt' n$} ݌ 6TsQvjxwfvUU3m"z/&?v lT:7t4,P.6+.+K3MXޓZH.3l|=\_7;:]M[V<2,j3*A|OlN4r1G]1EF7$Fbϣ cz׹ Y.B 9߾j}L5ۋV0y.i[ Z梓_2=tpCg<6!1FxřWZh^W$]FffolL 4A691^Hp2S#SyRDt?9n Kaz>;,N\4d?Ƙ [vZGCk1#sNRv5+$zJ!m)G+u{SHlן[[l|ݬ_Y19>6{Ocsk9gR4ܵګ}rOqJI0 (»ݢyhs/MME{-}T>^S/'K6g%qWxU̳!M'C2ܐ#l_ni#Fs F!TR2a:N'"Y&M7o1'~L9VL"=*K>svXb?8$lٞcj|sjeg +Z3*k؃O '蓦n$HZՏcѾo0m x$V(}+zVH 3WEyM [2M4;t ʺ4Șɯ%Dո}\6%Z[Vs,3{"Ak~Lɟ3UCGyBlu2a_s$+$Wߖؖ]iyޝ̎ \?r!FB89]H8`dI.b^|C#LcAP[GRV)?+ |cb ܹYt>83&{^O{~˕rKsTAOP(WȵU9=㖝%r=,˟"MiINFQ j0PnT# +,(qETV(PkUcꕎ6UB6nO=-+HpAN CԭQ{Q.)P_٬¢B`B86A?M_>&;>ZPxb!hqzN1>mHj۫M+Gi5Az9fmџ$|O%[sҹ*i|ԅlŋ*m}Ld DưLU9tDnćqےU~=*S.!Ȱ9G;jdeq?Y5f|knU_'}ˤ8tA3{+gv b$` (ntXқlO&-+JFWBjT_ [y^2f!1>uwa FxJ5[ܽF'}yjV}*5F*Jbs}m8h*Fl@` B+.ڍ|2g}) Nz"f$ .wu"(萺o󟔀]4d^U{QRaG`CBGzzi/D4ڙ} J[+#he@HxB@KəJ 4;|O E0 DLFcEP s&^6XxQP(W,trpDdؒXW5y;j !FE2z:z2i:ž$%gzܭ>Wg› ;ky2%zaTNLy/QQ]kd'ݝ\JwsI̮mdm*+{¦I6 /XؤL8ѩbYD~(=B1A3rhCѶG*㿨3x_&+ B0pWxh5CL-5F4yu2Kj=?h➆!6>NMiyd3`Sw,|:|M+ɔԶHxyx2j$A*CPy_6rd9B쪏^WnÔ]A;#a~gH?m?MmQQ 9 jGkQEiI {g &O{Hn=85)mlbZXalcOF+sZRKc:O?$I kKF6Gҙ SzWjj{g}lDq@P KiϤ?wamB2 (^Bx=MoީcvݩUg`8zD}!f>&oEF?Nl}N"6P9kI5Qbh+: <jrf9iA|9KނC^yQuV)7)?9C`Ɠ㺣4wvȪ j) 2uV8z;w\n3 E]9ujB3sߞ(CB |0]⾈YՂ#^kCVuSO~=d갿'P!=w"P"Emzq i@aB:ȝvR+9DO?R}VdQqk8(//kһoM8kAJR7 8~S'?lͷ(4^ZRZPgKo szq&~ xw4/4Kȷ~EoncJcڕ@g8чA<RA)vF~o[%<yTjx77$9<~տay8(x#ٿW W4PJH}n]?{Σ-WImo>s $<!0N[ +[3pǃcYHHiY`Pc,+x ,CjaxEMo.xJăOrhr9ӮˀeoC!5|,vW}P̍wv oK XwS_O, 1 Hj 4&[W +"A0wNrG|O710m++v%@n zi@> à7t2nv!5O< xc@ +v( 6ƅDc`` +|D:|؊O<<?<~Ǎ] X|q}~xXP?92cلV$FtZTA@;"0R')1ơ? Ьހ?o Po3y'ᾛZ+{g{½Mr$g'19ܯn(!ZV80\W"-ǭgaϠe'M.XTƀ.P0p'5ֶoP0r~\Y'0nSN& 8(9sR#o  败EtZ@>vfg V}x+"^w{ "}0 &Aca``.8'qb;oo EUeh~1ă%hH[h (V`g~ ش6`ͯn.`-`~0ݚ>!d#\n7(m\FS11mZw c) $60o) 6sbIx ZgѨ<|}2&`7;"@Mm'@J@dVm[o;P xmFwZX A x[n Io c(%5VF`5`]mik[Aг  .9Pĉ>@Fs=x= H<rXDDB e4u X H5p>V!u#Sc d5x20N#|*J`( `~VFøJx2J@>2xu@j,d `,h>UJ66ւq6| qK{:&@TZ׸v\Ԁ~:+X!`6P N ́JżE$;rM*hFg2gaR) 3'ƮH;i1o!ʒiEN 1WDp]Wr:mnp:@Vҡ[ لk贎J`K*~`eX e2!`%nP={%K(A`lpM¿ (w.aA৉:%TPpo=~wx78 9n\jw^muBwwN9?שSchT EBdxuT vkF`U+t**U4By7DƱvƱ QS1z/ @jJJp9`?UFi(xGz_ aӰІ2JSOk;ay2 FMJF |y4P?/rxh߸ `X؉Ɖv%\wTՎkk$dP NH => jd@U1H|C8VInX88- ϏJ@ WB PGx =<|{G TU -%|CE`2\<AU(XJPf,9sιsTcJF-ńr= mrsHj~ Xj9K"9r?*~և#L> ;w&5]X"C1 %L.HW'm-Pn1nc\<`S{2Ҷ8S:z"P}To=;y͇"}ܦ,vΟ"\S+$Qg99`3 -li\T(uԄȤc=\3t=F#s'7xXZ̅~h8evoTq >hf$$}WқN"p=^_=:cݕJ=bE6%4L#CJp4#X+hr"tRuvjc6*UViUEnd@QkJПb$V$'q9πJĐHɊ#BJ̇ǰ^~+j!1i$;%9ܴV˕gD,S,dw2mZwy?Gs3]xyusz)ɤ~qbRHHT&!{KbOba"I$஦ :(lmlƬc%:l  )A_mUgIc0xE"${#m0 US#$$~mZ.XM%^,V^R6EN`8YE"a Ie2}򍠯#oR깾;laQgBIQDY8`1e$KJ\+|’4 >XP&im=2 3'_/ NJ"glg>WI_޻Tb$5GZNn1\jIyY'y}%A܂LZTGX䐸[L@_Lb.DMgjMz[zJ<HGJK:W۪2ೡw?uBk/&+jNDranȋEp+9,?I%#C6|/2Hŋ!ر`i$c q -v*b{+`1\ "#[Y,9PPK,%H9ʐŌs7gÜ;nN–#T Q rOI寐nm |}:эO~ sޣRFX7GuԈX@L ?u6urҲ[_Kjt'$_ 5IOwXq/Ɇ2d=if'r?Zx$AZ"a[J:IPMɍd9L8n%.fԌH(r$d%&LbHf#r%b8 O>s!=d@wo-RVF>$b)ڨtOr8fc'}FlKy\8(bwk`B,$x@'@FCH}&mIXq Gh7sA u6T=,ْ2E<ĭ]ڛ$1|H"9C}0#ٷ,WgBJ3T3Dy:\P\&jrUO#|?ȚLIٍd{XО_%9=vl/M;XV?~Ur:ki$TvDꟴ\}s?$.K#WrA#a~(dO ~I| 'o [Ujm!mBFRo1vO2RcD$3'dUvZI5@.a$Ns#!b˄Ն,I&\PQǽelC%KrJ}|(t>P H9%d'*ZjCj된6q>(@Q&0 ,NIKjzoF~Yqc4at2aT W.`L?:^)Q\vܦyC gPfadޚ\ZRmWQ+\4ɣ%AAG V֒ 0Bj PXŖ$d*FhQSb.h j zhbFI هj|[_6Hb\4E,$w#b \`LԭcȖ(p;݊,,E&8%"8*a.X} d& Am-#Amio4tuwn&DJI3Iyc^' ',kpԣs$?)>s%R؆G1d"  '})ȄT+t-#l5a YlRlE)r`8N.$҉x*]#b8 GQͅ/_$u6N"X3HF =֖`'SJ2&)}Q,)1.d 33 !<"c} D~C2jK?)؛\qOTSN2<+'=zxF~vij}wtn3e@Q;v:F 8 a sri=$~kۨ8mN3 yG?4f@6_߽rEm+4;Oޤg&Gr8A #$*ɥ$&BPArF́o'p r4 pDCzEJ*QI|ᚁAe.-%Qgr?D=`VqV*Mkгzöc0݊0N2i[G@X:ne& 1Lxȝ& F{=΋1SK^U幒0@q~Pː~-i=]Q+ާ+ߨr@#} *tĉ!deVƻSpS ԑl I.5Re8 z$C(x%{c;U?rҹ܋xi+$pgjC_8cȩXbh[ŊE)F`|W&z>j9OR IRgdQ 9,&ՆpGGʆu,!@nhrk8g9iHDs.O$r8f+eAdl4Y',;&A2x$:(x*o=&Ry s9pzKlxxT}ڎMe[,CqeYm]߳ 27߱V<=xpĶ3OFm#m-2/2dp5A[ˀc9 %po5&rIYpS-뷵 Knms lv:#uEC_TP3gWxO;&kk e\hq%h\q%h\q%h\q%h\q%h\q%h\q%h\q%h\q%h\q%h\q%h\q%h\q%h\qq/~c\騰-3̤Qe\fS?}AS>|:#{J59L:sғ_ J;{Y>QI9I+7lar熸iG߬0pQŲϿO6ĻrΠ(A u{7٘3L Y;^Eo'%% >xΞ=65BnKO<0ksÒQw,}qaEVD +a?O5"EJ )ym6Ve_No;o~a_RR N>Sq鬖֩}S'>۷.>PSu{ چ%g0aOA/~Pk5"#HYаkQ߿ޛjfwÎ}s_2^+S=nxE{#;U; d˖fܓSsvٽj'h~ͩѳ8pR;^3br|DQӔrs.kO^Z{q}D|r$Ew脨FON[7ebX0Cm ϜxeN/oSژk-{b߷K݋d?˳0홎+,'Xl7 a&_߆1-^dIt.;wIẜ w{Kԫ{5}ȴT3۾~F.98Y& =ͻx,ii8?vu!?~λgoo&R@қ3fG&Fu~q=ܮjo3rM2Cs:#^!.iVr}ݰAcӦ^~qfcU}*r׻5Ǐ7.FC=&'lmZȊWXX>򁏝~rSؙG@c#?[w̭?{Q&jZ5WzGg-޻ݒ˂ܿ$zct4zfͦ}ruTM>3O2|A4hjSW0>͒#O4n 7T'x;7ZUU͟hB?< #o(e5Š6,%>aնlbMV?f ;v~tpzȑ<1 UUR.MZ)t555rPnF3r]gϚ&ʰ+>Wl([}ffӄ TvkQPi32#/Pu`(魞ۮW!R~gҧO.ˎhaȪ7evHpjamTnopUF́q]N 3_ZϼvAvjS'k?3$Ҧ0-Qtkw;3lz(YS2u2OWڳuЌvTNX:}KFm3Y_P-9vn~"{8ȑQ.+5n2?eБlorghˠB.vX|3ߧsOW4=Y_8h=:}tEoe 67Ƽgkw Οklh.i xz5_m;(CmEeaۈ/ǩ 69|KLFt~W8®L#n眻~']tշyΞϫ9rA{UmZU]45 c5v ƪN=zu+ 3W?<+tlʵ//e y1כeɈۓ`ea5 7ƙwY۾RҢC7m|12G28b}=/]U9c~ٷϣ;~sщx繉oU۞7:~c7"#?dl F8*ӫ#L)uwprvyIIwqa)=7^[3= JG+}~Y-uQײRRndq[aR\WyIg sذY "隸78d?M{'W %l_%>^':K߿[ab$V;v˖K|ʗx&1z&o5{RDÀym7oz%0l޽Ӹ`ռ;g89&:Mo}No'ICR/NKIo}(\eE:bГ3.Hn7Kբ/#~nF߯Vߞi7kkS5QV_7!5+{z ,SK'O3(d trµ>nHIy4x쁋s8\ɍa/TD%S f~8rθ%'>i}r_תlqٜY}ڪUR,]zfE V4vɚAGZ#?S{]~t/dxojO ;G249L>"~ߐMIʝ7bmy)7V`9hst{nY4bKm#g2ecO n]7CzIxIePLd>Qo*\;$-֠]+\MOv?v8'qɂ®ez]Nߏx5t:kVe2̮riR[liD)GvX Z5ާ#Bw'M+ejJ}/zu܍-\(HyÈXU 8$G'= ֏W2vEDqHՃN{={[*`vY<(cͷY~gۼ,]}qs}?ru{ߌpk:Rh}Zj^WQ]&L?pB5;:+tװB{}|85*d 7_6=7?o4I`­X]{餚 !':Ȥ:,:%t(a_NEׯk/ĽgCMct8QgF}Z">),; ;,l>ZQ>v{y,YҜyٌv]8QWYJZpݓ &Yp;U}rAG=o*.sƲg=Ez;_1K[ɒs?6mzTToR6X#쏪A+^ٓ+I45j72M߼ow.aݻ^Ns(u>I<Ψ .cJF?ˏ'w|~5˴ͥ۷1|]=={bEFπ5;>dZǮtP'&~:fP ϧݾıP[QQy:tQ̧68 Fy]_^y]O*$}k*9]_OW[xm*4/II3z滲G&vqJ?RT<0mҰGI..k"qqh}rga*|99nr·>[JK~?łSRjyQo5B-=k)4#[ Tp ;缆3\sۖ-支:15yiW]yP?Jwܤ3K.ԯL++?=pKuǓ]^z{Ox3gc}Ŝsl]' |EF[Xsg //lxympxၷJn ?.H8F|Æ;;uo;Y:R]F.Mq~W*N4RT?{c٣~s2Ge[eAo\]ϫӄ&nͲי;fl2},mt>;kVpܦ^=^[hǓ;rs ۹SO96A \UNzySg*ܹ0qdiz`髙OhqW ,[+/ 2 <~BWwGOn>#W=wʣ[jҖiTq37xGI@Aʕ l2#?f߱Ϥŏ]x"=za|UsAGuv)@xbA_{g?jm+y.<ֽ3'N5A#IxAYno|e'RSy_Z'Jz,Nս tú7HQ~߽Lp qeDu;Ȩjן-81./)\;O4}ڤPvt5v܉K, K^0|mѱ!e.ǧvWt&?.?`5}֕=I>#Vz:G7.{FЧ+WգGɊ~mbιAa˾N.]>|]êS4O:}㖍C*V{cI 7Qu܌\Ӓv{>,[<>F6а5FNWNK]w43w%\:Ҽ,<˧{j~m5DEgl h~ve{ (pވ]*n蝎M>xMx12w3_Z{Or7wi*Ro=c5;ywt4^:'P?_6kr˕խYi;3?, ,y/ Η=ݱ{DU̝}x[ݠOƦe_x9;+ui+ [ڳO7tm2 {;hרKC~0zzejfnH8% Vȵ*;;Nd}no_klȅWޱ]O}f?roHgܬPݪ/]K=wj㑵'RnIfv hҳ3JxIywCf؍̲^ր$Pq% CnЇ6׈.:tLȰvM 975W#hڞW~*xޭ=GΏ"^x2SDN%y^m\ڼxuA5SYQ_f=C=1_o.kJ!hqӎ쒇Oܡʇ=puz㲢 WO\&B{K;u-SpyAσ*[*+뷰LDzcaǨ}M(NoOoy; šGl]Ȟo55~^cb3Yl=&|xj7VHHiLJԲ!\uƻOx3I6tdnO?x5֤8a/wδ޽i:emU^vvڷ;/m~hY{YEAD_:ο]c@#!Տ4ba2@J/e5lgV=Lhs-/Y2E4{#';,;Qb'.{6Wq`׍ח]cnmi{떯\ױ8ζ;$)}}蛯}RW[ٙ1~?.~}pсg{˗ud)sFiNagiB7)y#GONSCsfĵ`?KӡЋO>cz.Ϳdxh:+;j^14g뮵y=I'>Ci6!Vۘ{\BygHsZ'_9I4;RZܙ닲⏎OWt Kŷ'ME^CKf7DДT]nQz>rգh=[ ;\]_sD{N:X>Nd`t34׫}4ahkPy`>-ǁCO#` iT@ޱktgf9kOUY*oZLku6= T^yJMլ{}U_7I*fozd'ÙCo\(7+zYի&ipvjMww3Ѧ}QgZj6HO jgޟZKN;&ͺo]o]n*6}9ac.{s^Ze - ngѷ՗ /Q_gtMv[3^ĭQ3/HC^L 4I4ҧ퍴]QY:}y•;LslpGcS4N=̓<̙]{ ˧ZA/g5k\ *;kԺv;ffWl  I=3/Z:-^Q6e;ϒ߻R񝊝*?sY./7]YlJmer|y.l5>ꋍMkD|qkf6Z28ʿr].oS Th S+tȽlǞݱ9ze M|.6C4zx=®Y+Ʇ7bć壂jGnzUYz8bɾ.:\`͝Sis$= W6|)αÆ[Aqcgn5v9OG]iRNOU[ɷ:t|(\19vCbyy7ogkШҔ#~}xEo~g|rKA|h.}M޾ƮUӎ4YydLj "#i:qUv_O:hO-t^Fst^KtyUV}l-Gշq|yl]_coväoMvƴvb{ZrpEWvΕMK+߽ $~jaгZ㒺j$~!?@};;l.:#=qJXkLF/-s%T؃|rfmrg_# /Un蓿f򀥶j_y%ƵuìעOL, B3;DL2>\P{s1h}C:YuˀK0Dz{ͧz^\:}ajڏ,+Xe•gN>c8Ё)]|k>\0LD+u桽rm;VfcOx˪{pbQunvvܙ&C#>gt/s ]GsjF-1;1EAN%}geֹj/DfTT/MA ^7 anˑ!%* LӏNO5[qAk3zi; o%O{_/klhX \`ڽ7lM#v@e}w~G}w5x0?Xӥ;,]old3=nZ;M>jUyH3kw^o:MS^yXЈS+}?WN:_"0ȉ(6:l$/W'p  6K_tuInuySU;$nyI7:L{jۑJV}^}@n owlaC {9SξЖfWfP'cOM F0T`({,G]0UyGp,7pX83G_[h z$*CQWD4U.Q֭@_1%\IWebXS&h?hq =}ÛO͔VYᩲPH -3Sc2VqJ$o-Cڑ8V6m+fp]^Y-bB3xW{$9Av, "(k03*Ԭfg5U%5;Cőht>,c-'AvNJrKVp'lUbqOԲA(دGj-bnu=5ۚef;zCtCo77̤͇ w9afoς l[ѯx?^jݲ8Np"Q詜#*@٥b"_Ae~]jH|ءRz0gp:̒s~U5ן0<4/cnq^чhx%A)JA)M T{$sC9C>x%]?9 ~U0ԫߛC>+澳.&LjvxO[1"X{_9Ng5M 2i9$Ym$m}[?xSe`TI fZVgfPr ["na8fx Y=Z>eK,&_ں&\+0a6:d=3* \D7C 4'oZJkXPA;@TlυxЦCtTXT~Õme- ۊZj{QA0DcADשV 3_+ċַ&c{KZԤUR}hPkϩmE-r(u1(p%^\Y @k^./V~a[> AZ@*Cй`X;OM0T SђsAqj[`׃Res%))"cZMغU~%C%=l?X2` P3ǶPeJGsnS$'گ ,z` -hL#ϙf*{7qE{Wx |kňMnr{+eV&d5o#-e [/ub&G=S%-d3ť.K\5>b&n>%GʗA-꧵%П[fg_OYEQCh|U=".Vchz0,jFRSݕ~W\a~8WT,$ei}?XɎ'TcbTxi: d"`Y)ѦAj6!'a[Jt^=O'NFsx2Y]kyjA[:@}`XUzBWaoxԟh^̰M 8lE5 k{['%hUt3\ӇidאڱZC. PGVs> 9ht+szwBIEѱ0XXZq,!BG~;˓.ln+ڍ]y|dksH9YQBCgb|c @+aҥ0a)+Xj=5F [85d[W()R;o{[Uˆ./=FR/EB4V31"틞 MY)Bb6O%n`k3'*t3_m#%".n ;pOUƮIh[s*4&T)e$ԧFKѳcaN:Kc(PFo;&zD)nq۔]h mȗm&28a욉2xfnXtAN:Dso򚃨Wҷ?ZbV)ͭ@*bK9R&Pb͇>̶c& _y%maD9C܅V[ę&q \QoL ӄX-d8QKmcu60q d3߲~l-(ntno>򀾁.Y1)uOOͶn"[նZ=d{_(%N#dcDRB P%v; |'ϻM_ܞan&*Uv~dN*b[_m!WgDz9ڂ&g&:If`deʂO&U7]jb E?|s_& ui\(dߌ!F B]v~BaL>g VZzi"y 3Ke+*TٳH1eV5%ϱcDm ZWxlǤh2ҫԤUVJ9BFFƶ&qX~?o%ghk\TF֨7eM`3՘vL)ssNCxjrX0:I5T9%2ejI+V3F3#e5z2=4S~(i!L&f Ll:VyP^_JEIZek1.uXI _)qy7٥[Ra^4.ڽP{ :Q,Xp$hp'5Y?iM4 `S}w?[KW2WƝ I8Ꟑ;;Lƿ+we_/d{H?W!5/ f]@m|Z4WJjۯl}Dy}?'U?ՖN ̭g~"f{Z? ~|~U~yD weϙ*tYH0ʸY^t|FʸwzyEg#s'0ozd=a`c?D:90UOڟcG?LSq?ȫ;+06ϕ407O9Ο 6c?+ҫ'?WSϫcR?%W<~5|sV d}_̀_oumwU*h$޽ws<`c.sSݍ@~nݛfv7kn9^x_9 tpw/`,9 `,9 `,9 `,9 `,9 `,9 `,9 `,9 `,9 `_,K{YCWYq=W;,9 `,9 `,9 `,9 `,9 `,9 `,9 `,9 `,9 `,9 `,9 `,9 `,9 `,9 `_,,"q9 `,9 `,9 `,9 `,9 `,9 `,9 `,9 `,9 `,9 8 Y,qnWgw0gYs0gYs0gYs0gYs0gYs0gYs0gYs0gYs0gYs0gYs0gYs0gYs0gYs0gYs0ga/I+-诰 !UnO${gl(8=yecSЗ*з }!{a[CL~  :Iwm^m.OL\Lh; ; >? (8 >]8I}'6 A  2]\+% *)UDrT3>ɌRkb?֤F{In~[ U&fc:WƆ;rA7$FQߕg6\΁pZ?rKҨ#wqh+= `nm&!&D {wb@J?Lػ%G~)ItAG8AѷdoϟboIC^Sב0%aLP*%q]ߔAho李(`:6W: -1d by#`/̿Gώdq{^6/ʳ ~,`yp'@ J @@ ǷB@s@*d2VN8*Nw4`,~c4"?si&]bK7sHx2doH`ϸoa`ѹX;:VĿr=V+]^]<{ss5йh9M|ymp]}} 9u=wq<_4'{ܰ>X~asp{z/CaUgu˅zlu˾E+b]%ؗDnbx\QÛ\q=nl̉|+O)];Zx<>ϋ %nq-bzvg}HpR!rLÅ5͉4UA~Gٛ)NLkbf(u1r3˅kMIi] Ai E= CZj_mst<,|ǢH+c?[\nopchiX=hzqvb:3MsS,ϡigXB&6noh(}G$&s~bm:*mb/}9 bCjR;MޯCךT初/hrtBJ*ĩg[ĩ2ȃ_:4y4MthLk yw;ưqLr,3EJj}6éݿ36pT(󑶳3A^eEscqz5Ͼy_W,;XFg,{' jWujyq mtHȅ[9< l31@j0">,z!( )&.+n2@Ē$O-^ UÄ)kCDmxIQ@zh3yuڝWYiidx.J:qA,+ _:Icu \X\[WE/T>[du^ O^i"]u]/3mKI875α g Zzm$_g8<6>˦'/viGe],E4 cŨa]?vIm2$5 ٺO.Qh%~y9W R:61?גÀP+)AH䴊Ā[rS 1Dzz19IVss14|lCđD{h79u㌫ffh0?<J3h_5%rj$  _3:):,_‡O E5'ԓ&4O6 ~6򴈭lr#im/3=_ב2ɜ zcB=v,H׈֌'0\@&lA >}U;*bS_5vD+176[$: ڙ`/^x_1|BEtF~y5~ʇݓ5aJFyRuF bF fV_[YEE9LOH>EAv,[|9YKkJVTLOD77y57:H/(^1] u*;-lBϖ zly%#=p΄,2tEq'I% _=@^ +A~G#!"+RWߔidZ׿G[ui)O~cb5qD⦍q:P4>Ox9]yFtSuQ$xŒj+ΌFv^{o3 "$H' Қ0o?p?Ut;o>LYPנ)/Pq֓ղ5 ; Il}UQc}s2ܧ.fzJ| NhJsy26&ҖSK I}ә+SNHXA+GcQQ[ 7 mQ}&jQx:[{ﱣ7+)3Ygd##[GӏiթF$YED ]AS!饩 *TrqԸVv<"hh'!h6~P+L1't'Ey%zTc}w:?*Δ&ތxj鹺d\.3clɤC{vtY O*,6V-) o~mYf]UhZXNl#ur&rAޝv32w+֓݌i;Tѳ-TycrK%kL=x@0]Hy<T閍 Z/oٜi*d n1t 7qauuE;A%$lX ܝߏ/xPlq/s)Zt}]Pᷝvs޴Qqz3)=[ѐiZ"պ y'jt^B^Է{q2WvtA^xgM]8g-":me=Ӿ83uyTC4 y/~WNiXۧ\*ڧsi>!F]@+<^|6N@ZS,7C:R f'h) '=?w6x|vsp rh@J3ypIčoL{`([~ dUmݳ2d0{,LzsM]-$QeDHY!{v0$ i!5Y;U555݌@ ShxM N>9_f1;|[_B7_DQC#^p4u `8giw7D&UԠޘ] R.*^+ى~XDUb2cbx ۲̙Ę#5C*X_cN^`.1= M# vM7(Ƕpk]0Mv4dz hKtBWOdbJH zof7 |e>z=IXARG6O&PY$=By}_6)Jod!+HqƧa0öQˆ\;Md$~n~8C sD} ;zmGKX+}R# x6L'Bi7 S"oђ̱+y]ٶvGGGQVM_q-F(e**N:ЬM,;Ad_**~inH'陇:lDAj"PXʗ=L_S ߕA1/06 @;bYlR?j j:AnY{ rڥga6d j uDW ܎k6S_W5&$:sr,ŐP rӵ& AxZB4gE/2T`[,-P/+hʣ'h,Z=ۙ1vj3n~nAۚ.AަwDt(< :f?o {@Su^ձN;%sq"%ŚFqHi su19;c'b0My٢pCzؑgKPStcG!A| 7Tz )68_8wo_)-Wx}8PZy~hyRf1Sڛn~W9.֡{}kzުkXpgg";jQTaG2eerV"n]5#R Q)đAWdĝ5Wc]6YGpe2FhE")HpL-SAc? /!dҾN?^9_ԛ}zha@uѡNj>FܮU VrAoI><ia;^uCù7Jt-Χυǃ.,ÕBeSP~q%Հ#^CL,N'16P߸iN=BT*(MyOUntv(r07uc5qܮ%RҚ":&srm>DIIro\lmmq`=lFxw> dˎ])d#fRsau0P,cl7DCra~wuiSh[b@( q_ZI$(4u1uKʼ˴ !woƠW.Î{Hø3Ŋ1"ƺK2G",PIV)Ia,7f<6d,Py:G1u͒QfE7+`Y3 KE'O=LܜąU45zdq;hjNC X4>8bG DW֮z]>5;!fEPm"\ ^͙āz:٘I:FfI5DžؓΑA Jl_x\~ yJWq|DPJ)N, jǁؔnE9^SwjDWdrZXh5>aV!%WвV oꠐqZn(s*Q8[ƙ*>c̫7iqϬ٘£i 8N7a5A,s#Hr8ĚpkY4\X% {ZI=%5^TtWBi$UjEaV-YA wd~ m8mMc-T@j/iն? Jָjhhؔom*+;i"R_cXUFDZ;{K'ijQr$&st"6-\n:bRͅ>][ u |q96L0趦0y'W_)]"Hk[1iŷ==ruG9^SQB;=ϛJڄt*g~]"hx闈&8"u?aJ5MIśO7]#-aϢB {%r)iۮ$>}QDU !DI@6{c* ܒUPjtD7z7y" ![*7Dafy3=̹eBјF0I5^ʱA;v+C+Qz gXEYeȁ ՠ̄ cϹqӏߒiGgzL8b5syKjN H$=9^ v=o}jcnIS!Az˖eQSytF.0.12Ia/zҩ of5:J8Eݪ**eMz <ĖX :m Mh\TD)IֺW{cQ KւQ\*krm"QNێ8/|Lm T!v@|l|ݎ}!|ő%jRhHzf˅ۓ82{ή> GQz 2ۍt#szU:G& |gTs |Kگ/"3kT셭sM'bs:-|X\=)@N䏊w"E#MšͅTEFZI%ǻhR-\v¥.N`(r*(kcYͻj/B0q$״ݺo{ﶟ;6|7v=4Z(q 1A}jی72Qo0zAo^p@V͜g+K3Bq3SQۙ)x$P=㼑 jN! PxpNJyrC=G$_w93U9 3 RwH &eͿ`66XZ M}¸,&a76G,rj4} p^7Щ9wɲiN%B;h]O뢡|R^ yO$Um; /ϕF{F{ S([o̎C khS (V+*}- jӊLN`CTC.9Nx1Y31v#Ko#fp/w˞=OJf.4Q/{HJBz\^Q# ZA}="Iqޡ{)ADF֮r\GGa=:6yA}R姖̕ GHFI_{BG=%ܬ$!ޠnBkjdzgLАGŔެ0f(q ٜ9J_,Q/?M2tE#qXe$Cd{ǘM#^E,G̫}MZ. U?`zPqVj^ =JAsaLZc0V|V(դQ*B<}VOMT]n\̲G0Ԣ,í묥lT5jE|2U]Uswa\ 2 $RI@+>H:H@E߾ؑzRyf'%>in=`A!ڴ ]@f{oJ:~E3^WZf&#rƒ0_pĆc 紱{?<Z(5q(~Jz1x`]tH~˛jdv;RF26f;a[!zWnlnF(Z߹J.•^YI  C-|l#hkhpWDbP^o%c`P87U[f2\xOtgH<A۪ѯT6_jq;_>;M1KYmqS J˗xq )ۍZ-{L2CSxv'+ unPJ9?vQh}u!WrQ$tO [u8ㆅ3c4w ߭NjjqJLpwS~b{WGYuYz?k-HƚJ!(MO?Rq!ԒMORJdxwTtq i͛ XhEij\g4Nֱ0ߐ Gـ$xr88*q%cb{SD?oYj8tbL䐊اdz?QءD"0#:K;Օ_cZcTOÄ*ZiTU}`]nήv@ySIoN"m.X e7K˖bs Y%)2WVLAx_jLoKo-w"\Uv_.C4/C%?"Z"B6`K#W=gTSVr+ Xj1GG}a0~u^yϬCe^[&u F$[^sDʱTνRxyDiudd:'?x?K!UXoWJ.k Vg%(V" v B¢,?FuG&qu)K <_QdXSz*ߌP ַpBV8Ps%>  +"ox ,"d|ޅC1UeG -SZDl^e8R~VhdОאYt)C\*KkN;DnvƔQ."iep_ƪzyFpC`Sxpَ'J@3kthGX,Z1cӉDIQ wB ?1zR"KmFb5cjL|Ic2oW;!d)ŧIWf "Znāy!VaBZ;ޡ2 9;k5j1ikKR$ܐnMcvST҂KK" }zook܍i%#$Klu}SG㪮\OONQ)۸h%?Eo #nzϾ ❼g46=t͡NWST%K:34|M zXCVA6AK`)‹Rg w‰aId7%PãMA@Ӧj@zI L(OnXm&ëwu3I+y|sfaOA32ʒ*r'2%tb6Xݱׅ LڏTz6ݯ:cvX9RBwʼnl[w ,i,>u_Wf F>(@ JN 4 : 'O5xD.Q',tS>2Q +5sIy/WЋluIvx43McTgq ]xH-'׵+ ^ޛ6yRJ7̈́LfMQJ[qT{3gr􁏊JR'JVM&V o jt/(vaZ"ɣ&7w`̪ɳne]д="bku̔S#ȯsRJ%X0Je }n ,|5UDG#VZ.L: S\NE%*P C-̋}|(IDDT\qo.feBLkBoH>AvZr. :qov<ΧHȟK0Xt&JĽ'QĢ_BU~C)Z7=ZFr?虥6,$8S+(%/b+W84sqx;^Zkq4m~ zHdA  i^(KoQ[1}؍,Wd=dD!uK g6j6|ܴcTN[ - ȡyLhb^Sx^PB""V(cl"fqt1&Tq}Y Zv;b_fG> .gm{?1heb>"~?IQ{^Nonnڂ]|cz1kd~A<ԭn{<U5=Z%J|:hv'w "ZÓ{S!ת+=&` \}snǾi<.EYHoS@${6%a㛽)^o-e-_cUo<|Hufa[{u/e9 % X"L݂2e$/gՂ}m;k~#r}ߥLCtӷŦ!'ſOK1:bսm٘B,X+'M, (y,&XNi㔒WFho+4{%e+g U#֧ۭ* 8M= Nd&o5 {)gU1~cpxIsM$95#sCF3鄩;Jb.7Ì[v>V\y9zG9-lJA5[>mpZ[)sX>`'ݜF[&~emb%.7X7ɜxĹzi-eZr 7vCm|&6w8Nl4iݭ'',N^p:O>yjy{pn ^n]DCC"0_}P|=z};(V~?X/Ԙ=$JG0kmS~%[dSnM҂ vtNPFP-׵QQv'Jwn`V3R" $lhQ~X!Ojf㾯sy% ;@&M4>%>)Ϟ=;eAC8ӊc /@OvqZ37}ʆ{x)mڮ9?&)! E \mV}\fRFlN1{B|ɺ'EfO馏#9}J=R40LŋOee MaPlv(6fr_^=PfRHD1o[ӓ.w~3[a2[1#P(eW7ਢL9m{Oio%)N#1n3 Ž_'EHA`һSzom"ZS\"22"[V641fb{aΰ*VՎ mIuh7Ж:|Pq"9U5!_LSYfN>jXP2QZyjW0y ̊9vO#u{ iw7PC?F$Rxͺ`VQzXO_>,d40~ 9s4VvNыŀQ:P2O2[V|eHmzvAl+\{V+gSfROt,i+ݞ78-۰Ɇl]Gdb8FAW.x݌*N^A\ ].784+Lk&ޑwO}!-TCݥ.jOa-'W[Fgd@wmo[Μà.7&:RJ*Bt0`5cS. mVG,s3{\ŇP5Sʕ-m71i; ڈbL&+Ɗ=FhYG:@>a迠PܚS^PU.}ɦF_5!AMO{ N͉D<I8y },SU_/}qxV˟6LvvmSt͇eRHO{D 0|I ˊ當uf @\gn%!5[1ZiV.A*#mO؇*ҘȆ2z@XحJ/%klN]Zzݒi+pQh')ֈϏ~"yQCZ\-5|rN)11>P=eQuo[ƖR#/gN7j}t:.y됥phסved7yՐrSNYQ!+%>yrث#c,# *O6_>99ǻy dGG\2[ : xP_"|_ymXG? zVvR`<Ŝ,ٵlIv.}8h6?†k }ŝ~;?{)vب粖-IX{E6IZ~NfSKb =v*ن2y67 8ytEMlܩO[ޕ~1;iZ(#N,[{=T YսμъMNR[Rk99Q+6wn/+qernO`NYM8T>Ĺ9VW]gP:)^eՓ!ღcNg֦ UMZl<ި}ww 'vгu%u{=S"Y-<޷ǪrC{c3#{yi42rxǮz_ZH\CqNX<\xpg+gۇµ%T׿R4 I]@-D83}}rNK#ݤ,w#O4D&oym rޠ{:/{gg\HxlysChz1d^ o5YlYH\k؏Hl5VÞƓ6Mbot8ر˝o㖙P}r<Q#3Fx+nz5D馈!/i6HX>٨ TG_ԎygqƑ5Wan&8;k\ݼs>xmtĵѥ9 pidjocw{QVo֭ [~ə=`pޡ-M_3Ի10u.]v6AHRykZקvk+?u:J?>!MegW{pHٝBҾ1-];esh<4ME/\r6il)m 윭NϽsbk \=G_-ιܴ񕾩+->sAo3ϫwG"o37ie[ C: F1Vr* V܉]nccO#6 S 9\9&/ڊFvuéꌔq Z]XDn\N46#;.]۴5/>Z'31sp-.PM_X斸"4xF,wShXƛr~O9sCԤw3w~nV|O֚/G{  p_6zaɵ{,);5u9=.|OaOcM'i.2#XOw#\wia7S6ச;ٹn[n U =7k⇨Ƨzߙ[༌\qlc㏰ɾJk÷3/\rW}<_qiaILpAMUP?,>T&F }nrv [ۦV>m4KTY[%5rsYQ3?^tvRagMzw>{'|Sn|"xэ8n>Nb>Hfn˨yN͑UcJBbn.$Nǃ _s.d7cʬwoǎs={1g5߯ٓ=h3+҇1%\(^}ٟ_x=յΥKKW+LOVLbE\"^/~vM,'ͪkx{RQMok,GsƗWO)S\|LtQ,^kOuz>gui}H7=|/NiA x 1@iR-X~94[7'pPI}"߿?#4qC;mj}׎4HjͿ;JT#|L=xzh~y,k9]#֚FMEaExu!scsFLYlS?/(ݰ{Eq? SnR_5oerהwQR D+|\ڮsuSYҋs.¹ aV_"W.ެb6.cަr~޳;ŀYb 9L]ժ!>U8n1]5fa{$ PI[znn|{;_LSnG֪'?~(\m(CLnfE?M"u:"euaaac塙e鑑7m+Gb)BpG Y$UMpzkip₱cŮ z>{k _mkAA_-޷v@nnZi:GfE-8٪~W9v) L:7sk綽Owkwl_?gavBGaBC{4~Q CL3Mivq~-sђ#=nā Q -U -Q/gviks20Sv۸[NUm:-t<3,c';s/<$Zm<ݯn:OT3#V ۬uS9c<3L%j:v‰aY3cO7m.2\^zҤL5yae뾔;u:)Bytq{+-sw| oyJ)Lj?KE#_sYuxJW/1ÃTY7 YNP/ f1~aq˜ zeY`U5K3_޿{蔇C&s>mWo~7۾mZsߪ{K~37[oac}i#ʻc! .qy*0q113 ER~YauBOjxC;Iյ"á==.~/ij)-5 "V^7%(YM:eчvͺ2` յ)K6NQs?ا_|q[,ykZy]&vS®3Cn'My"pa?Bo돾f5xQʋT~/ 巎̒2^^QEX-^˒|Sus }Z~V=6R3Vp~'fT̍}HC:-L MHziSnoϕj-aqӫqGiPm /8Mw9%ҫN0,rҁ#\\枻q v>9pW07?OH }j򛪄鵧;j{e߷-J:ۣ?߿~j'ǧ.j3jlVviQoϷ<r LþǭZNřjXPi:jZbw+"L9tA Ya4\Lo#iP:&Ȯ=_?n W~k<0vYnzՃ=&?p]vS߁ر}%e&oHG&~> Cmhr̶i=IΉikN,eWƞ֍=kŽ6\ۇa5B-) eU{T]#r 4qmZ1D!N }ٹƭKuxDx '5ϝ[]t?Jݧ< 5[@3;Jfc_+=2g h짘QWmemݥvo gfJ>Y0a<_ْ4ѣmK1WLL3:ǿ.>XzrܲjC7`^}tqo;th*%/1<>f8bBoo0/bT眗HmQ>챽ee`mTjf,i ̖)Cy"[˶O/$]87duʙCOWia&wtʢ#y&$'ܳn͚ЎE=O{v'f$ՁE䂬b[Dv>+ 7h€;m:;K-v|q15wXڱQm8[# rث=g7OX"M~f[.B99ۓD/Ft^YcINZx|fSۣ7.u袗IO1q|X~6{syWt2ʸ1we. X+8Ϗ>Ό`K rDY7Rl Q<MC2)%8J;!o`W9ܚ'?:=xq츝k)9ם!uK:/~Ca[ߵiՎ֛#=&߬Yګ3=A`g( / yu㺹/s=L|ckՔv,wJ'/Wm*vncWmٞst.OWNj8.\c/IaPvhPOxQ/!sN4dc'.}]ɯIGW$^8޺kD~rͭY>]azۢ5fѬ7K9kD QJS_5}~kv̂yIV>&ڲ .n?vn̦~lK bf6YcL~R )۫N3 %tʱ%#y?im3wOF8c XO PE1w]۲;-KP5>Om=\2s0TȬG[4-̳mYwbv=LOeQkC/HDmLxf!ӐtΪ//7暬c^`u*Ч <ܤ+' =;rrob~9*ӄw|;V3F8dݶogM};,E255ۆ0kG¿2ɡ,{ϸyE Ԝz\NcvԴDRXtkϴj=o?PF {_cݷ/='Ztv2KWq#<;cG܏f'^L8퍗myO9L3嚈'7gW\ZtϚnTr-hOkA7Ks+ulZ;yfØJ̌ CӇB_NJ$uk0b4K;/7c㮨O{>y`]XΥW#-plE4?n  6/w'5Ek|IҝX^@832vt’̖cގ;Ka).<ƫqIuk6?{ޒ' &<$#8%H]8_y~~o?]]qîg1뮣y奀I|]mhCW ꌅi|N}?ڈ~7ӧ|L<_17#:\S,y]v[f"]XmJ̓5VA47U^WB_Ƴd`nfܙΰw~2t{g3\;vA^VvpsG;M/_hȒfcǬxS5)'oH@ySb̩o0)zaO[Uϻ7jǜiYdYzL9G,E6 b[;В"wOQWgΟR4OVqǿӶ{7ޗ؆+r7?fUٹ|7{MuX95آa3bBnPsyOHX|k2mIgɇ0▮enWeH?&v˛~M2ӳζBy=!}+>PZVXuꮫdm VٮR&>uJlnkfЙWrn-̺|v畓DX'5>tQ!}p/}I ]I9/<;ڌ9nT;&vϜ'CxvzĠ_z]2<#۵*3m˘ֵRkrSDݷ}nL6ˍa}ڤ) {7k+5̻43])I't1pJsWp3_mzH!YcyoHY73 ;?g]ioYN7\k7;+۷8IçmKEiũ+*4֩QF?c̳<Λ|`]CGܽ7ovoOx/g$՘ÎgÐ7s3}xўm?9qG6sw 95<kw 18mdV΀) tK|ea1Nd_pGeon-N<ή G>a]tt{ٜo#!q[> uTqR}K<{{\wuZgܰjrjϝA0g/4p@ݦé l= ksl0nՓ{η*{w=ϫKH7*7rs};\ ϕ*w e67<}V=Gu0<o0mwϔOmOzܺF`nc4`1F}^]O@iL!︧VlYx.Uo̺x"ցҐ%gi29V*NJmF7%xncD_q|fsp"Q 6c,^y司g)Pj=f0r2%NJdR1NDV^UCCU5*)J J5D{m_w׵l%@* t iX=n0p)r@*`"[> 9ƨTNwX8#c0̾g=ilslYlJ>78a-=b~Ye0j!I-83 އœ)2@Ɠ 0P!B ?WKZp> ?ĢN%x,^# 8~募킻^ fY:Gbqt.n,!eRA`_4G ~Gի*Q]OcsdQ݁mi9l;c|P,TKb5e2IY l@֬yYB8GX#eR[r=.͊2YY#SKY$aKfK,~Űaji1!GS<+5_֐iJ7CcyX'LY-0ȉ9r`!)dc9>p6l>crp:hBt<\>NDD2Dw4u JRX\&). % m,%4  U4 ?.ĒKCm^J%nhAU@nTfŴS3*Yq~Obqa*5x_o7LpHR)r SHrА H5d>)\M!@7GC$jy923VlcBW7)jJ0d]L=$< ϰjquF jme}?S{1`E9v l>C-9eΧU?EvjJ!}tclᗖjIjПx`(z{<: {sx_|dw>C.@,iD B2k)xZ!x>VB+$R dqve2wThuH2wDCI!+yjZjX 'o;NnN9GfޮP(V ;d<hED!G2%NA%SA&&8E ҳu|K|.Ѐ:hn`ZTm-ԅ8z8bc/!KX"\QƐBZN, }DsbS39h.3Cg}"I43s 8b[1%DDb)yF$eKXb9,bʑj ki|N,FT ;.cw $4.݀K -(rY:$Gl=U@$IQEU)|PŦLXKiehjMsgИtN7͵fH Ϛ9 l_$F|Kh!qLX-! 0\PAertƀ(JL2[6:ъbZ:b/A:Ef2ćBqQB!RjDrtf_Ι&2uJx Zv-B75@D%/k 'b bIf4cLFs%~`g \qf,.q!Za 6af`J?{5 wŖ)и؍tB42걒jYQbe?zkװqYOʇp odP&/Gc0.:wA\lBUqAd oP'%j)[bf2yع Nj_+2s%>vcmer=VNA9P-82Aq[A,x.xbc !b 4Nx˒*^EAth eFss5@i[ 4 `hLȚ]pPΠC1+^))dR:TzGūGjhմd;?SS֤v؄E2O18.CGu&8?ύN<7n2"[rjF2|P ^|PAkS]qFP# 8[p9P.zd bS1$j+ hRlH-lSP u%'b[?D>$3^$'c nzQ ]D 8]D_k(Q@JpJ {!@1UICT6QA?ov [͖JEX-fO ؞Q6޴V\rх4 zeWaC&gS@m)t r.DH4 MCAu缚VEC)mZ5KGJ3+蚀15A@۰b`d^QQ%[2U-U+*+ćjVK}6 _C(Tv;aFsa)B2͎HT@3t> `Aʄb8=pGdn.I-H{D`XIz'X3w9 纯ceH)bmMGxDZ PE*I87$Je*ѹL*Áz'"€YPTnZOZIxb4,Nfv"[ebV">}ךP>kytN&l>EzA)ԫ@EZKL!ŁoLoGe?9lY!le@}]MYcUS'YZ&Z$`}H><0|Ghb*t6\9!߃aƖS4r ʆQL+fQZe(O4f+8)t xt !jcXqJ`;%NJ{,9TN rݸqCuB1CL1#(OU*Q95d/ 9~^C+ӟcEVr2Z |Vܗ{c=+Aq620ـz_Xd p$K5gq`Ö[A 0`\%t[( ƂhDt@PV y :G%ipV>~)i-6sA28UZwgH)Jho'dJ?p.A?oﲫNIAk*ӊNJsЁf'7h^qn]z^)!@W66F9.עm: MCKG4V WFkQ) k4n3` QP 6kT/PzYCA%BN,K-OZI8oLV FaQ0DcEC)--hb8$'VX@{t '&Nfb59$TBH@ŻCPA#.щ4 3Q>2ߊn+I_~&Q\:*:N$ d5(hGYJ % @‰QI@PBs>π5(J?Lx-'xn:Zf:V)rZZ-JpHs|q8NZ84vI$l* Ż _// \Pvg [s8f;x Ʀq'*y*v46;透|Gc\>FsKL^ 8К{"~wH-E2Zhwh/dK;)] P6K\>hh(-6Z 5͵#@`\;I9P,8fB;l#Lq L6c0F*W yc:A.B8aIzm+ūNhQ$KCrq8; CZ^ f*lt"Q 'Dy;dųnP!Tzku(um3| %\_ԐOZZ\fR;W =G1 !浐ZWJLecwv5hG`^h/ H*:3 OOWA\㘭}vFv\*DАБBDFw@Y_"(V݈ @΄ XŸ́3/ x_x!-)(WUR[/VYpbDPȳAbA `)Ł].G$oKxp=l: :I[#Q<+2ڦ\AB3Ni+Dm*;(.F2 j r@mXqb/kU-#7CFPK\T3V\3|89[a'% t;V%Ȥ`+Z~p5c?G쿭S^E&62 kvR!Ў;>е4Tv Yh=ИZdH&@`mrt}>`;\@!!Rڭr}; jw D+ןG倁΀F<ʴ= SWLII~^ x2N DlBMH f+$",l@OŶP=E"U4Dc$}MNA2vJ-ʰX["ı;;> NB{nh<-:@20;2SKA6Fo^ЇL fo'\T~!t}bxhM^׀7jcYbw#ji޵p i<\Mrxh緖y5~B?R6,NŁZ9AhZȑ,U2}ѯN,I_Z'jbiq SUYl+2`+y6-6d >lAN~F":DJ&}aH# |6GҫȑhGXX o;ק}7 }|WEe)w tKڽcNN>@@ZGZUHM5( 8`ۡ*2 *!ߏRl'U3P`6EMtvɤS3) *୞_Yjq^UKX,]ןҜ|P5Ǻҙx_ê 'oC{cG7!%S`T%6p.2X|5\ih9Jf%a P`& Rѽ(^OABiɣ~Q&_T-~ѿzF:J@$ٲ !Z?mQ[-'']JN'vADC{UhG"܅-^7v$¡L>6)ឃyk[orb%Gg>BpXwc8(Ⱦcdq}=++yoA+WLS{bhgʯ295ԸšSX *WjgDYrD\NjCgp v|G`KlRJ kdhʔ!L2414e0ed 2 )cy b)Pْ -d Kd)kz{~AApQ98"Ě ӡ\Gfۑ % g3qph&{W U|6ǚCkmךCKGϳKg~<9mܮRu +A&Gge "xJG +ZZae@I4i_iCNpa'4BEI'-T);$Ir@69Р׹nIL)9`[/ʧcjgX>>e Q(' }2n?@}eȂ}Y6!= tI"B̐FW")R>VzT` {\J;+ <0(,3y*:05=||A $>iG`*-O`X@JT!k|w>ph9!ېIlZezS09r3 HxR> XVlm`ӶlD `Rd ]ׇ#}$?:d`K.Р,hIlCOsXhӿ 4@YMD/q:y\LKMf,7" lf[ UXHd 2Ao!CJ$6ͧ 2{e T!Q)˥6>"qJy4A|'qwa"Cy,L;1FjA _p"lAIILSTL|L!uF %Pil!2N bUJ!jq%IL{JCE֢q 3Y|:y^c+p".ME@5Aw`{&Zȧ*7ݟg|Dt[{7$ \+Gϧp辂ޔ$e +i)plI$X>Ttrd Xݠm1c@?`0+$'Ɠ ce/Bl Peh*x1If2@6z±d֢3d'{: $s/91n 3}Ou,fQ)DU:E썧Y(1mX/dܸ(J8  K͎5z0;PVJdyKrdѨqhm2UZ"e+m2qCo*KEKh|=*z |L[6u;kl(NgFqsZQc ͇:>Gԇ dV ŕ(U%.ƬQ sJ<X%i"z3M!1Xt *FC!S߃V8?A+@i&l sdlHy^iO.|%OVTxN#-8 h?n Ha>ȅmfx\ PS AE`<\F+%z:|j/::u]+Qa"T$]ѪUVd0g2mdMm0RFA c3E (WQDHkAjd\3U@E!(})Sڇ@sDHXi"}T Cf!DCt(\w3`Nց.c@ٳ`vnOP )E)I$:F Ȥ$1E[[mȄ}誅 9Szg r#DoCl );W?]A[Xo@_&鑅BCz֔PqUD6.N0l[Q)dQd28f`%fG!MvJ!¸sƋxy֠Q"7iĀ?C_Kl[R OQT F@+.!@ m^EVLRD&lI0ۀS2GO.[P퐇eT@4I8KQdq(tR-Vz=z2`A~uGpҀD8.ɗ/$<m(f MҦF@RM @?X0 f)1=K [(   4$)5ᒧH:.黝ْ`F$vA=fŠXql<å*(,6)IZ(ܰQ I- [AC@Q0ЀKds 0 <?фGgQ i88{ r~jPgOɖ yM4WЃ<\ i *LBL Ib$2$plӭk*X,V, ~(<ll=&oڐ1ЄfHt-$3a%0 J Y".#уۢҽH'-.%V x52Nx cl|V9 =A`=A'Ɵ¬vCO`Ґe!H}3l(dP&TO0m цI| %7 *T_ô' x?C_ySxX`m}<ч'Ja,Z60zM`Z5%+F#_J 喸|I^-_!@W nlqov6ҁoQ`$APπ!Ʌ35^;0:od:QJz|xD.5œ xЋ:!<H HE[LDv4uPo%aM rK@gc , hG6F`v@+X%(@6т8vpHXtYu&F&y$p'S!IdX \H:`):eȀ\Y$6d1~+K`OA3n h=afl8A5K<Ő "J왅  g׬-9 3ڲxxz ( oCm$p0Ț׎Z:?S_5C%AB5J䰨x6Y) ҇&s hXe#3k_59 M8s(5F;{AkI<;IR8oـUJ0] f;DX X\ Z@m7nO3d1>)"tm0F{ ">T)p)X09MftqـlL-lU v#ddw\ >jȔ*WZ`],IMVsm']` sW > 91Ό1_q:0;9 D_ _B*ئQ ݡmnI< }._B^a`Ѹ BEzh!ER]8LhvL #M#&#B{PE&Ld7jb:r1ST tb((5аsS~6ycK)qtR '\ +5S1@n8  fZ0N9!'CpKл"b)!~PR3͗c`~I!McP ݑP(K ـŞJ_c<X ۀ ۑ-c&ѳ +wy^EUDĄ7=* g I=^qcBTL;abJNɜ{WyPJG$^ ť4U<dd ".y+)2'<lq%q9鑭5PH6 ySHpv$CV'̜Lhtrg0H4DzӆQxX5Y\Miimr,}.'w.`t|HCx8I9ut fOi" ;xAu ` l K8F(fp@P/*q3( yL3d!*01/A:g]!0T\8H8\JxbU@#0a!SP]`R3 ;Xff3M5b'fcaD+1FΝf2#=cqp'7uҟɛ0`Rk{#3™fpT˞OQ ٟc=3,Yi\ {P<Ǜ g*+X-Tүia{X9 yff xX#Ew`-<R!`K[Hf+Z8Z<[Q,@[0·9,ifv8,+fgTN*Bﳞ@;NualPqgI1,ڹ%K[zQR ZEBAmlN^#iǐ`ɖ[5{ gC -F R %433kzL@3;)}#!>n1_ߑ䘿/}o9_D|)Ľ<)K8g\ͅB1Dpx|!\n#@|hLAh߁fS־?A#f\VnIw]q5#vz=kdfe*9=gk XTllB7E>fӳ;`i0w5Ӎ+#~sބAlНEo_%W{z:Wm?ޠ|FoGoKAzS_MNoKfHꏡԹW&Q4C7@z[?DoߔAɿ7x қCo`{+%8=6 ~maߴ`]wp|ZO&hR&MN@ MѤ/^/2 Z@jڌDҚouvl劣5yJ<&*36 ~ vqw4ĘHKfjt%2KW@0 ]e]c>v hB,rpMb5YfߩF5n,u?r?COX>W4Ggi։: ]+ fj_ľNO։/ă'+@|[gh -c S}nῙۢ~|[D":sP7}:K?ms6#t=7>LdURe^n;z-x-G#MMg 淼{1 p6'w|ЄB4şsf3 u?<^)OB?w}s:, ɍFiB~J ~/y>\kiz}p{g ~xO{|>WHx~;$,A .=c gŻ^~=ޥ|އܼ~t4{i]7ٯ泭-@,b.3 yT ,mM:D&XJnnbw.+$d ,8^[IΊżTϏ0>tuk->tޤ3>u,4E ~q$wBpO/5(5:Ғw+GVVܲ/Xv^_̛뙙h4~v'/ozjcUrJr%*;9/L^Z\x㡂G}Z?sǮyt i-&'| 'Q{OtO 6|Ft*RCS)B!C&eR|,N?yl΂;+?oX:R=ҧgR܏]=Z<&}E(*BbsXWBj.mg^Rʿ2MSdjꞽX^[WÅ]ކV/a z2뛩_ |?{stjFxyOkWa(|"^GW>|a $PhADm %RݎmZC]N4/=CPTiАszǜehkɺ].p\z4-RK d^40\.ȧyS|;LXdM MEXZZՒgVW=1hot>aOC?P|dߣioͽ+Qέ=cIMk*,뿐>&ZQ/Y2<~O13 g%w^m{]q.+Ӫgene/$;GXo~^ zKviOn2aЦx/=呡*|/4i{ӁauT0RF{fږ%N^ok|vixnb^wOˤ2{>Yfc_:9hE^kӒnbi鯺,?+)l?>-+iAʅKk.wיy{<ش}}V|8%nw|s;T#NE, V5d<~gcfYUwGT{Vr[d|RrP=[ֽ\iUZZ*VJT39|LQ6KEԅ}NK1/#}Ko~/.uw:xBXeRiPyٍbU"Vq/o&|TQX]jŗ+õk5nϜyc[U%=(.x.Mgh9/ )4#ߔ| !Ҥq]nJ9& 0ۄ#""(Miax%.=N+kak*YŬj >#o)4iV~1f;2ɦT%5նvs!hg9ߨj9UGѬW ;]V,,h2۬5WPeo_kӥF2BMt%XmTpT8CçjQ\s0݋IyU7Ǖw9us-ՇwUXhZt#aƷ8c8^b T&qƐam^_[+qИKJNb;MW?}R N\!Q\ae7^''/ə˕xdS~]>4}Gui_^ZF?užWeIkѴF-Ci"-ywSX^كW=1k3_weK:㩓IiocoI{<}:Op\oXuéwF^0w`4[U`W_=H>ғO)m}ȕ|n:Y~ ♭y]-  B.L'9kVXzH 욛n9xN+KJ_0ol mHJ솋Ǔ~8f*Spv^Eb'_*n3 qzv8yoizQ#t:e Ӓi'%wsM^ݯ# tݑ l&凸{{{C)xDwBxaccã-*h |A/-"D̷#ap 2fIJXnJkJ|FcwM~>$ImA,Y^ޝӑu\|sIjZZkV"/ոpUҹI]1.J@OaE1Ʋ^\ͻ4etLT~iI̛JIKu7K-<K+I-6 73S%]&Y5-*_.:uY {IfؑW~^, z۶*^$}c?a}@UټIm&i8E;1k.yVʲy'_ܭsRXgmgǫ5{ +#^`?X]*Aq]Z+n;0LR0eڥ dEshCzՒk18ˣVO?~ܿc\RO(MG\>S 72b1M۷'F,Sk>ؚ6U/zu8n;ek20NBDE(?v)ϹϙoBќޚ5! :b6&J\H7r8=<ĖMWN lNܵP %Lmɭ<%z7,iGɜu:Ľ>d9^%{.4EScq(}&ܖ/{G&#$E_$zG8o*e6SȪ6wBJ3-w|cw덦EnN0;}Z, %Kd_P-NI/ͼWxXN~3'q"W[I8J{sX:o;Tx]ͩÇB}8 PYy‰AaMe~sB3<{GoFƷ:+biY,-iB}aob1\ŮE>p+:Y'4N^LmCw%E;]sd'j 6|{EYg^A[G%,OS>՝sVCOZ}`WJU/{Ҥm1;Ne[+Co|/bbjԾZ6rkhj(Ga EZx=<˙Gk4vp뽓Ѧ^Cg_z}n9lPv͚ܬ9:śki!e 7>ؽ9QhuaE`Wte^m[D̶ԱÜ҈S Wۺv^CB,d^oxn32 C}mɶR6i܂-fwa-.ȒbG$U>1)3ʲhmɧ(=^V^:TX,"Լ D'f__)5Oه|qZU{B4lmc5eycݱ%\Otm~zg5͌⧍ע^ouw?)a7y@,';|_6g~ApwP$On?.* K}ݹs ԸwEwkPC$(t2F]B=K FBmNq_2J1ꪤ'#ml̚}9˭:OBߧ/ՃQdo7 QFYOGF>ikky*wwA}#^YGrIfJEOޕC?-U?9Մ<[Q}W}3j2f~㗷oLAWߣMѸqgpsnbz:U-*qʚ྽іh?x>^c[NUSЕͼBCLU:Lj⎗r(euش<!Lv>[DSZݤ#oqlZ]l{:)z퓜,I'Ne:eꮋ{Rߠx+vԒε۲.<ģبnz@w 7OE 醚(굻 縦G]*T]p8+ipbVli'b҄L{u,TzyV־iHyǴc [w·/LtJBwX[1s=,ZՋ b:ߕE|Q7|.!qi;Vi^Kˏ/2@癥 $3dmǫ<5C{Q=JJDt5^rg5?Zս_8[ZLToEqj,v퐤;oĆO/^)"Z%I57(Ws膆7哋GWE>yh@~KbhQ`Ї/ g)zmOY|^sD&?'&6kx}iς捯Bg. g=y7Unv!{wpO& ^i*Mne&/6Av5T:227%-I;:~+t]#pΊs}oGlP}lb_Bn:KޣoJ7xmkѩK˔o¯^]!5,,+rYDQKKJMFƝR6o){*2Ar1mEON,,@|)(gë֏N#Z8ӛkGۏs?VLT# 6-2qeתJ&o |~*tծVwu6v?iw=}L3 Zyw9߱M==r+ЗK_[oR%( /Q\e{>gY /t|S0;vѮ3hm`y>!ZtŢfߋ Yp !|OT4k=Ze*Eo*plQ{?\WbP{uEmk傣זb#vH$i7nP mSb-T[NT'<?n8,BLl5Y3vr{7oܼhlMY$Z-9Ʀ5s_zn ɝgQ(4yVחVW-K=#,y=+¢bUyER9/%!Ξg2N<2l2Q?ZTE gjNƬa:N5P8SL,H;Y_._yp5*=El^{\xn/**Wdje/ylǀ>&wMcn8DJQZoޒ68ti 5nړ#[T qc#7w6d\ڙF#RK踻ӓ(aM5jRQT$:f0\NC~PN|D87!זC{=NMKtU[OYrʖZ腓]y^r&RFPNQ" :R~t8&>&MA>Ov>b$dTkJw/xK:EbH4ޮ Ϳ7odwUDZ{\h0fyB%%byDCKCW'!uԅ*ӥښɉny-w%gd޸۫kF2nmnմiYuX*i=_~[}ۖ1?Ϻ724N;3R-n쓙Xj{m?HL{ٽu?K~i{}Z\;q̗AR>˵Zn]iZ>cY;ZsDǰ.]2n_Zsf~N*i3bq]56I8 ]|}joa>ܵ5y6ЇvN#EKvƑõ׶eaԒEH6 5޼rek nM=U ܈h ~W~;}vO_}n>| ln|ՌH1ژqA.;Zč ڜr]o(t 1I92&1sӲGm]={v 7w~|+y8zoùɈ$D^b\q|t+Kur;\$"&<%Z,Cb_ mntT;t'l]Hq_>2 VlQʦ.*&'a Vqv!;ϦrhZ\pa쓎eC~^[0LD{V,}D;LRᗤoMs Ena WEJ9y5{rqJrju\{!qP/[<7l3_%җ6Vh&iP ,őH/I f+5Nv^S!?[L?yZK$#K`qz_b Fa(ԧj[.kĮ g ZKǶeS}3xm<=ۖL|}yMkj륷GN̦\cߘ<.)/!)c`ꏹ%^?%9-gzLKWY02G*gI>˽!%%QN U4X} l2-K\w8VrHݿkywxw7 =|WǮgn֖pgGZv ^:J>3V. T qºA^uўg 6fnZJ~HwmU6>X+ԱU4 ]j2³%KL~\ǭ].n%w><öHXƦsʿ^{l_14!eRnJu<1m|hsR't.KJ;d1n ZtshK]_mysF>mHlzyy$”V #'Nt9S74Iug pɉoL-=_^+`AO hrV%C ;`n妶Ehy˴n3lȚоS[Ӊ8ۭlg6ʳC~~obm!~+_&,ZѢ]3FimgywveKy>&[gSzxH^9. Xvɋƻ_,Xxnۦ{_2dy \; ~6_ѼӶ uZY{e4V6 YMl}?>㪋ɲ%u;wr]Nwԯ'W$~2.ȝֻ^v;z|k\!C_𜀏o íDlЈbs%>{#65_Ъ{f. =O;ovG'{NͲ,ٍ+c|0P1`BXV4+OZl 6^8jEW ;߾(oCSҺ=}=rWF*W\1eŨӊoyuyE ׹Vi&T{79y;cBô˺ qhe{즵!%׷= {+&군a2dnG5xO[pR\ױoyG] 6^Gl{pִ^^hŵڟ8p`hGũmS_AhзeL|e63\%{ȮwW\]zYJDFF/XtwѢn3nL,=pX|rLz>l#o?طF]2ʞ|T߱N׳^g(5v-[J{^ƈAu\#MGNyq冧}NikvӦuڶ0,3l7ۃ+M/PwyНzjn%viÆGF8)thCu럞h"3⛦>tmx`cy[qg=t­;?/ylyҾ)~wϿk1_xLaCNl}I'+|ms|%k7o v41&T~Qm)e$`Xj4&>:1W;vHͧAmuuhZnOX<}3EرF=݈{D{a]عCiߊ\,w}ُp|YvO ,ڶ=3α߾ƏO/NfEZ Re|鷀`K +u{6gf;-ڳ睍OVvF\zu?8lU9s.eFg Zg8Xvucs6Y4l{a-ڋuuX}%g\l׸q旮{i<>hox'/_\uo6ìe#6Yij1p'6cБ8g nՑk.]nEL~?h@Ǻ-#Ԯ9Q&:d[mW2ԈGuŪ ni2O_P$ܻgϦ.sϛtdg <uo0o'-7=_z/ǖ7xmv*ܶ*̏ 6lnӝv>َWli3kT1̖a꽱4ġ@ZY>e%4wչuO97\VSĺQ[k?~)ܺ]sOFX]Nlo3oj}ϔ4m2749=Ͽji}Zj/ ٩f%9fO]?,vPeZ-;̫]k{xN u8p)Y,8f~ #r7;duN5SO7 '{~q /Y7kV:gCk.p=kCS9= nn CIQy>G\2G鍶zFh0_w!mU[O[kև^w?f M6=>>~l?V̝~紧кw<֫(ww91kip ^khXVLUuZxy)Kk"TV-gn"VRu_̽VKyo#UQKX. TjjmheaޥjzޔY8*zc2o0k:az[&?-[Me>aװb`-g% #̢3ӟl,Jfr29ۭXvLNUo'8DXLDHwB HDʰHAM`9|-<@'s_L<2r0#$Zn*N&剾;[6D :c s䒄4,(V%%N>:ys2V+E˫FLarF~ $r)szL>ȥa$TeeV <RQ\6 10<Wʣ sʲ?*x Z.Ш@ٞ#Z5 $폝܂|?/UF-/$ğh9}lW]S=1Q Q&XitOjA"oMڵjkT9ӜBqk:ZL:Ug Ss(h)HHjDˎ5Z;ɥulqUbηʌњ($!GE t>fCBߍn.ȥ9JFJղ!5s$ WՉ2o~ӈL}>sOi瓄L3;[}KyHIcCվ#Dr6 $ttoR#(hCtF[2o m$XzDPmO)JD+wO.H&L_W+ [$4K(rH=nȱ2rTN>=T?TYFRF*Vqѩg5yH/^6 U(䵹JOP:X ։:WIs\*?Vda 6HU`SsqH)$K35(vp)oy6D~&e g1`)9 mNJ2]&R)i(8Y؊e  V6,`Y 'fe =HО[`vvRYRDKHҊIQ7r$IL^.)%v0<&BfPUǐ?]v\k,1E%`ʰ!Ve)giX4OdR>-5+Tߪ (Gaj۞$H!#Q2v된DzHXE mHkͥLd![hӪvMscWC{$OCNɄ4׏$Rѯ,}N ˚fU}5ZzDG!:*FAJ Y 1FFOC(zMJr G!weC^M#C+GˊiV%$U$,9-,8.Y"TB K,h{sDO aq!4*8JVғt$!ql#XZ$I-& )RGI Eb99 '@Z:F1}fb2 }Ty Y&F3T&rYBlę>RXRj2!/H,#RFAҾ07C8 4t;' 5(]I",Ss٥VfFzdj"Y50 CbW[e\Ѐs$Xx/Ս/1i1rS;=7KX,&dboDz%QXƻFZ,a.Rms-/շUu0JpGIsCV"H<'eKK^&;J@K+ϮƍM%˒Hg& R%eSxV-OCi [Ҟ3-˖yIJ}!'񞎝hv2ehdŮlq> ]&@r&/u'H;Z~GIGt?i` YRM} 6/^0.M|hMUc̜Qljױ+lI{d_qGGnq͉1epz+tzzt!x,86#TMۨh:"JDw2v]I "}$)k$EArY&$c DL='B%Qmeă~IBz<IY=Z}8C~5:F]b`VRe);tȇH"13."h F)!=Z7I{(vpIUиٻ)O.4jE9F%Kcf*+^tR2H$m?h?迿{OΑdot3IH[+Ҿl0%"G2҅FDۓ,9Jʽb.mql` Wi˔-4>hbj+%}WқD"p \4.~10UtĶ|TKDeKQ8lRn–'l%*)Y6Y&d`YuJ?;gGh@kҗ#9),]b@ZFʕEB!9i%$;%9ܨjGU˜T-q- $)4U;?䟟#s8o:J'H?GDz9$ȲI=##YY!M$>% Er{IN$Y-(i߀AiuctJ#hiT#!}0dh_-KdR̲e D%$iy\\!!+]c53ƙX!"r5|(% ev]{' ';VRa|"WUWvOA9LJCk6a#IP栉`x\-ɒh>Ljf8C,UV~ehHڿ,QI+]NJMi$ΔgI_^t㒚""-0[(iI8'y }*OѸE3i RRqsll}f idiՔ+б>㑨HZ8~}o;bg)gWu$u"KDsajW,W՟<ۋ˽($FPʋI9Mq"2/!Z#GF2ir/>WSc VRh/egI2Fɗ7q< 4n7G!1ӱ^]e|iRWXE4 /kH[uY r̃'vD{tN㭟! i슎MP39},'10Q3ёg1 9xt 1:BZ .3UL\o2Z a39l1K"T_k lx}>Vx0'j4Dz4fhWXm&-.j9,>e;kɹqTߺKrfT. "Y+bb|iioKz>LIk'i8$_ )is|T^F?C ΂\󑯳-8J-62BC~莇 &m~Ȧ m66p~=4060>}~HQ4#}ghY#G$1u$ؖda!'̍eҲĵ5<H &A;;r-*8Oǐ`T9HM@e d9֐N$Zۓd9\:o*&LmI(+*K(ʔ=_-~ЈBrUJR.)9H~Kku$߮v2_!ʹaiE빆*Os9H Ym@X4+ m5 YyLMv[8 9* *I.qұ4}"Eb{&i-|~DhK2/nJ"d&njIQD%Ȣl!v̛XRZVROrR% !%1rdw4͵]z.;{HӔ rDR]ZMk$c+E hO[ȞɿL:vd/4eFH CLj;"5Ҍ.6IuB2^/""}/?(N$[9*/IS礡EĕD45r>fڸZdܓ,MV:Jh.GHF싯ڤS)$rz^s_ڤ]&HS*yh3Ӳӱ-GaMy] stHΪyp 8DW,d4Y83ZӋEt$zU\_3::KgsD utL?= ZϋELF iS=JM[Xzu 37KG魩tE )R&BF5"&QI 0Bf 񲝋H./ҏ0Z}?2QRu%:OZ]}hmגO;Ok$[?.K]m%1 :]\m)cb:SćBEJit 4謄RU$6mjiF3Rzx+I/3gƲ4g#OxvZ#:Hrk6JG~U!'|4tB̌263I3sE =U%q8&&ę tn#G%~nnΜ_:Fs:lC7}j&gacPPe~V$O~18ie8HF͘TEzF<,EpJHTD@ght.΢:Joȹ IUz<1d[6h;(H L$;"+[%&u5k%)D{ZJǫyQD~Cy5U*ѱ7~OH''=zF~vZOgO0HJkoGq' HaD;ڳH>sR@|ފFC/ "1ћcI(PJ}g)q |\50'^CZt1_+&J 娪/dTh$Y9s~d?s=DY_V{R6q6ΠD?_;+0MQ6q]əJў'͘q8'uKq'3jxvtN%IH8 ҷꏊ0'QОmY9' mi\tX09H\BgQH#uDl5:mmeWq$m@4ש$UHKhE=I_ޅhߙΫK'I#-;s\)5A$FHyI$Q>NvΥ]-j8JE_ҷ311z]] J **^%,={dybVa"D/ph]m:Mah6NUTK6i+Z1..9q.suay/cZ9>՗+Xwsc|ZMzfڍ$#-aDE@Ɖdr)54Q$aiVQ}xFj`ގX,[7y?3#-`JH=0 m)Ul);69Hڛ(!?.ZDI~Mrwh=q0ۅ9t,ef@\W;RSHVGi4G !hѳlu1+zɳD^% _T64_Kum|mGUyc%V$FN.N2ލZ:NrZM6Ť/qI"dFC :WI=l& 3& I|]!I&>YDɕJz} m&+fDZ?BK^2:zZ9N:IRBb <. m^k>>ZEʓ2-+e~es+5OTmVY,*+f蜭W֍I_uss$y?^R%tD=zLhϝq̈́.ERkNN(}j: eMن ۠Wm݌m1؟o7Q?H]~м"5cz?;3?;Vtdm̋6b5m0(0(Wuac諭sRuU:]LNM۠yR9UbHan~#1 >T0K+,bVg\b1ƥ.K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸qw2.~ɸ4qYZOo[La1g]ƸZW-0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.1e9K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸qw2. ɸ4q{F2%]u=ƴ\jU0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.'q%K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸;Fd\ĸm96H_r9cZr-j%K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %$1.4q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.Nƥ/?1.w_V`\ҫc\Vøq %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸d\*0eyC0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q d\qi#-w`\h j_5.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\b70.[X06K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸qw2. ɸ4qYc޶P_㲥cZb3ej\nq %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q,iq %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K'藌K1D~ re#ƸZq %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q?ɸ 5f0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q ɸ4%g}e3~ƌiikU0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.'SƸ<Ƃq %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.Nƥ/F?1.=1gL^ ƥ)cZa1ej\øq %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸o`\2eI=0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q d\qirwaƶ袿ׯq9.cZc˪u0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %KXC,ƸTԅq %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.Nƥ/F?1.+\m]{/),ƴUq%K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %U=Ƹ4q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.Nƥ/?3.\rHƸ`\ǘKj%K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %$2œ1.Yq %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K'藌K#-`\.0gLKc\Vøq %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸d\c30.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q ɸ4%' *~t6cZz5`˪u0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %KXKEƸ,o%K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\¸q %K0.a\K_2.~b\z18~ Ӳ!c\VeƏ֍Ra)=vݳb|n6yë/M͗[4vd~xI/#:]vi\ؕO:t(.{n¦,_zYo^2@edb{s_wh۵O±g͏lsh܉ʊo7nܰeGsHUQ>sXK98Ä-(ydӆLi8>O##ehF6l'{;úoaϢ+Q{kB=ot5Πv+Zμn@UelcBwibG{(gn='SE"n\A:KVL^>V~au=LYʆnk}7FZnr8JYb5+R-ȊZM]=r֭[u|*)+E&3Қ5҄Non6䧛nÇ[WԗW?^imEnKZwKS {9>f8y}VkWӫ=56$}::X{ú5k^z%_W%vA fϞϞo.8NVOlJ+}7b& PtƪQ(g켰E~g^ QnsNmuJqٚ©eEyDu's^ޝki|n(ϟs$|)^*>=#=xෑq_Y[]hۢe2y^Qﵻ5qu~b^OSeQzSqau 'ݿtiw׏yyZc=u['\2 Sʊٜ d%5krO ī?\XnƬ{]#_~};]Xo w}np{AUHҧ.38㷗k,vjr(OnT6Y%NߺsC޽W~ƖJU[Yߛ]P{&GpYoҷvVԕu٤ȹlGRJe_3u%K 7|#&&krR*˟p'Ou ~:uЕI ?[p}rNFVڮ6V}LWZR]{ي'[dߔ[nXOE:)gnkϺFw Xݻ#dIcM}875k?NuhDZcM_|cc鏠:{=ֺrF-Ʈy{l(v0{xÁb3su{jn*ٵ(/]~pڵO~op]߅ M=ڿIJ:mmkd)GcM-͟_0c8ӒnFMYn۸W5_~uuo8'EW5S>n^ ذ0k /v(70'~tr^+>m;pرvd.Ηk'=H3at,nwnl9$o#(np ixb]L)mr#ڬsvpA ֚\o>((P9{إ@wtavV ~3j5Eg p[n-6ڰ\j6/ۻU۴p{G^8\ZvXs5=?oX}~[w{429HVaouC-lZm"IR8v;f o]\Z|M~3V6x}g(ŲfW^;%Ii_-]m:ֲ}s汣2k0]i6 N_l~mGs4l'⳿vR}-S,G<>E$*!Hws'bʽZ&v6>3U9/w/)mx悡 dz[?xAΧW],w=M}7T~W}:sԥ8RJ^ \z++e\zvq5B%?ބ;wuɎ-{şV]?k%'hznyrt. .%1JZ_9N\ f{wb酮|ް o||rA ,Lyw~oK[9#`܂!]T]VcEsƁO߭Zqw*Qtm+Ǻʳ.Ѯ+tVQByd9;Q[7}LjSz 9бZw84&}_uP_ZX8;)fIo9g]tuiH隯 N+>jc{c9^M/mtJHbw'JO9ë-41N|Xwμ/;>]6B"髣^cG*xd39bѓ̻}ueT}i6&wkhu.Yf+9[<_sRmMxovx ~.I a-SbJ;u4+-|ѡ7{nKov5_ܦeSf_*4]{b|صO*\֘?Y՚'f˲{VC ޕ*07tǎGwWt{jb ka'1⻔ON<3)B9_VK2! A=t-y}_XpWQ ^x?k;6jΤBg3+LH_mkeϣ-6hv6ismCm|n L\~(ʈ}{LQ}+08hR9plSVɓ7v0@ rƬ~+nuldȣy#߈sߛUwfsfi/[}-ycumJˊf67k]G5}lۛbK/?0~H1F n0[ cli'z$E[6KAt {#)+z?`ɐAnh#ykcL7ĩ&-&#<|$1/h|ܱ;ne:쮻hYϽ=\f_|o-NMЬ ލ_|ooOקvoļիݬv W]y>u\V t0B܀˼KnXј^kܖ>f>n}|`Ojr= _P9ljdX!O{uoh3"KpxgƴM%~v;م=wy{e|L˩z8O&nn68Mرu}Kѽv83*SR>hCێH -}z^.wZ|۽Mb4D~v_/z`ӊǺe6TlD-EwWi^nФgB_,KsJd;ѶF4z|=ONֻU뭂q'N^h`lqpefa}8>kEޛ5{Xڇ>q8|帱?~F].^:?|xv: }4Rw^~|T\hV_%m _9 W4&z^yw~۾`]L/_ސoHfXlG׀%ܘm70Ɋ{8Ňxk_@JBƒ% X>_ۯg$$}|ݤEƉ?ٌ?>{E;$9yޔbLTh?s9ޛ=B0Vopn?t;߃~];~yKJ֓܆}sâ#%e2m&9wmmڷtA=.o-fu\սz8lƔ I[tmpӥp>)q.~pŽ-g{u枘C]?-?P?ǶYߏe}LJd/MYTQy$x\QiwGzm߬,v{BV<l~`wwMp9=Ǎ-,K*ԧKԮ[rvK N>:5Сǰ y.Z:>KgWxFu/ҸВ/^/vzA>`!q{_IͭqvqnȲ7xQ'N\uڮL^\oڐW*K0<\y}W.4cclɹu}⤺{ckGG~=0Œh}4g<3sfcÿia17϶Q瑲hm hqJLm1ӇYw?$݉g7o28ߪ|zVUGDF+y{ٌ*_s}x= 2o+.׼|3nYz_FΆQܘgO>3ĮSP3nLEM^d"Dž)*)3v=Lk;DٳACKu^צ{>rWJl E[kHZ:g^v2s|x9M.8>IJTؙii~.`ـ4}o(#j//ω.+s|a|Yfd{q폘$U%b3j+7{6ڧJ .*8'*4HZ?7_˞#p_vxOrS;I+I篇8c,Qߓү(GOG 㠝-llڅk'֜fѸCߢG~kITٻOweUM}?x`j▸&=)h_ߖV.t`z8!pWWKr,~> J,akk63ϩ? {^wn@v{A%LD|~'osɛThڻ)3f/9|I5/bORK]*g~:_GkhX}iOwa19'O]eP}6kA[FJ'Iw,|O(Sf$iE۲?oSVUl-hqOZno'`fi/w]Z)⋶wXd$/z:m^,nڗI'5;{i$_wB7?xW|ǣ\}qƨj // oP[0B҅ (iKY^W%eǗ^}la::V9 ÚNϿ?eɝmR_#,ʲc{ZJA]sQ1_r|D^ҟaјRcG(k\#}{4K.9->?ll-Cx5pĩs9..5_b^:+-Ifhב3;򶓋=FT]t9^Π'! iT6p=\&= s[-O$D;$C^+-qsMzj "7xRgF^HRAmKzj>zfv-`| ~ 6!Q8]fNn 7B!n4vyW}u)14ӢN0THteuM##.A›o:TjZ`h;ymЯr5?a烾!*9  ׼Lv ZτѠ(ԻďȰ>0ٔ;|'z-sOJGx [U(۷Ӣ {XA^Nq.V{YJבxX=vk*ʗ; -2w\mxDsx?z ҢOzfynZa405g1v[OK)7=t{K]bCڑ-T$jk,pRo}A0f,ظMF}k%b7n!P6иsa)qZm< AKnbEq%V?`۷ݬ֯O0|s-jrX{rP$ _o= ^?ĞR11)}I9.cm.͕'8^P3.cUFA[${ _I W9jaW JNgU4t=G޿L `6}>t[F)V::q!g$Ljc19ʬDJL2֍#hZk ΑWꫫw9ԽxxjqgoJzwb6Oo3x/ߩ5fzǩ5>UbIܠn4L- 2Nάylg#BB3tFP+к|Wv^ۃ.>˺gM09&y 6v}]RR e^S#>4eݳgk@$7g'MIZZo(,}r}71 ,ȣrE̵ Fm6īp}*IM_:Ka^G>iy`E]Kwa^ a)ap ןZfMꠓ,;v-EGnRa( =cOoZmOj[gp9ȪcrɐlZꎎq|2Ǻja՟ 33Y`+uRP]оdvd%i%5]Qdd}ph8b(ڞPjgL3(Zؔ4%VL!h_O^s=\5Ltk3(DSMyGdes5_Xí]UH\Z"Q(/ȺHdal)f- 4[Y̋أݯ([lԌ=^n'L>e,h{s؏!6rK;0gRqMwfr]7=<"پVU3:VYhڍ[ˏ(%(uHz!X{xC_b緽vM("ǷY &uٹy0U+<^zBjӶX{\_nLM8o|5㞙;{U_Ƹ{|.JStaJiܣ\S*$KCIBTTSb$ZauA?Lf:t0b&m\{O͝.Xc͗pħT?y+HZȴTrpךHZ*g3r?a!7D_|ca|WOK4q/>fo.q&uB?h(78(R?SD~?52@<-|0R%=iRa5r>XH~|TAvAJkz]5E=: 'S6 Ǩ4-+`,tt @Al<`C*Wj,|=!&m4e[YjngqZ & _9K(uˊ7u*jPGlW')aE8k7 9@YS?]! V5x+3W+GXQέ!tUa@ip'=*dbդ,H=Rw,.ѣGZ--y V`E5%HU-"QQiZIA@#16Bu)Je3PzPKjJ7PM`T j4&G`)hV=`G<9N`Eɝ4S}ZuEDt\сI ;9qrpe9YA >+P'"-UUD<CUBRx)PX jW$P iRJ+~;F?XY XR^>$Ks:wҶF`%YN5:㬬5C/%zU,?jq<0Jm/P1H+KmA"=0ʳR򄠁=Og +Z  XK!$_V[5 M.ޒ~&_ʹ)H T= h!J hu&-hA8@=GGojyJAZ!@ Z ЄhR/|Z,rlujxX:sHd%``-Kz$Vbhf W̑B}(?2 qjJ-5k%~AZ<".g+$ФAJP#XsJHt~AQB%$Yzhp#vU*+HY9`J7X ' r%D aFAoI@b|4 !R⇪@G,"EtC uE4Vk}Ђ߄*"V0{kM-_v+| QXh)uő0_0B-8NB"RMHPC~@iwqFJ Ro!Ɨ ?~&Pz$ׁX)(Kk=r8S viBX-iE1(NEP+ @ thɶt }@gJ|Ne PqYk2yhvJhja u Fpk;M~kHYLJZYHvjf@W84 AXҔ!U@x`FYXj^"VC*r$ajԂh܉LJ0,Z: (O=VO+?ȉhU[hz{ry¥r bM@,-CnJqCK3iƨF☤%pL,r8i-%IR޴lro ~V~[϶n#C~r I>ZlW?ڰ4< I0"dRh0BcԂiqrZ9-|[j=X-Fi#2k0M zJ Kn_Xi(PXsYw,+wPs?T~ڮmZB~GwiBbZ@q]]H8̭=&, Z) ;cƽaLGSrU@I IcUf☔,FЋI F^'xD,n@V, hm#VB5E!˶+A!>iA:ɣH| aI'jr .=ĠVAQ"0Nfb0l#;P./9|z;bnr1ZY_P VlMO`z Y(^+OQH"P9nUD+u9NmS|o;!8[e38lQ=V爳_v{rͺ] IzX-OBX-MI,:Mb&v2)/;MRf~.ڪmx{,|IEm(QusACPeb59 -׿"k]{A$&[:\׈kN85!r냴 Dyt|V%Qz1V*VJhw_& ˿\%Ȥ{$9VR:N!Zz4!%#N i{:53 t![5+P>Ĵ Gt|pO# P~@mF1*>gznΠޤSOcdٮ[_FXIBQorYHB ދ0zB@|Cyvs6p2Di$"D AK%j`& EZh@ukE]:Ctg@CK#4(Lhx`BmH6O ۮ(и@8AC!?iṟ="ll !xS>܊Ƶ#n+>i#?+ .wymZԓYkn};;ٖA*mPjiip5Q,ЈZqj1:QFA ol$> ?`".00|g cGȽ*e;&r'px JrDLj<' 㖎X-"ZZ؂o#xHjCuG]( ID+I()plP'iW>-믽+%heBjNiEl3M"ꊕkIqZ;:yse[pv=D;Cz<Lk#q6LK'h2xԂN8ޣB{UsȽ]9GcJ#6>\8!W6A6M1ddm\B. wĈD#1Bgx"p9TnYTTiO_K)X~>CH$^uh}Bj;qNl +S`x5\6^F&;g*\893plr/JM #~䚖lَBa-~" 徬OtY| F}~Lr@؃iVP՚Vt'jjqJ~c<qM~G2Ԗ[T<#l\1LNtﱢ!?I-,c#ΝK, X痟*a2^ϓ8Q0eVP!݉xۯTS@9}m 1J@!8-,(9b&OG?q5~D uKhIDGng%'kǶ=6Ŗh.Siufzp\'=>q$-+GA]vV!U>!ϼJ0KzTD3<f-"PG@߈w?<NeEbE^<w "y34/.PB/l@qxNcaO($,Ip_ϧX@WhQ'@n"X*#D7XP0Zz( V {0{aq0]Z_ W_@f w*G_WυA`_ fLWsexv(m}Ϻd5?5[qBf{VjeG}`_ 'VffU`_ ynXi~yEb ],u 6 +~@\-0턠;0*|z<];ߵG<@kRՉGw>~r")O; /?>|OIv|wp9{ L4kvo>ҍ+ wW7{T=*J+rDXc]XSrGE|5H?N&{OT^~ 0a0"ݙ|O , ݾ%V%ߖ$s@| XPHpߟ~@^}EoeA9Dy1p(?xx>#ץ\FX'Gn"q^N|vkt5|wr"m F=ipVY c~up(a<?r6 >7moiX"lDő$O͆r"'uА_?H ~| e;`C6[R/YfAb#@_ ?uj6br y@@ڐ~ {ί$Fg Iۙ?tPr?YN|?r .4T ҡUPD/ 4P(߭@CPo 1ʋ\3%NK(kb8<`xB:T0 --U x.]hnY~lr@+f-1[{ ow9#ml"vuȏ~3ys"#iSi[[`/lŋ_+), sWr@tSzUۧw,N>T~#^R`gL3vi \=6ڬ߂\ױ?L1EU5U93>W2y]IenH Ȇ:;qL~j߫4gˡ_W>&:Ћ:0S x'whg]ߩ+%l5LL 2TCy|噩F73,۰ͰKh/ nX+c٤:ƀY=.ɼL\*;Rλ5}K[^ۍE쑚Mz8axc&)CubwE1XZc2ȱu㕞Z;+4fDZE> \IuG(Td Iwe`|s,5IU,u׋WԺIg,5 %$T6&K~fn+ZRf;J5=6i6l)pSyj|k2eW{޽˯=CZAMI47u8tǾfkW5g703Ɛ\.x 3@/~Y:,^33Uyarԉl~&2.}-Tv.!Ldb3uآX8ۦЦE'zl=T67ݧgf QA>W".}%ғrj_} os%Սܳ?*UQ|1lLNm>ym).(.t|mYr^0Fs>{νۜǯD7Ȋv}]ت1;,SĜ]ORa7UG*je{< ?cUF g*K]uL'8U1 JzMs|73ǧo4:}i'klgX7n:n;,t3R|D=FO[LXE38̽_Tv7Nn_eF E=u蛱sjZ?eVIXY樷eo6}YSoҊrMrz2j`qqs٠oSM[RɏG+$:RRnճ8qo'27_-=;4T=:8$ު3oV b9&w6,\㶲X/Zg:.Q&a=s!F>C&`FF߬qCMbx/Y-FTzۤ ]ԗ߇ZLp{xbF^򸇯qXNm}W ?:՝i*pxdV8W3o`lꛟU ;5Pk/dVhs,l{^N-u{t/\Q}U^sUOG_&\U7ɿW{^]}ǹk:D  ,ӵ6jȫ3x@d\3bo J`*,ixt%$FwVJw6XmB/jw~Tw:YJﺷY\ UdQ|zY0E]/Xel9+N#Ų#lΥι[Vc=ТX_̢x#F6Çsa4[lUKg$J!a>ӝ9^&zh~ U< 4i=b86pͤ˃`<RCvJmk!/U/u8]lͻ t|/ɒL |vı9+YIa/=f4/ 46w%?;ݝn ʆdzO7mei{d3Ǜ ڦ* â(̐|Ƣ˔ٗĵwfSXںW]o_z4r6͊*2.4ukm{Kh>^V*4=/h5PiV 7m .gmI̖k9nW}Vdn^GY-&"]8~653C͙#( $\o{ϣfcW{oxŎ&Ѥ>rq|P!]|P{G{f~|eZjC4<]$0^&޲IDӅQf>!&BiǸ`K:c7y%<iDFt$%χyĘdZ. 8.l$mm-Ҿu7$:_b;E։v|x^l+Oߍ93nJɴt 瓪=A\ 9Vܪ).6OrDo;|6 (N~uԻZɝT6Q·:CVӆ,64ջzǔ_ﳶV-1ېѳtY%ۢt_2=VvH僻/(售,l藄^I=NF{8i.-4]YZ36y-T?O`{Ξzm/ >8r|%oznm c =^w78vIEU=i rłS8Kc5gczvQɚWk.z9 ,svUƲ) wjFu߉6!Zۈ}9,nj[T9i{{'vZ(kJ޹Vd ۮ/}{wiz tSCq 7Spl}Se݉A^Hbl7aۂXo0徰-e?ѹvΖ]Ow~~Qr-OvzRp䌭Ii)-B5tg/U2,Er7FЧ')HzXËsO?Z^D DUrNlRu:&2W\ko `w?(&دQ<*n+n%wcldOk5xdUPYaS#-兾hQ}3~Th}O:tnv("X9OŞ^fTJcm\)|-kSVLԮւ(ֽM5uXS~ܸ9=-g ii{O[Uj Imotv,qmr5imh'8^VzmiT ZsQ|'s4Gҙ|qkp5br<֔W5?}y ǖT\}|O&8]w'5:RRp2TѤdPK|)ߵc_cruizaIkxޖvRu<{SW[ 0oR3h_Bg2x1m;8.WrPgwz=E^LjQ/|#["d'jlE^qhug1Feg6r$\NF/eV9gWEUMpKfO O E\;=>NJ {kDmW:^6]],( =O;kbr+HQ IL{gH]zR: ሢ} MϖyIRMٽ{Pqt]]#Ή, ̌ zvR /-`ȧhs.Ǭlu;Ʋ-"K IR3xPOohc['Ty\W7~YOQzݛi9*rU ?P[1e&oݕiNxȿH*G;ho3Ξb6ٛFJ{%{ |EEU'*nfv֒qcטcxv!5j\9*|jw0d:e˒NZՁt_ᒦYE?tJƉ&99a0UPm _KI%m!Z %ml$v.^Gu\-Zi5TYO;2ѨeX^|DPqھ%Ɣ0"JmqgI^WRX=~1=q|V5èw,xC~=fĞ$yK㓩7 fh;3yq,߻溝nK;\S:5~4ۥ'9zqƟ6L4(,&3k*Smtc\̝A {xsGTÍs71uV~'wrD|{tK*;b{^8m骴/o=c\4G;iGOPg߬E}RZZ VԞꉽ8> ~o[,\gZ'p%l/dt`gvwKq<|c+%dRs;6-~31Yoӄ93$ULCԼ)ׇ? ʐd>l>{W>\ݠxG-YwŸWRU25I,q"˻0#O[[m%,TzjB~ 7iw)^![OoksYӞnoL|̮ܺB]UN $*=-38jXMsXggMRp* 8 Tn-.,)wޅE?ݼggО}A\'FvE{?P4z0$ķy9jo;vjֽSp2u8K=`EwB˩v= uwdS&VsN1c^{/2u6~NB`D5{Sv?{S)ZƱqLl'gƒ(hx/}>yO|ߎܽdh<vٕxE$2/<j6>kHkڃ[:>I9LqZm/ \q4[iyO.jʾ14ƴ򎅦^ =8MA<5g|mbT%pzQ+*1;#TU_|$zU& ?2Pױa*6\'>.}tZSx;N՗fHzߛ'v7y)>_\ F9^0eȦĪĴKIq.mnOb*{y4⤼2:H7EeV7:ï8c¤oUݿ 9̀H^n"ԭi l< D9\i!t8vֵ6]`f r:aꕬ_0 *f,q4%e=m̟gI|hÔ;RZ614CZZj3jH@pZq/Ezvr ̴yFcμ/#>nG|"Dž}tzpww>f 6X)OP\-r,Jeua\h0E̒)=*O2%.u֐-弃Jn %G(g8ye ?0%F,`3𮝥1X Do9wjV:'*.V,>O fL1̠Ϋ!pz!mm]Ғ;iUQͼ!G >S W~Qe8o-["OSftEiͧ=% e*Ǝ1D_QpUg)$D$]~o|:*8UOc\:mZZZs+­sO=+|JuOHn$81x>꣎ :.~ַ #6:r8w5v#օ<IG{{p}E7Skܮ v_dgw۔QQEAw& ZxIڜ;)ʓzt4>WiNػNuVgS$MaJ2t_di_DiwR^zVcx8 D h*:5' b('$U'ˈ1.֎5W$ڹȿ hwDk6U·Sk;m󿷘%?r<ϐUy3ɭ>6~ccEi6m۶m<۶m۶mٶ}myn&3;IW*UݩZg*i&I%HE;SLL8JNlӮ@S]$A‹ l7z8`4׽͘."Hht#i7r3+G`tU5o4 D \|xuYI"Me,XGY՜c /(QW!w%CE hZa\%[i5ªeįqkck~k_3-+ԏ֩U@e\_Ɏ]pկ l;=]HBM<> gM̱? fБ5FM砛zm qL5j_|@@m$ 0i(fY:gmalUQ7"Z9Jq{]qVVU] E)] BR,1rR:d}bBq)X]%O[f)P|92W܇5^XR} W6ſq&wbr@2fvf)4:䢕!\lA*~e `ui9[%,E624NEE I-/BfJzjn賂EG:#Qg(L5=M'>D&E,mqk>s,]jkw'MM`;>~L L92x^ajl==L]ޯ s~tJI˒hrߴѵqky gOԛD\l,,6Bt5 [O g$ɠ(gu]xinst>(B34 "-? c= =0t>ޡ Oilpac0`0l_,'UQ_$}z&MVpTm;HXw6)k [?SBLۤr4T$6l쒁LyX+(ԓ3Zv;dpMS5歯Pgׄ<=q]y>_ƬO3&R-)')5фCXD8I0>~\\c#@ 4]@,]LUJJE SՙDk&H|i5/,UQ_Q128Q 5nI}9{.[*f-UkֶA5i6Wd:s_ 3 [)"  vv-u6Gẃ;j9C;bݭw<.+'&)JtŸ(5L}. / sp!Æ %_{|a%mͼ|%UdЕ +@֣+(]:텓t@tړHlcw 0w ؒrk@8<<4֣M[3.DyxFt#yɍ#2`-a[&7ip";昍s!RmECf,}J`/*ö7$^LF,x$ O[ڧʼn3HۜHst=6w(% 2@ƹggHZ}~cy ^1<$='*n9O-@ 'zE Ђ{0ۓB2Fu,USqY.*ZZ̉D9|m6wFhyޅ׌IXeAS)*U]k|Ϲ nL˜3QIMҝy4_#YK+7];1ZWbcc\3S;+_*a!Ij;勮 -w&t?>7 yQsSE'̰3 K^M4M%^F ,V*'`fZ6KŅlOr:cW'A У: zi1JOPOжgQ"Q{7 vz뾝Ҵu|&Q(t[q|Ѡڂ>+`ei̜-yof:^a4(O0H1F-%<A9F-&ܣ.ɀL엌K:rr09K)6_dMmg_[b`yS ~N %^FZjZ o3OwcȜeS IE) ^  \8l?+w9H’S$QvGz-I2kTNXfLU jn^R+Dis{KJzmŹ~z 4ρ x؍'4yp4H]r/k:zsNd(@1= -ryawjvQةNCf_wxS`XsqN3KZvQhAJV|>Gh/ꨤ2lg baLSPO^bsgeDa^O-K*"0NnFx|D,槺۟UɫW@ Ǔ|!*Ѷol5WАù፬da勊YGƔ Q6;;ae̦Q Zef~d_R_`X ̕CBI2%οp:= ۏC&@)u]J4 K#~ĥEgLuЊ}Xaά>?TQh3<^~Lޛb0gqfVWp!0㊕8VŌo8*yuPܾ9x~4GaM2TZanOI%PE;@s5<,\U+R@CQO >=sF]HOr=gF*C螵 K翺(/H:=J'V6/"BSG>h)|G > OBL/B˿~H Ld@eR7Xb)'%dsdNvivv|xM3蓢TazS;'m_f]6ZN1rÕUo#%hc*{֌/C;+JuQTva䂲\[5W|<@FWyWʸlST,{и2(42 Ͻe|{ma:SUXU#h:EwH5T}CxCWw#8J}q303şW ,p֌gnT&9ui|㸓gG4~oLT4{ϧ'n:o.?aM:PyWPHIhLM!粞8m&ݡ\>p ϟ'Y7M?62vJp25Oz5|i!"d4+@掵&C̰(F5Y?oUNI&$)rRYa(iX8q iI*: #q+aKIik'yHL%1CE+:xۚ{Ț-K6躽`xT;lr8A)!➍c_ D헠cV`cIm}PN ]0ː-poX}~&e@*# ;t2dW` ܷ=}|y5s$,c-أ K;N k tОb h'hWTUkIi6sye_sl}H^sC3QW!L\nvM`3JE@ƅU' g 4DgFc G*ӇC PUX]f`Yh3 ڏ'$-הόKOZCf[OT/m㠝PpBRf'Bs+S*Cў0!Qcb߰ Է:㞔~nbSv RS;cߒ RϾJݳc5?K }I5Jr4sFu,zMiC`@7?l49KTuыD<&)gEׯIjZ(gK>(o c֑Kw#UQ "'khn$@EQq_ ;lwʦ3MdTTKW˥>KIXGsTgY&D6ʌ;T'=%y]*-w>h6EK܎8љ9i<wi47>iY˭~Be$hP-|6Y+)fvsLK @t_XteN`x :QTjTZ?5B$ɾŗe\<9PY8uiIkhFLGK\ 8󶩍=# ,EV6Cl>l7R)̇iǨG2|#}N'&C'˾l-`0٘CEeX.%C  %-C"Q};ްPxP֑NIy d Bpl1H`Dō%ScDt4i_5"65p1w]xISB-):`;C1D:h[uvk͒+ ) YPI*yl+,JdzJAiǔW.4:wȋq=E_H!#)BNJT6ʙgms><7cʖ2}7dS䢉Wu) ROπJN|5eʥ}o%֫kF|Kꥌ=H ]W4۹j<]6ڷK[gx/Ѐ]䟗هOæ |E\MIS0#c-x`dS cs8=@=#IqyfKăO6&u's9.XўLqݐ!Zj\<ނrϛqjrl$RGeB, iv?injh C!2|QG. ͳѕ0!#*k_ZE>ۂ5 Dar>Z^s+9$+ tSC~on;Og^?",嘝1n~ xzTEx&= (980KC0$jŅo]NϤO@_AwT&'/yھ҂Gm{(tԆM{(qufnO0g=[е`ۻJ[Riuk7><ˡ|:{.L]o/oQEuAmMce,Z5_Zlؤ5_Bӭq Y[zL".[wu9!s"PLSJ"gY<&ͽ{~~ ILv\Y|y^xWIE!Dj?@iHErv̮tKPӠY/,>Zg0Q׈\ Ds JIT=@Kg !Q^:3zk!RxxoL9+FGr"h)OmUttZ(:γmr@*rF56LOvpD! Ha6yT-, :O5.i/g*AAX +V*(oSEbn hrl]q :Al碔i`iKpfV|1Ϗɫio3*sXŒppm$ʑ!Ӆ,RhR:}guijN#{l\O3vu:HKo 㜪:{b7E1`FcĴLQpL\jUba P(xy̢5U!5YavS2eؐLGo0>Ah`&S99af[&gNJ\ٹO0KWEY<2IKup}蜈8v^ؑLc}u -C~Nn7 |D+(!*1 N4^{~$/=~2^Ye݌ j<3\diH%M *UpAB,`+cXwxjqx_E R ,ݍŸ9W>hp٩wz 7 _nC;9R2w8K:TCm Q, _ۚx 5v1{}"UgVuAv.ىa_޶{<< ] "D 4܂p  E5NcmZNLݩsDT%Ft1{xE>W"۩Ti bغ`¾@G :eJk&ILy[B?ֹլēקRgA!dF1MΫF?x6Yq+;hؾ:uS$RFc+fg DDR} ۿݕ3KaoAX Q鰽.']جNa *^:fEqJ B6xT956֪^.}>131:k1oZ꥜q4?}ճ.ikK.Kjkڅ c =?XCX/ۅӼg,2qt&Fn7e ¶!6oدW~"|b&Y̋4/[ǩN6N B (aӥ(Mc'pZh onsZgλh"7jf 2NVat$iM!IP]ñi5=IaEt&L]b2:Yviꌭcu2Ta%N8G^n1f-g CRh?/L2Jy W W`a|y+.R IFBliBOK <=}V Ωvj/q>xsDzޱZMCL؆ 2l mT_R}'l"k&=UWU? ,Os"@e҄K"Vꃃ$&Xlu1I] f! g ;,ݯ+١xNGFa(X03W~Is& je0mJvئg^@|!auCl& }W#(vxkx^]lCU6w<[E%Ẃk\dTN(,jHBuR!HnUSZ1M^`DŽtPMwEm5\e){ TGfZ`ͰfR5Xun3YWm`MC<FƳQ+NRELR MR1d&Xg0w, 9"zd 6K(:LZ{O/[jw6ͻ6nFEbos]v};F-e"?LMᓚO[v2Pb?ގx<`1UgꦊJ3{"1&5!U濱tbhqB,2th]q42lI^mz#DX7A樛{+ͪPPM0/Su2`k'}V  c̹{-rAM54#Qv]Ρ_3ә&ɫ1$Y2kfuẃE0o>ox l'B(Lv1&BuҀ^&7U !Q nL.ɍWW娋m|]Kt,f=~r `RxKzEab5bAfiboeҗRY?멈\Mה2sQ@dAȤYxND [0U2$uyGi^P^IV;+.!iqL^4׈\%nvOOjMc QToI9[;LB6~߳XU`GVSy`b?n s 4x亸ߑ`5A)01094anfQΟb!guOvAK5Y AŶN0[bc+KRhI_rBY]#JKHS g&2hǝf z+GM"YHH7Ms4W7z6{y @qQEuLϐS|abdm3>=mU)^Ały@jjjG͏c-3JY4kñ (ӄ6rΙk}l 6$* (B񉈙Vj%%(~N}R@2$/Oӈg}@M<`gjT6jRRFg'>pebQ]^e#HP8Dl5w#ӜCs;uқ[vS!&留χQLR<+T*w!%:纎]X|2\(Dk,A?zE/v;+,)B LNI}K,K~NY|N6hꤛm9Bѫ(ʎ?d#ak>)O7 $WV>u{ѡ#E  ]6Җ#m9/l1 BG +-o:-Tg;6S6Mq*VMH{`T0^1t2nkfcY~+i >V 0mI@kƲʞX ͛T23ONǶ:9&24f(9e"!Yta?Xb]P oqZQt>5XAF0"f?O"-e<%P" D-5CѦr ߘj ޱNv׭" 2S>'`,wM|@/\.>agH`pG"Db,Tf `{0ho=y)f $ƙΔ}nrʿ|UXGJƢuSq#0K. ڸ$\I3:b :ǭ|2ʍkQAg/E-U e=(z@ _*YQ$IǠDV%LrAPm;y X !>EH ]EeW4!%u )*bKop G~̻wS=p'PLktvA`ocsUOOEy]J{(PHV765mx{v3cxDz}AT7AB"s)J`u|#ҽ6ܿX );@3>R%P^jfw5!`q>vmCfD|7ҞUfӈ*g 䨽==8%Bfr%JIUjU2=ʺjPsԮc6x~՟U45ȦC]z_FGwv-)v5yy q.v*?fC P44 TF3eXZ6֒AP^<(, ˚o$C1{hڥz;(ZBXL_nj"XxyxX[v6q*]ט.zj^4y%UL8,𼡩rI,v7#r'LYpHJPUϟLĒXt]+>GngB_gBYRMΣxDx; 8JobRud+-ӿ\:*)ˈjc- =(6]*a 9`T+=L7?7SW>]/ 9F=vc0cl }+]!k+Z!bQ*~ZwS9=`n+*qwVJ_9V5]wd^>P>v(A)m4a&MBԨk6roa,O*t&{/RqOԏd^boѸ(;' 7>DũBaDQu[ XiPl,1T—HCrHO^C#&@*fpo:+_ +_8egшJ`,)5st#F2$ a]~n#Bz"H ֗E.m etd J`:>=âcpkhe/T5`=~a[seցE!wKNxEU)f4 Rgf2]oww+J7Sm+r2ڽAxzc5-苭/aIMm4?IDEM89IXm8IDSpƓcʂ\O;zjR[ԣe%-'r.BTDgA/VF>?wDusup,jLppQzq,N_n> w,5BNg 9MĖ,QGw*<#uEj<hλ:V֏R2yL1?GES]iGP3żEX(P|ΌD.s vа/SAAF=|` Qz=& +QVoYJ^qjædL{a7QWhH%F1.e B9b>ں48#4Ff-B=չ7|@%q AYߵ%)0B_D 36Y}vw62~1>5>wb Ф`]Y\khE=X-2?nvEs~vsQ0:K nzbO,%Th*ZrYv.{ZSvӔ <s~hjTܰ90HNksoӼLe5S3&,? h}ϧJ44\*z8*\ldn4ؑk+ڎ S]E2zvfSXӒjz◲5Z Y~_OG1-ܲq l4`o*TV@}gʙ߬+X4կP9diêw,< zduP0 w>cwBԘ ߕ,݉@$s [ txˊ ޕBZv#3: l]EbÁwK*Al[ݍ,E39zqZqgsia [ [ܦ&d ,[{-B/jcP!g 080MXIbUGIYt l~mv7:Enw펔K[x_x 6~ JYh A۳C65k*ѻcf 7*SOSŭLL; sC0_-rkzLej00I߾7Y@k`"a5أCBZ&YK%<&"H lIOF5M8%~A3[$@u 7p2L@|@7zI֙g㏶lPß/q4ki |Ʌiy#޷7GhY0ڲ̤͠m$uftۛiL̳ )$Fh?U5mw禮ѓ'+J\ih\*pވ\AVV2G/oEqa?/~oލ|O~W@Yq/)*'R+|_>/[*}h;B|!S+Wxg:P&ߜ)=_/ PF"̧̇*{6+O 6o]BĈo.Y@kg =SMGRY)DL8w#U.1?"P3KHorU9ʨrkcuC{ H#g][5巚_t7y޷e8)nc'4@78Zɢ922wzoQ%HivZf[J2G^58Q(Xsw<4< bRQn??vTV~Ayh8OY%'S2gR aﲷR3@+X g O @Z!q@w8bwS/_G.x{/(jKn@&QVqM;PZ%UhO<#$ՍMcelg#npw2>t>Zߖ\x(9mY{㩤隖Vzh&9s>E9l<2uM.L=]N3t> zhg"ZJ/:W_7Vi崚- >+ss7|ٱ ׀ä& 8X_;S`-TqcI1}ךRTnXOfjZvĆgeaz_P6rU ̗w(dOO=krI9;_-q$w I5JRYJԫ5m(kGVD+I -'ydݣUIWmg:lg o qRƂ͇eM>fɵ3xD]'{'r?e[9LxhUN?tyWAC)C{!jk佰&^%I$~4-y@ye#izzN6<6ۙDX '@!m\˥amȥҀu5zƣEټ 4`} GGӀ@y6<]}m%>~1k3{}V_rV¶ʀ*.e1k8N|[lF\/ Lڶ:nk* "L2>2͒ W7u[ݠ6ĎgUUz"Շ|7]YRF<;Z0^nw< [l0UiMfIC{66\WeT|;khI|P˓pUayЬwW%qz *8W,³t-)(2Ѳ$dMִz:܄͒d%ɒ) %H%xQzZ;{QY%uD )rxw ?c"VY8PC|-9Ľ` G% OZrZ/I5Mw(HB{Ahu58< < T EiGfY?v [ᮚumsme"H&kq dw;Pb"hkRUFX6"8=v8̙$ºNZ&(q\#9#.QȠJ=cMΒ+l̦hFjyD$w9_A'-^7K@&C:!EH竱ZDz#/;&9}?r29Ҍkh(2dU*4av)-s o{඀Զܢbۚ%x^b=ui=vƫMKkps>*I/HA{X P]V~.|}8J&ąxZN뵲fDAP$6/N= Bߗ\Fzh2Gʿ!!%?R#HG{`izO؟Q߿}Ҫ%MV/Ow8o&?F[`N>QV֣hWOO4j. W`%pތT8`ALPlm|e9Uf$s.^૏?&g9 +ף*:`>@j\æs ktlP@hP`n\rA"&mXA"?a}DM٠31՝"3f!o rJ%2AOJY?-g@OQ.{MH !:Pߑ^B$M^CҢߑ%dl_#-P]0K%k&a㡮B 77 /^3B KAzg靑֛Ab,sbHbDE7| mұȁIz{:Hx1Ѷ}}?2iMA0ps5@ T`Q@JK闌Erh/CY4D ߤ AhR)#ex4}if }sx4kPc/x/Ra£%q-t5;Z6U/4U[VIi=_Vk; %|uEЦÁ]CPCP&^?.B-%[q)U oE%f#iK0+ҼA"ug;Z:g`@4ծz`DI’kTWvSC> G t~)l4Kh_gp$/N|GKNoKݡN96Vbyl㦳4Ǎ?x0E '/iCNJ'%W'2f\e7F^0d<:қ)h ꕜx(gk6Lۙz7ca̎|=oǰI?|B Ӟ߬VHdQ.ɼ )IMX%fRjU[JtՇJNB ]7&yFg|d'}ol0P kʄ9gMtWmpiݡX_S(ެ*7J&E3JV(Vz*E^7ٺ>3/c8K )EC:O7ӫ3 kI Y*.d8kW>ՈMLL*yH*;nDNO&&vSdU1YW@3re`Nhp]~b[N;(U&%lLgX˓R%i"i%HOfA_z2AO1M"|%ՀVۤE*8倻=R_sTcy2UUVsg4DCĚqyZ\vFb~j\>LE!ߤ2YoRJx˨^ ["(_PEILT}T!+1/=#ښGJڳF74CFw AMJv=nݧEݎNľc|;ʌ'#I.ͦ%Q:φ.;]7 ?_7A]{~R vAq6w|:Lxlˤ;Tf4uE?whx_IX$.V%POU]bL':L_dZ0kij=;x43 HDuhw;Azs=F(Ƚ$$೅c˽Z-9!+@,',TjRU*Yt,PWm/^*{%*d"~Nۡ1 y&2Q=;#^"k&>LURc @0C7,yw'g3 RJxOqUW;OUb"6N4x{N.kdtKi)WwSQz_ęk\\~JVjT\cבV=iE3֨ ^#@ vd+5K_3V tz3j*J;0ìS7Y2Re=( htt⪒gLEɜo 6T=Sd ,aT>{?`[қ Wxwyw{ -xV=7A7o aMzˤ@g|U@ЄJ) ʋ7 ]rF}5dCY!&@vșV,Ajcdg0uH~a6FRamZ[0Pw rs3KNi% ,$fgiZ+,spAE^ sy1&'feWc s*B\<^ sRt<휴,@l#0&~78rjeY W_UnjƓY!U!^ Ttep5+8Ҙ %4Qx٧WJ}1P38KYrn`6KԀ) sI8{=zEn6<}jL H_g6F®]@J܏̰ӿW:Hqܴ:H=眪* ۮgǸؑ 2}ы<VPC%wt87XwET H8hFreI=0`F`;3Og:"i#el>M*OFk]w2W @tqjwWah*j ]>7ZMEZ5dW@{/*h7@37HR")xAUSR")DݎX)cT?GjQ)*̛jbU-w2[jbt5QUSܢT}eK<=K~Ϣ:` ;G @h)T4/!"0'}8]9vGZ)$݄1KNc Yr*H3AN,2㖆3̘22g^nyi<:gk Fp'kdYxƅ;c1| o11.^Ax^U}񡚢%4#06_`W8& fDA|yK$AI)&Ehޝ;H ,3n7MX$g]/?MM^T"XMASC7\WPUZkɹCe3 B6cd2=@E;G4JRBWc& 97{|I;*4aè̖y_LU.UƘ3=i+ѣLU)scz7~@\1ԵPޤM/1*IC5g{)׽Z(֤5󼶡ac3Q/: ?&19je{Vn3^QN_P6|SXa 릝wj㫦UDʻ#;#yȳNo[D gtMm&g>1%lY?ʔ -U0&ͮ0l^3{t>utM Zkڑ|tRw 3s'׶궶39ˉ F/5Kϒc5S3,q^ |} OE|Q+^hx&_Q~*=āyڝdGS?iDV~O,U1 b)q4b;KƏs8g2> |;p Tvt$@t@{UbzMƀ ì{\L$#7@i#+1Rjg^w{Z;o~ޞy/N@e5dKXs<'%% dP#h5%g'nʶBfb2X _o%@f`)^&t;LM~ll:B411n+G˓RyQ*ܛ?7$Eu5n泠U TrƈI{B-Xᦳ˔1 l-b廙kU Ĉ\W A>猁╁u*un&/y.xy9իj/7Qc}WaWR&WQ)M c2_U;g;: YS ,VyyP[^޽H5n \m193r\Tp5Qd𱾨+ g}`\ aTx:-4ՃDD[G4L)s9K".EEiYRUVUyU)ЛlL +LMiOm CY1>\4-E¯12* .O|?s+˭0k[ig b` qzf \*oV qgRLB{Nϰczⷥ5&ĪPrȅTs!2s6rI-^Qk3!YdpJ}+&s͇s'@;6MVmuabxww3[*F`ƺ٦ UX$\ؖe!ӰW܄$ l&aؕsv%3P iI?CwqwFJ'H+;Ñ,oqUˠ'zbJP_'/yc> IY$-~HP/?fRɡ \J:9ʸ_y6zʹ>'WP=O:xruXs0'%:mҘQ%HyWQ>ʍkTKiJľmv٫Un70rM|)z&|MvίzC|M. U#2s+@sYѓB&=,b\ cTɰF&#ov-WMz) 1+a]O_q;;t HTaѮSE@"xqqBӎi_mvf, WǵKFG[H7,MblZǓ I+x jcDj7RŠ1̞|²&d;y:πf9GĒU*{M(pZ\Ċt]gzU\TxȸX^6;gyNfR*p V$8}'hâYև` [ G,i ǘhIX?5Vz1n3̨lƜxj,TLn!ϖ;fLI9n=?Z'G, ZU$)H\Č!K\X+"b8h~_#k%XMIbz~3\)$no[h:1` 8ӳve?I|-В7F0_M'_\ge-:)$Y4vspٹk/R ;ذ5BՒb-*zȞ{`+]֮Mms A;ϓSJjχ_Mz bQOh6rM&wI{XU5bBL9K*,`wU/lَ @do8^$jV.<Ⅼ `&V#%ْq80ݜ jӹ4)lbQή,s&$ig J`Lzȋ˓G y.+cT_JCnj212-:)ON \12Sۉ\J,SUna/ѷɳukV۱WB{\n˓URIӷTxaXa%ݒ8 DZNf:9 K:rf\k*)ki==oW3a,w2 Zojp+&5 ĬU'ARXIDR1wc;x8P{3U[jLʒgMuĩ4欲gVDUm[FV$ՁZT:uuqT"W99B3 +HQk߯ J {> mN;_)FgB|$\IEaH ~@ `aw 91D}Tuz7 \ 1Vg@_2SkΙO<-6">O 1iRK$Fe*Q$*x{ 5a&K0G0+*΁έpX%cqU:Ոor~ᶜݖJEy" I`L5)N 7F^3}+,_J՚d5@W7w$bBUm}(c:m70K*y0@!YT7Ebbyn@Ea["Ƙͽfjy)kWRxѣ3E*)g[qcy Xɪ{Bfh{*yTR:F v:eY14qv溙D|kvETLrݺP"j}rUH/ F'I_5 T%(E&}(o6:pCUI\%;Հo ,c kYW˝|hNyvw/]2E4l*f4ӪyF{9)7"[KAzZMci\(|5j_y}TW{참Ψps Wwv+aZEj P;(ɇ4[1SI ^lO;"Y@ 268GbĊ?2Հl2 lb@ǮžݳWimU֋^Л]b:wxkmDZ=FD_ .>-Nn׊d nU;w6ɋœq[ׯx-Gޱ1^qcO<ܛA⨇2sy_ x4W7Kj%H;h[H+8iuWYi#MǛfor:Kvs<O2^q>+fƍ)26 %ͻ+}tW]sƘyqM͂*o8 ^/ U L0M;/ѻ_]8|J-~g) gw=~W1&2bg`,}+yg&&SubegdÅk۠fkYHa;F~ȣHoEH*{&F/(1Cڔskv Ȋ{<> $|N@9Fdp^!Ȗ-qNFm#/ٟ٫FOY鶴pr59/^<2ZV3y}2ScU1P5%a^XWWL Q \q S^;CG$uVsqep*#sUg~njwzuSI|ZjfXOP'ͲT|]*g=bH?y[Ih{L>C*8e{fJf |;tp{\9 79osfGAU53W|xy'Oٲ~W0XM<9ݳ,]s+,LqOIh? > W[+\ƜmcT7^'脯>Q٭gu~u17ywvǎcd}8ֶ֯wO4#\TzewzQ T?ᓽ!M_0WVw?rGsusg?;'S~SߺSݟ}/__`o/o;?8no7Aێŷ}~{׷ ?+oÙ_tn~_{o3osool~o߼Mvݿ_vs~")߂-bs[7~p '}(@:8ׄ#o?xw}C} n|dOOwspÖñY~s 귓o>M!i~s3~|VRs/U7~y~ss[Ggq|8axÑǼGڇ;OS}WG/:ßw8;qt[;_s>;xEgxw"UeWAsǵgx(gO1~jW/>G[⫛8VO ^|{6O|E~qnբ}Ѿ%mlo}1<gf{|A=n6y̯M}oygG/j_: =&~M䀳~moq}y?x>~Oǀrph Jo^0X̧5Ssٷߎy|(4Ľ6Dd un0  # A"96<= o4k?6¸) @= 6<= o4k? xM+<+5x xս&2!/^ ^L !"D!AB1kMjE,ZM]n^YMGZ"6JorE}+3gfvww3/=$4}y<%St7^ cK˛21c{WVe*ifh_-oç}u|gz{$v9W?Oe#Gnj_Һ<+]♐Z$W2OiS^5)tJI?g,M(3#&:g;fϏFd>EꕸQ)his{1?Ktq݆vD|t{7yln3g>ҭ]kKmcޝKզd+}зc-U4׭&~7+ʬ6!ڭ}w _UU7Cc6txi|@cBSilݟkngk>?,WL?2bh1v0&eeUJ8/<1 feIVg䫿j*+bea+cYyfr z+5`%lKMSaeͬUa:Vy -s@jVMcJghJJ@lQX?:VƎPae0aAX]Irr VNJ~~(,iOw5h[JIVY3mVYSYO*\X;3l_<+'ZcVfV+oбK8?YVR2J1+yqF6|*+*+%XyJat?XY'+1V6`%lKu&WXͬ +ܧc%p~-mY0 cV>lTUVf >ʷuʅÄrʧ X9IV±`^y ͬܧrK8?w@jV0LcV> Zeer9q*`r SXYٗ$+*`Y9R2YJ~~c5:Vy %Pm5+L2ۘ8#W{Zee,XyRaf+cWYlS$K X ےaeC+cc%^Z~?`Πy)G&A4H'Pq.$K3EeL~Y_'-wGgFrYd uPŀU^g%)/8H3Dl&A@>yk|JEuSsK!z 3펏SIX׺.XYw3 4tC@dž=%Pm6c 5\Sl>cV5t&7ЕMCc55t0|XyĀ-V.&`V"hދУJ\b ]a:k|#zihWҁYPm5 5\Sy/ZqHC=LCkƎPo:z+0`%lKKgvj跈i"hhmCJ\7Cg4ti( {wDCLC|WtWw΃m54Ϥ?}5<'V54_<4tiL:VJiZ L^*}?u˯5l 3ZN.Z^㹅 1 Y %zy#kI )$hO]Mk -kʣP\B=N(.եmLa-c]Znތ/F^~67']iz} dV]+~U9Hf4g|wAWdPNwU$b[o C;̫~E+:X3rOF4_nΰv}Da6=Zlu87wP۲ ͳ1qI6V%\ L7._SkKU*+%bu8+cY +fȀΰVz[V֌r/Z_ے; fp-3uZe\sXm+a-q@V"+<=| +a]%`%d5+am 'X *+a]%ʮ啽-@jV3*}5e$*+UV#buF8+cYTNE/cxVn:Jk`6րq~s+VtW-mY9$+$\?G'/=*+OZ?X pVƎY9(oi_<+o9JX'X X~]gWuRN g`׊ޖPm5+kL}5\O^*+SYYCZ'X pVƎY9V^!1V530וsa%08+V46limY$+%\C'&hE*+VԱ2v?c=xVΰC o'~{A+8#w-ZkYo˱f(жDL2e Ŀ' rL$D gbl#acbg {?zb=-?21|ĜC@y6F?krd'\30/eVzTV+GY;3a%x%ļX >bY){a`%#/ n ΂mY5J_b13üVYSYG 1ge8 V&\"+a'-V7X >bY){a`%#, ՗Wn ΃mY)d? jH?k~`>b`%-9+cYyfVVo<+a[25Jk`6R:J+?ױ:T]5Է fe$+3j0UVf #V2vbe 0|ımɰ2<|J[~] ^XX >b`%9+Cǃ[CeK@jV0>b3züVYr`>b`%9+caazAXVb|Xg%lK[L#VJ:J+GYolV2$+j0UVfG 'ns(4Reb`>b`"9cKYLls 3%T%|Af>V}>-}ĵ8`8#9^Bd Dlssj9F>vCl.5 2sNsw:ȴnY܏:Þ]m|9lܱэ`;)?zD# z|e8_d#o_kfAe묽g5#G\[j>r3<+\d'st} }ǙŎ'Hd 'cw[ҘOchݶ939r{R/)?u6Xyp^yb6h[&Y9Ƙ8#_Vr@2V>ce8 +,V]z+5`%lKMSaeͬUa:Vy -s@jV5qƬ}Vr@lQX?:VƎPae0a(*|΀MKml@UVWYy@<{:VƎރYŻ|r+a[23y~ʗ`%?mfeXa>+ἄ3o!h[`Jgmj *+ brc:VƎPaaJQ˛V>eI+*̓l`%?mf>c$p^#VlhJg!WZlXCaIce8 V^0LX)wOaGge_\eHd +m3+aK`5:Vy %Pm5+ MȘ8#W{ZeeBXyRaf+cWYF _C)qq .yPX4÷yh( s -(lVkhI p=|ϠgժNW5>h(54vPCQCJ' |Jؖ +kн"hhk>%.>3   PȓNC:;A:ä%̓ ߇B jUCT Aʓ yƎ>E  +c|Jؖ +Sɓpz'gOPȟ 3,3^;?4gp =8,[ VkhI 0gPYZjLUCˉX3@C!q*z#jRb&ų%Tg86܇Z tz?1%g4򪀆B^C8W4p mlhlVkhI 0| 5MjUCgZAʫzR0T4PG5XJؖ +Sɫd\pR Yoޥqq o'*%'8U9GLN]Bs 4މy Y %zy#khO]Mk -kʣ}^J9y]ۦo{G=4r*|~ۘφ1v*oG s`#06NWN\P+PG sA#04NWN傎cAnjSSy91'2DƜȩʩѼ¼Ęc^yT^h^an`|~E }2 {c~^F51G.MhyI#`:yjQ0O-M"'aZyju1W,Mx?6I[Vb Mcu bքc;EбsŘ`:9S"LŜ *DбŘ3`:yK1o)-ż ""؞Ř`:C1w(ܡ  lB1%;4aN~ wb΄c/c 1gLnbL#;#33=^ /x=enܒ⹑9ѪHJj3;^U#x-i8q~I#[Dy55W2^$Hve8xהX #E#Zq2^͎K/"xU66^-zZ[㴶.!^ac תJW+M/=_VT _~A$HE/w`ZzIY R׿#7=/;Ip,a׆wJVbs? y/~E~nދOu_Eg=w]wy;ߪ$t+i9{߄^Wt0M;x4>~bRГ>CysjSʉqnwu3lj*遉9=p-{g9=1.I<7n(51> !dD% kI 0A3ѪuP:(FXqCq׊ Ž-{J@<W5F_zU6L7.RfOe `ᬌd%r0#Vv;JXkX kmXJXk VZJXkkEioK,(жIV YL)#w^KVYWY)V:〬DVx+{YVzJXjVzNsVz]++{[ZAլ de!+>2VPY b%՜〬DV/{Y +!_JX3 VšuZ]ҷ f4TLzSZ2Se"zJX2vAG}OYyV¼t`%KϮ:'X ҁ0/xk`oKR(жY&Ym|] j2KUVf"b͉Vœxq@V"++$YFr0X s-ܞ+a3>sV66thlmV2$+G_W*}&p=7AUVTYCĚw V j9#w-ZkYo˱f(ж4$#s AWd<½A*Ga/gbl#acbg {|%zO_ Y/< // _/(Z5 C֮::v[$|@ڔckтyx 2ļ^Ƌxe32/##2ٸ8TxkhO hgccMq\δ0t5q*+ V2vg+ JJ+?2`%lK-Q+Gi5+e%|ҭYPm5+L2?qnggrM)t|yy8䬌ÄKa<ij%J8N`%x8ftV2T_^5: fY&Y9>qng3)#μVY9^eYG㬌dZAX^Wmɰ$+!_J?+? \P}wP"(ж&Y9!qng'!3UVNPYY wX Uq{3uJƲ%: JZ~] UX U`%xW9+Cǃ[CeK@jVmv>'!3UVNTYy`Y`%f9+caazAXVb|Xg%lKp,+7ku%f`%fll76l mimYYhEs;u%p= 9yHee`]`I+caaʥrr?-V±4y Xa$xv9#w- Z5t M2HpoPILw|.yf1A& 0&~3Dؖ R|.Ln7 pFw33żgV , g\|>\ÍzaWE/^h^eQyPzB*'4WAe/S\_f4HF7R\od4ПD'?Q<b43ϭ3hgVUBoBo~+[ VB^ VB VBo~+[ VB)\ cjzl*i]Hyv ?_֕~-탃KOtG*AtNq8K9gס:Dd 7  0  # A2qqS) `!qqSy1؞Qxpuw;\zP 4E^B(2?lґXTa q1)5h:3Xd?Z Z::TXN c?hEjz}o{{{wI.׎{o{ǷU1}z9??}86nEAWGuX'üӿdِmY;̯%JǍ0E WaTB1|GҞBmAvBo ·pJ[nek*K2Nܖd`$ 7:!-N5pЦۓl':G,UӓM]w$كX*y&o~Qw~ߏs7,ϝ׀!snfZ{p%O)_G21 n T7զM1FnOfBmKE;àYkМ`Ly5ٵpsuAY?vtCww0tsYpڽUzCY,!$sNjOfSϢs~_m,|gY7-0nZN iZgOic/~aWLz8e ұ㦞~9ʟd)VI+L ŸئMLvP'gsS#!y8zz zKOkL}N[6I,W'GXMyՈn&Cq#{nc6h5 wzG^xai6X~;t7 mXow[eJޖ[Ī6x~njc%Wjߦ}YsL\OUqrU\uSGy)mruA 9ɲeaM~ O$c?@$_GRFed yH${SGRF5xS(? j p[qs$I'FRF57q4ydpI A>?'|j&#}.$|oOEӮCI=2C^?|"_UG Z2NYOyRLҏjl\Lϻ396ΞO1m7Ӌ k:/Ae_2^J,ǍLN-L+!c .q1w ;d<25ik!d!Yƣ(Zgx/D 'H>[[es6M!4v~y"…ӝGw?}AwtyMWΊt9gkŸ6N{V)\@]#ւcl,񙠻;㴹%:WT7N)}(s_Ս+S/W<46S[y$)e޵yK!g 'ӀCZS=|O{<鑶-޲!Up} i'Qt | ^-qZA2VO))qߪN})c~k;3[s*-]5Tq_ kOEu~qc;̅vcjg>R)Nq~,BBK3WQ'liMVK'JNـuoIj\i6Nkz< 7yBo)Yߚ1p >_yt>\' -0yk󓽤4iby ZJ700LsO݉E9tDrIs~gA:(JZ ުo3r]Ox*莈y>j;tp? އsV(vj5T/}vWIڏ)Si̾66t>Y^)4!y1P* ɵ"fBp9 cpl0pfN\6o{رVئsJjl%vN)Y;7%KkĠlWwVxNxV yN;z>Ưin9O9!;{BpKo3;k zjU%sHRpAYvNcZ"n MZb;7Ej,=BvEI+jOx5saNۅ"xtq]qmb]1$g+vMFnZ ۅ4+NLۥpٮfiZ#nަ{v楴guʞ!o̞3$4aPFK힩3d_<^)AP#h$vz0.PwrU."cLSTP[F9#xZPP%q61F5;]ܺfU/l~I1_#2Wąⶮo? -~[jqcGIQyNW={c=1`fKr\sVӼ˛[b( R'!`5l5iX1 c&S7=ajXuj\ CukM}i>on]%φtX(b KKo8);a'5+BL`~a5젆r3PZt9\̈́|dG=ӻ[nUW@ᇞu.!|,5cܖנ[ LCx-gJ E/|-1Ӛ͚5Jn eyռ4Ǚ8?*S}7[AYa}ZvSh5;(re4CެioմiY[wktiY5DkMk4(5P5m9(1G5ڹJʖjjXGra]/;Aܹ)/wh7sS^ЖqzGܡ} NyCĝr[s+&Wj-0w&3)+۬8};0x\#;+ܹ^M)/w8Ɏ^vDZ2*QIzd(RVmH8 Iyù_wCGsQ^pĝ r \g;\+ᅤa#koln߳(A$eWMKf7$sP^p <ܵ忸Pn J(w#]pvSpND\QjN/Fj9"W)>|#[,J϶ ުc$#6QДJs0(L٢0-6f̶1)ibJ5rB,u %d@(Rԍ6˄헰t7y6:- >k8ܕ]kmWX&@sm5=DV;]]}');yy,[xIW Z1==r~s^}ċ*phfX^X ) Normal (Web)dd[$\$B*OJQJ^JphHH d References  & FP CJ^JaJ4"4 9A0Header  H$616 9A0 Header CharCJaJ4 B4 9A0Footer  H$6Q6 9A0 Footer CharCJaJHbH va Balloon TextCJOJQJ^JaJFVqF PFollowedHyperlink >*B* phJ`J :;0 Footnote Text h`h CJ^JaJ@/@ :;0Footnote Text Char^J@&`@ :;0Footnote ReferenceH*PK![Content_Types].xmlj0Eжr(΢Iw},-j4 wP-t#bΙ{UTU^hd}㨫)*1P' ^W0)T9<l#$yi};~@(Hu* Dנz/0ǰ $ X3aZ,D0j~3߶b~i>3\`?/[G\!-Rk.sԻ..a濭?PK!֧6 _rels/.relsj0 }Q%v/C/}(h"O = C?hv=Ʌ%[xp{۵_Pѣ<1H0ORBdJE4b$q_6LR7`0̞O,En7Lib/SeеPK!kytheme/theme/themeManager.xml M @}w7c(EbˮCAǠҟ7՛K Y, e.|,H,lxɴIsQ}#Ր ֵ+!,^$j=GW)E+& 8PK!Ptheme/theme/theme1.xmlYOo6w toc'vuر-MniP@I}úama[إ4:lЯGRX^6؊>$ !)O^rC$y@/yH*񄴽)޵߻UDb`}"qۋJחX^)I`nEp)liV[]1M<OP6r=zgbIguSebORD۫qu gZo~ٺlAplxpT0+[}`jzAV2Fi@qv֬5\|ʜ̭NleXdsjcs7f W+Ն7`g ȘJj|h(KD- dXiJ؇(x$( :;˹! I_TS 1?E??ZBΪmU/?~xY'y5g&΋/ɋ>GMGeD3Vq%'#q$8K)fw9:ĵ x}rxwr:\TZaG*y8IjbRc|XŻǿI u3KGnD1NIBs RuK>V.EL+M2#'fi ~V vl{u8zH *:(W☕ ~JTe\O*tHGHY}KNP*ݾ˦TѼ9/#A7qZ$*c?qUnwN%Oi4 =3ڗP 1Pm \\9Mؓ2aD];Yt\[x]}Wr|]g- eW )6-rCSj id DЇAΜIqbJ#x꺃 6k#ASh&ʌt(Q%p%m&]caSl=X\P1Mh9MVdDAaVB[݈fJíP|8 քAV^f Hn- "d>znNJ ة>b&2vKyϼD:,AGm\nziÙ.uχYC6OMf3or$5NHT[XF64T,ќM0E)`#5XY`פ;%1U٥m;R>QD DcpU'&LE/pm%]8firS4d 7y\`JnίI R3U~7+׸#m qBiDi*L69mY&iHE=(K&N!V.KeLDĕ{D vEꦚdeNƟe(MN9ߜR6&3(a/DUz<{ˊYȳV)9Z[4^n5!J?Q3eBoCM m<.vpIYfZY_p[=al-Y}Nc͙ŋ4vfavl'SA8|*u{-ߟ0%M07%<ҍPK! ѐ'theme/theme/_rels/themeManager.xml.relsM 0wooӺ&݈Э5 6?$Q ,.aic21h:qm@RN;d`o7gK(M&$R(.1r'JЊT8V"AȻHu}|$b{P8g/]QAsم(#L[PK-![Content_Types].xmlPK-!֧6 +_rels/.relsPK-!kytheme/theme/themeManager.xmlPK-!Ptheme/theme/theme1.xmlPK-! ѐ' theme/theme/_rels/themeManager.xml.relsPK] fi~m,"~m,jH8H b s s Pl$&, 2+6,;?\BGJLPSrZ<`Nfjq {aZ*AFEWYZ[\^_`acdfghjkmnwxyz|}~$1>BKK7KGKSKnK{KL\k%xȗFXYnƪǪ=?X]beilopqrstuv{a 4HLtBY\$'QeXXXXXXXXXXXX8"#@d"(    S @3c"$`??   C "0`??   C 0e"0`??   s *c"$`?? z " s *"#" `? B S  ?{GN\R sj$ &*< T"! 4$z4TT|)d*% % % tC%  % $%  % d!% <"% X#% EHEsEwEEEEEFFFFCGMGdnCL$$''**++r5}5;;>>CCCDFF!G1G VVttSt[tttPxYxyy||ȇ·V[SVϔԔzz||}}OT?G33333333333333333333333333333333333333333333Gaff]] ==zz||}}ݟݟPFFH ff]] ==zz||}}%?< R+Vt^ ^x +I d' _}jJ ) ]|1F&F tTZFR`v+HW[zsHg! d'G>$ |n&,#(',%=)1 d'kU3u*p == d'V=Znʃ(?x }A6\eaI"|JhMFvN]bKU,V\FlY!jo 5p_YtRBt^u&&QO} d'\$_A hh^h`o(hH) ^`o(hH) ^`o(hH... x^`xo(hH....  ^`o(hH .....  X@ ^ `Xo(hH ......  ^ `o(hH.......  8x^`8o(hH........  `H^``o(hH......... hh^h`o(hH.hhhhhhhh hh^h`o(hH. ^`hH. pLp^p`LhH. @ @ ^@ `hH. ^`hH. L^`LhH. ^`hH. ^`hH. PLP^P`LhH.k^`ko(.k^`ko(..0^`0o(...0^`0o(.... 88^8`o( ..... 88^8`o( ...... `^``o(....... `^``o(........ ^`o(......... hh^h`o(hH4. P^`Po(hH4.. ^`o(hH... x^`xo(hH....  ^`o(hH .....  X@ ^ `Xo(hH ......  ^ `o(hH.......  8x^`8o(hH........  `H^``o(hH......... k^`ko(hH. k^`ko(hH2.. 0^`0o(hH... 0^`0o(hH.... 88^8`o(hH ..... 88^8`o(hH ...... `^``o(hH.......  `^``o(hH........  ^`o(hH......... k^`ko(hH. k^`ko(hH2.. 0^`0o(hH... 0^`0o(hH.... 88^8`o(hH ..... 88^8`o(hH ...... `^``o(hH.......  `^``o(hH........  ^`o(hH......... hh^h`o(hH4. P^`Po(hH.. ^`o(hH... x^`xo(hH....  ^`o(hH .....  X@ ^ `Xo(hH ......  ^ `o(hH.......  8x^`8o(hH........  `H^``o(hH......... hh^h`o(hH4. P^`Po(hH.. ^`o(hH... x^`xo(hH....  ^`o(hH .....  X@ ^ `Xo(hH ......  ^ `o(hH.......  8x^`8o(hH........  `H^``o(hH.........0^`0o(() TT^T`hH. $L$^$`LhH.   ^ `hH.   ^ `hH. L^`LhH. dd^d`hH. 44^4`hH. L^`LhH.h ^`hH.h ^`hH.h pLp^p`LhH.h @ @ ^@ `hH.h ^`hH.h L^`LhH.h ^`hH.h ^`hH.h PLP^P`LhH.h^`OJQJo(hHh^`OJQJ^Jo(hHoh  ^ `OJQJo(hHh\ \ ^\ `OJQJo(hHh,,^,`OJQJ^Jo(hHoh^`OJQJo(hHh^`OJQJo(hHh^`OJQJ^Jo(hHohll^l`OJQJo(hHh^`OJQJo(hHh^`OJQJ^Jo(hHoh  ^ `OJQJo(hHh` ` ^` `OJQJo(hHh00^0`OJQJ^Jo(hHoh^`OJQJo(hHh^`OJQJo(hHh^`OJQJ^Jo(hHohpp^p`OJQJo(hHh ^`hH.h ^`hH.h pLp^p`LhH.h @ @ ^@ `hH.h ^`hH.h L^`LhH.h ^`hH.h ^`hH.h PLP^P`LhH.h^`OJQJo(hHhTT^T`OJQJ^Jo(hHoh$$^$`OJQJo(hHh  ^ `OJQJo(hHh  ^ `OJQJ^Jo(hHoh^`OJQJo(hHhdd^d`OJQJo(hHh44^4`OJQJ^Jo(hHoh^`OJQJo(hHk^`ko(.k^`ko(..0^`0o(...0^`0o(.... 88^8`o( ..... 88^8`o( ...... `^``o(....... `^``o(........ ^`o(.........h hh^h`hH.h P^`PhH..h ^`hH...h x^`xhH.... h  ^`hH ..... h  X@ ^ `XhH ...... h  ^ `hH....... h 8x^`8hH........ h `H^``hH......... hh^h`o(hH4. P^`Po(hH4.. ^`o(hH... x^`xo(hH....  ^`o(hH .....  X@ ^ `Xo(hH ......  ^ `o(hH.......  8x^`8o(hH........  `H^``o(hH......... k^`ko(hH. k^`ko(hH3.. 0^`0o(hH... 0^`0o(hH.... 88^8`o(hH ..... 88^8`o(hH ...... `^``o(hH.......  `^``o(hH........  ^`o(hH.........k^`ko(.k^`ko(..0^`0o(...0^`0o(.... 88^8`o( ..... 88^8`o( ...... `^``o(....... `^``o(........ ^`o(.........^`OJQJo(hH^`OJQJ^Jo(hHopp^p`OJQJo(hH@ @ ^@ `OJQJo(hH^`OJQJ^Jo(hHo^`OJQJo(hH^`OJQJo(hH^`OJQJ^Jo(hHoPP^P`OJQJo(hHk^`ko(.k^`ko(..0^`0o(...0^`0o(.... 88^8`o( ..... 88^8`o( ...... `^``o(....... `^``o(........ ^`o(.........hhh^h`OJQJo(hHh88^8`OJQJ^Jo(hHoh^`OJQJo(hHh  ^ `OJQJo(hHh  ^ `OJQJ^Jo(hHohxx^x`OJQJo(hHhHH^H`OJQJo(hHh^`OJQJ^Jo(hHoh^`OJQJo(hH hh^h`o(hH. ^`hH. pLp^p`LhH. @ @ ^@ `hH. ^`hH. L^`LhH. ^`hH. ^`hH. PLP^P`LhH.hh^h`OJPJQJ^Jo(-88^8`OJQJ^Jo(hHo^`OJQJo(hH  ^ `OJQJo(hH  ^ `OJQJ^Jo(hHoxx^x`OJQJo(hHHH^H`OJQJo(hH^`OJQJ^Jo(hHo^`OJQJo(hH hh^h`o(hH4. ^`o(hH) ^`o(hH... x^`xo(hH....  ^`o(hH .....  X@ ^ `Xo(hH ......  ^ `o(hH.......  8x^`8o(hH........  `H^``o(hH.........h hh^h`o(hH.h P^`Po(hH2..h ^`o(hH...h x^`xo(hH.... h  ^`o(hH ..... h  X@ ^ `Xo(hH ...... h  ^ `o(hH....... h 8x^`8o(hH........ h `H^``o(hH......... k^`ko(hH. k^`ko(hH2.. 0^`0o(hH... 0^`0o(hH.... 88^8`o(hH ..... 88^8`o(hH ...... `^``o(hH.......  `^``o(hH........  ^`o(hH......... hh^h`o(hH4. P^`Po(hH4.. ^`o(hH... x^`xo(hH....  ^`o(hH .....  X@ ^ `Xo(hH ......  ^ `o(hH.......  8x^`8o(hH........  `H^``o(hH.........^`o(. ^`hH. pLp^p`LhH. @ @ ^@ `hH. ^`hH. L^`LhH. ^`hH. ^`hH. PLP^P`LhH.hh^h`CJOJQJ^JaJo([] hh^h`o(hH4. P^`Po(hH4.. ^`o(hH... x^`xo(hH....  ^`o(hH .....  X@ ^ `Xo(hH ......  ^ `o(hH.......  8x^`8o(hH........  `H^``o(hH.........hhh^h`OJQJo(hHh88^8`OJQJ^Jo(hHoh^`OJQJo(hHh  ^ `OJQJo(hHh  ^ `OJQJ^Jo(hHohxx^x`OJQJo(hHhHH^H`OJQJo(hHh^`OJQJ^Jo(hHoh^`OJQJo(hH hh^h`o(hH.hhhhhhhhh^`OJQJo(hHh^`OJQJ^Jo(hHoh  ^ `OJQJo(hHh` ` ^` `OJQJo(hHh00^0`OJQJ^Jo(hHoh^`OJQJo(hHh^`OJQJo(hHh^`OJQJ^Jo(hHohpp^p`OJQJo(hHk^`ko(.k^`ko(..0^`0o(...0^`0o(.... 88^8`o( ..... 88^8`o( ...... `^``o(....... `^``o(........ ^`o(.........^`CJOJQJo(^`CJOJQJo(opp^p`CJOJQJo(@ @ ^@ `CJOJQJo(^`CJOJQJo(^`CJOJQJo(^`CJOJQJo(^`CJOJQJo(PP^P`CJOJQJo(% JhMs!=)1tG>$FR ==+I#(') \$_+Vt ^vNQO}}jJ (?|n&|1& 5p?< bKU\eaI!jo`}AYtukU3HWV=V\%%0 P        f                                                              ӆ                 0.7                 )f$XK栳'jTeJ;*bdla         a!Tq%tt/'0@XK]Lm[',{ 2=J=7ai!W#h%1{5F(- v } )'`E N#1sum9Aw >YO{MDhL4:2t|tK &!s!1%8& t'|'>h() )k,E. J.oP.@h2B23Ic3q4'4_5TJ6]7`7890:: ::;J;<=HW=\+>1>u>w?eAB^B1D HVITnIT[JKjKX1N6N}LOMQTVUB\UwGV/ YY|zY2ZmZEh]'^^^"_K_eVaQ\aMzaDb wbf|gngD$hth>Ci|CjEfjij?kP_ktk'lmL;n|nJ pApqpgs va=w y_|*,~:q"{E \Z ?{YN8z}v^\ +;RL&h^ <EVwdahc:! NxrS:q}:r? |+?JMpE7D$Xe'l?OZW^ nMra5lZ3AJLA_UV0s<N(!] ^h-Qbgm6Xw({E W$ mh)*bMQvaYcRn j-s-B{j/\<UO^Aq O'u?f-5dc S~Dsdhsq"P[q7Qv,CU4bex$J%IDs||F 0i@GGGGp@UnknownUnknownCatharine van IngenUnknown20080908T125830953:C& G* Times New Roman5Symbol3. * Arial3* Times7K@Cambriaa Angsana NewArial Unicode MS5. *aTahoma?= * Courier New;WingdingsA BCambria Math"1hkŠfk 2U Q U Q 4džž 2qHX ?pE2! xxAuthor Guidelines for 8HICSS_PCCatharine van Ingen%                           ! " # $ Oh+'0 $0 P \ h tAuthor Guidelines for 8 HICSS_PCNormalCatharine van Ingen11Microsoft Office Word@#@lv@::@3{ U՜.+,D՜.+,\ hp  Shidler College of Business Qž Author Guidelines for 8 TitleL 8@ _PID_HLINKSAH}!mailto:rwilton@pha.jhu.edu) W*&mailto:vaningen@windows.microsoft.com) Atmailto:thakar@jhu.edu) smailto:nieto@pha.jhu.edu) T<mailto:tony.hey@microsoft.com) =Jmailto:heasley@ifa.hawaii.edu) kmailto:dan.fay@microsoft.com) L# mailto:randal@cs.jhu.edu) s mailto:alainna@pha.jhu.edu) ^qmailto:jvv@jhu.edu) Kxmailto:gbell@microsoft.com) _smailto:szalay@jhu.edu)   !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~      !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~      !"#$%&'()*+,-/0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~Root Entry F˙Data 1Table.զWordDocumentHSummaryInformation(DocumentSummaryInformation8CompObjy  F'Microsoft Office Word 97-2003 Document MSWordDocWord.Document.89q