From ee6a814a40445147239652ec76165de7375870c2 Mon Sep 17 00:00:00 2001
From: jeff-colucci <116673918+jeff-colucci@users.noreply.github.com>
Date: Wed, 25 Oct 2023 14:47:02 -0500
Subject: [PATCH 01/19] feat(NRQL): Add new NRQL tutorials
---
.../introduction-nrql-tutorial.mdx | 4 +-
.../nrql-tutorial-advanced-dashboarding.mdx | 293 +++++++++++++++++
.../nrql-tutorial-control-your-data.mdx | 294 ++++++++++++++++++
...ries-nrql_screenshot-crop-tutorial2-1.webp | Bin 0 -> 54372 bytes
src/nav/dashboards.yml | 4 +
5 files changed, 593 insertions(+), 2 deletions(-)
create mode 100644 src/content/docs/query-your-data/nrql-new-relic-query-language/get-started/nrql-tutorial-advanced-dashboarding.mdx
create mode 100644 src/content/docs/query-your-data/nrql-new-relic-query-language/get-started/nrql-tutorial-control-your-data.mdx
create mode 100644 src/images/queries-nrql_screenshot-crop-tutorial2-1.webp
diff --git a/src/content/docs/query-your-data/nrql-new-relic-query-language/get-started/introduction-nrql-tutorial.mdx b/src/content/docs/query-your-data/nrql-new-relic-query-language/get-started/introduction-nrql-tutorial.mdx
index ada29e68c2a..9e54e252386 100644
--- a/src/content/docs/query-your-data/nrql-new-relic-query-language/get-started/introduction-nrql-tutorial.mdx
+++ b/src/content/docs/query-your-data/nrql-new-relic-query-language/get-started/introduction-nrql-tutorial.mdx
@@ -425,6 +425,6 @@ Grouping or "faceting" allows you to get dimensional with your data. When applyi
Congratulations: you've completed the first NRQL tutorial! If you've made it this far, you now know the basics of querying event data using NRQL.
-Have you read the tutorial but don't have an account yet? If you haven't already, create your free New Relic account below to start monitoring your data today.
+### What's next?
-
+Content forthcoming
diff --git a/src/content/docs/query-your-data/nrql-new-relic-query-language/get-started/nrql-tutorial-advanced-dashboarding.mdx b/src/content/docs/query-your-data/nrql-new-relic-query-language/get-started/nrql-tutorial-advanced-dashboarding.mdx
new file mode 100644
index 00000000000..f679e45affa
--- /dev/null
+++ b/src/content/docs/query-your-data/nrql-new-relic-query-language/get-started/nrql-tutorial-advanced-dashboarding.mdx
@@ -0,0 +1,293 @@
+---
+title: "Advance your dashboarding with NRQL"
+tags:
+ - NRQL tutorial
+ - 'NRQL: New Relic query language'
+ - Get started
+ - Tutorial
+metaDescription: "Learn how to use NRQL to improve your dashboarding"
+---
+
+import queriesnrql2tutorial1 from 'queries-nrql_screenshot-crop-tutorial2-1.webp'
+
+Welcome! We hope seeing example queries and explanations in the context of your own data helps you better understand how to transform data into powerful visuals. (Try queries from this course on different event types in your own account(s) to really understand their power!)
+
+We are now moving on to even more advanced concepts. You may not use these functions and features on every single dashboard, but they will certainly come in handy when tackling specific problems and requirements.
+
+In this level, we will cover faceting by case, advanced aggregation functions, the value of the EXTRAPOLATE keyword, filtering aggregation functions, and how to override values. Remember if you've chosen an account without APM data. You will see backup queries that may not match the lesson description.
+
+Specifically, you'll learn how to use:
+
+* Advanced aggregation functions like `filter()`, `apdex()`, `rate()funnel()`, `histogram()`.
+* The `EXTRAPOLATE` clauses.
+* `FACET CASES()`, including how to use attribute and group matching values.
+* `filter()` to combine Event Types.
+* Overriding values, when necessary.
+
+### Use advanced aggregators
+
+
+*Calculating Rate*
+Let's start with the rate() function. It allows you to visualize the frequency of events over time. This is helpful when you want to understand the frequency of events in small periods of time within larger time windows.
+
+In the example below, we display the average frequency of requests per 5 minutes for the last hour. We compare this to the previous hour’s 5-minute average frequency. Notice the query uses SINCE 1 hour ago, this is the overall time window in which we are calculating the rate.
+
+You can use rate() to calculate requests per minute or requests per second by setting the time interval to either 1 minute or 1 second.
+
+
+
+```sql
+SELECT rate(count(*), 5 minutes)
+FROM Transaction
+SINCE 1 hour ago
+COMPARE WITH 1 hour ago
+```
+
+
+
+
+
+
+#### Funnel Charts
+Many New Relic customers use funnel charts to understand end-user behavior. Funnel charts track the occurrence of an attribute value across multiple records. They are commonly used to visualize how successfully users' progress through defined paths, and are especially powerful when using custom attributes.
+
+Here we use the funnel() aggregator function to visualize how many users visit the home page, then proceed to another page. The first parameter is the identifying attribute for the unique entries we're counting. In this case, New Relic Browser assigns and retains a session ID attribute for each user on your site (subject to cookies being enabled). You can also set your own session ID using custom attributes.
+
+The remaining parameters determine how each step of the funnel is calculated. They are written in the format ", WHERE attr OP value". In this case, we provide two: we want to know how many user sessions visited the homepage, then also how many of these also navigated to other pages. Try this query on your own data to get a reasonable result.
+
+```sql
+SELECT funnel(session, WHERE pageUrl LIKE 'http%//%.%.com/' AS 'Home', WHERE pageUrl NOT LIKE 'http%//%.%.com/' AS 'Any Other Page')
+FROM PageView
+SINCE 1 week ago
+UNTIL now
+```
+
+
+
+Aggregator Filters
+filter() is a powerful tool that allows you to aggregate multiple data points in a single query, offering more control over which events are included in function results. In this example, we use filter() to return the separate values for total transactions, total web transactions, and total non-web transactions:
+
+
+
+
+ Example
+
+
+ Copy this query
+
+
+
+
+
+
+
+
+```sql
+SELECT count(*) as 'All Transactions', filter(count(*), WHERE transactionType = 'Web') as 'Web Transactions', filter(count(*), WHERE transactionType !='Web') as 'Non-Web Transactions'
+FROM Transaction
+SINCE 24 hours ago
+```
+
+
+
+
+Since it returns a number, you can also perform math on its results. For example, we can divide total web transactions by all transactions to determine what percent of transaction were web transactions:
+
+QUERY
+IMAGE
+
+
+Histograms
+Histograms allow you to better visualize the distribution of your data. This assists in understanding how data points are grouped by frequency, not just averages. The histogram() function takes three arguments:
+the attribute you want to plot (such as duration)
+the maximum value of the range you want to consider (such as "1" for 1 second or less)
+the number of buckets you want data grouped in
+In our example, we create a histogram() chart for all duration values between 0 and 1 second, grouped into 50ms buckets. We do this by specifying "20" for the number of buckets. All the durations larger than 1 second are grouped together in the last bucket.
+
+QUERY
+IMAGE
+
+
+Apdex
+The apdex() function calculates an Apdex score on any numerical value (such as duration). You can calculate Apdex for one or more specific transactions, account for custom attribute values, and even provide your own Apdex-T value without interfering with application settings. In this example, we have provided the function with an attribute of "duration" and an Apdex-T value of 0.08:
+
+QUERY
+IMAGE
+
+
+We can also add the TIMESERIES operator to chart the data over time. Notice this also plots the Apdex satisifed, tolerated, and frustrated thresholds.
+
+QUERY
+IMAGE
+
+
+
+We just explored a whole new set of visualizations with funnel() and histogram(). We also learned how filter() can help us get more specific in queries with WHERE clauses, and how rate() can display the rate of an attribute over time.
+
+These queries further advance your NRQL ability. Apdex is an industry standard and is applicable in many scenarios. Funnels can track progress through desired paths. Histograms visualize the clear distribution of the data; and filters let you get super specific with your returned values. Next, we will learn about EXTRAPOLATE.
+
+### Use extrapolate
+
+The New Relic Database (NRDB) receives and processes enormous amounts of data, every day, at lightning speed! When APM records a large amount of event data, New Relic agents implement a sampling technique to continue collecting meaningful data while reducing potential impact to your applications. This usually only happens when a single event in an application or service handles extremely high volumes of requests. If multiple agents are spread across multiple load-balanced instances of a service, this limit might never be observed.
+
+Let's discuss what we can do when this happens. The EXTRAPOLATE operator tells New Relic to mathematically compensate for the effects of sampling, thereby returning results that more closely represent activity in your system. We store an extra value to record how many similar events occured over the limit. This allows New Relic to deliver statistically accurate results. There is no fallback query for this if you have no APM data due to the fact this is specific to Transaction events.
+
+QUERY
+IMAGE
+
+You might be thinking "are we hitting the limit?" Well, try removing EXTRAPOLATE from the query, and see if your count changes. If it doesn't, you most likely haven't reached the limit.
+
+When EXTRAPOLATE is included in a query, the ratio between the reported events and the total events is calculated. This ratio is then used to extrapolate an approximation of unsampled data. Keep in mind that only some queries support its use. When included in a NRQL query that doesn’t support it, or that doesn't use sampled data, it will have no effect.
+
+Note that EXTRAPOLATE is most useful for homogenous data (like throughput or error rate). It's not effective when attempting to extrapolate a count of distinct things (like uniqueCount() or uniques()). This clause works only with NRQL queries that use one of the following aggregator functions:
+
+* apdex
+* average
+* count
+* histogram
+* sum
+* percentage
+* rate
+* stddev
+
+### Use facet cases
+
+
+
+As you learned in previous lessons, FACET is a great way to both segment your data, and understand it from differently grouped perspectives (such as seeing average response time based on different response codes). When you use FACET, NRDB organizes data into groups based on the values of provided attributes. But what if you wanted to group multiple values together, such as HTTP response codes 200, 201, etc.?
+
+FACET CASES() solves for this issue by allowing you to choose how facet buckets are broken out. The operator takes any number of parameters in the format "WHERE attr OP value". In the example below we've categorized all transactions with response code starting with "2" into a "2xx Responses" bucket. We could also do this for 3xx, 4xx, and 5xx response codes to group our data in ways that increase readability and help us understand what's happening in our application(s).
+
+QUERY
+IMAGE
+
+
+As you can see, these groupings are useful but difficult to read. Let's clean them up using something we learned in level 2 of this course:
+
+QUERY
+IMAGE
+
+
+
+FACET CASES() allows us to match and group attributes with mulitple, interesting values we want combined. There are many useful scenarios for this functionality. It's also even more powerful when custom data is tagged onto your transaction data, allowing you more granularity and control in navigating and grouping data.
+
+### Filter by event type
+
+
+
+We're truly becoming NRQL wizards! Next we'll explore something few New Relic customers are even aware of: filtering to event types.
+
+So far, we've made queries that pull data from a single source. But what if you want to plot 2 data points that are stored as two different event types? Querying NRDB data is not limited to a single event type! To query from multiple event types you can include each event type seperated by a comma.
+
+QUERY
+IMAGE
+
+
+To make this even more useful, the eventType() function tells us which event type a record is from. We can use this to control our data output. In this example, we see the total number of Transaction and PageView events combined, as well as the totals for only Transaction and PageView.
+
+QUERY
+IMAGE
+
+
+Let's look at this in more detail: count(*) is the total number of both Transaction and PageView events. However, we can use the aggregator function filter() we recently learned about to do something unique. We tell it WHERE eventType()='PageView'. This invokes the filter function to observe the event type as part of the total result set, then filter to display only those specific events. We can even add TIMESERIES to visualize 2 directly comparable data points on a line graph.
+
+QUERY
+IMAGE
+
+
+
+We have managed to locate, return, and graph data from two different event types. This is an example of how NRQL can allow you to navigate any necessary data quickly and succinctly; no complex joining or join statements required!
+
+### Override values
+
+
+
+Counting NULL Values
+Sometimes data simply doesn't report in the format you need. For instance, sometimes integers are returned as strings, but you need them as integers to perform maths. Or maybe you get a NULL result, but in your case NULL actually means 0. Don't worry! We hear you, and we've added functionality to address this.
+
+NULL values on attributes can appear on both out-of-the-box and custom data. When you use aggregators such as count() and average(), NRQL automatically removes NULL values from the calculation, only performing the function on events without NULL values. NRQL lets you account for unexpected NULL values in calculations by using the "OR value" clause. For example, if you wanted to make sure NULL values for your "cartValue" attribute are counted as 0, you could use "cartValue OR 0" in your query.
+
+In this example, running count() on ApdexPerfZone only counts the number of times ApdexPerfZone has a value. But if we add "OR 'Null'" to the argument, we can count all transactions where ApdexPerfZone exists, and also those where the value is null.
+
+QUERY
+IMAGE
+
+
+Coercion
+NRQL does not automatically apply coercion. This means a float stored as a string is treated as a string, and cannot be used by mathematical functions like sum() or average(). To override this behavior, use boolean() or numeric() to convert arguments to a boolean or numerical values. In this example, an average() function on "httpResponseCode" provides no value since this attribute is a string. But if we convert the attribute to a number using numeric(httpResponseCode), we can use the average() function successfully.
+
+QUERY
+IMAGE
+
+
+Another common example is BOOLEAN (TRUE or FALSE) values. These are often incorrectly formatted as strings. When this happens, don't worry! You can change how the source sends the data to make it a proper boolean. Or, you can use the boolean() function. The example query below returns the same result, but that is because we're using a value sent by the agent as a BOOLEAN. If your attribute was a string "TRUE", boolean() would convert it into a proper boolean format, allowing the query to run as intended.
+
+QUERY
+IMAGE
+
+
+You can also convert boolean and numeric values to strings by using the string() function. Where numeric values are floating-point numbers you can use the optional precision argument to limit the number of decimal places for the string. This query returns the duration value as a string limited to three decimal places.
+
+QUERY
+IMAGE
+
+
+
+Sometimes the devil is in the details. Here we've given you the power to control your data formats, and tell NRQL how you want it to act. NRQL operates in the manner we deem most logical, but if that does not suit your unique scenario, you can use the functions explored in this lesson to override values.
+
+### Use string concatenation
+
+
+
+There may be some cases where you need to append and/or prepend text to the returned value of an attribute. This can be achieved using the concat() function.
+
+You can provide upto 20 arguments for the concat() function to concatenate into a string.
+
+QUERY
+IMAGE
+
+
+We can limit the number of decimal places that are used for any floating point numbers in the values of the concatenated attributes. To do this we use the optional precision: argument as the last value. In this example we are appending 's' to denote seconds, and limiting the value to 3 decimal places.
+
+QUERY
+IMAGE
+
+
+Values that start with 'http(s)' are automatically displayed as links which can be clicked to open a new page, which means it is possible to create integrations to solutions where a dynamic URL can be used to open a related page to the entity. The following example demonstrates an example URL where the query parameter values are set by the attribute values.
+
+QUERY
+IMAGE
+
+
+
+You can use the concat() function to combine values together (e.g. city and country for location), and prepend/append additional strings to present the data as you need.
+
+
+
+In this section we explored specific, powerful NRQL functionality. These skills will undoubtedly serve you next time you're in the trenches, diving into the nitty-gritty of your data.
+
+You're now a Level 3 NRQL expert! Believe it or not, there are even more features and functions we want to showcase in the next section.
+
+
+
+### What's next?
+
+Content forthcoming
\ No newline at end of file
diff --git a/src/content/docs/query-your-data/nrql-new-relic-query-language/get-started/nrql-tutorial-control-your-data.mdx b/src/content/docs/query-your-data/nrql-new-relic-query-language/get-started/nrql-tutorial-control-your-data.mdx
new file mode 100644
index 00000000000..7ca9b392154
--- /dev/null
+++ b/src/content/docs/query-your-data/nrql-new-relic-query-language/get-started/nrql-tutorial-control-your-data.mdx
@@ -0,0 +1,294 @@
+---
+title: "Control your data with NRQL"
+tags:
+ - NRQL tutorial
+ - 'NRQL: New Relic query language'
+ - Get started
+ - Tutorial
+metaDescription: "Learn how to use NRQL to control your data."
+---
+
+Welcome to Level 2. In the previous level, we explored the fundamentals of building queries and manipulating results to retrieve the data we want. We covered basic query structure, defining time windows, and how to select specific attributes to observe. We also began learning how to aggregate and display data on line graphs and other visualizations.
+
+In this section, we will delve deeper and explore even more functions for creating interesting aggregations. We will learn how to compare returned data with previous time periods, group data into more specific time windows, be more granular with `SINCE` and `UNTIL` functionality, and explore using wildcards in filters. We will even discuss how to rename attributes to be more user-friendly.
+
+We had a gentle introduction to aggregation in Level 1 by using count(), average(), sum(), max() and min() to transform data in meaningful ways. Next we will explore even more powerful functionality. In this lesson we will learn how to find and count unique values, locate the most recent or oldest entries for an attribute, and work with percentages and percentiles.
+
+You'll learn how to:
+
+* Query for unique values using uniques(attributeName)
+* Determine how many unique values exist in an attribute using uniqueCount(attributeName)
+* Retrieve the earliest(attributeName) and latest(attributeName) within a specific time window
+* Calculate percentages based on a qualifer or other data point with percentile()
+* Perform basic math using attributes and aggregation functions, or a combination of both
+* Cast attribute names to something custom and more readable
+* Search to include/exclude using wildcards with LIKE/NOT LIKE, or limit results to those IN a list or NOT IN that list
+* Query within more advanced time windows using dates, epoch, and WITH TIMEZONE
+* Group data into interesting time windows using time-based cohorting
+
+Again if you have chosen an account without Transaction event reporting, we will fallback to other queries that are similar. Let's get started!
+
+### Aggregate functions
+
+
+In a previous lesson we learned how the count() function can return a simple count of available records. To determine the number of unique values recorded for an attribute over a specified time range, we can use the uniqueCount() function. In this function, we provide the attribute we want to inspect as an argument. For instance, here we display all the unique hosts:
+
+IMAGE
+
+QUERY
+
+
+To optimize query performance, the function above returns approximate results for queries that inspect more than 256 unique values. To return the actual list of unique values for an attribute over a specified time range, we can use the uniques() function.
+
+IMAGE
+
+QUERY
+
+A second limit parameter may be provided: uniques(attribute[,limit]). When it is not provided, the default limit of 1,000 unique attribute values per facet is applied. You may specify a different limit value, up to a maximum of 10,000.
+
+
+To retrieve the most recent value of an attribute over a specified time range, use the latest() function. In this sample query, we are locating the most recent response time for a web transaction in the last day. This could help us locate the latest value for an intermittently reporting transaction or service.
+
+IMAGE
+
+QUERY
+
+
+Using the earliest() function will do the opposite; that is, it will return the earliest value of an attribute recorded in the specified time range. In this sample query, we retrieve the earliest response time for a web transaction in the last day. If data is consistently reporting, this will simply be the data point from the earliest event 24 hours ago.
+
+QUERY
+
+IMAGE
+
+
+There may also be scenarios in which you need percentages instead of counts, sums, or averages. Using the percentage() function allows you to calculate the percentage of a value in the data set that matches a specified condition. This function takes two arguments: the first is an aggregator function for your desired attribute, such as count(). The second is a WHERE condition to specify the subset of data you'd like to query.
+
+In this sample query, we are finding the percentage of transactions over the last day that had a duration (or response time) greater than 100 milliseconds.
+
+QUERY
+IMAGE
+
+
+It's very common to view application performance and/or customer experience data using percentiles rather than averages. With the percentile() function we can understand the experience of the nth percentile.
+
+For example, let's say we want to know what the worst experience (highest duration) of 98% of our customers is today. We can ask NRDB for percentile(duration, 98) from the last 24 hours.
+
+
+
+As you can see, aggregation can manipulate data in powerful ways. We used uniqueCount() to count the unique entries of a particular attribute. But you could also use this to identify a count of unique machines, reporting containers, or even how many custom data points are sent to New Relic. And, if we want to know what unique values are available to query, we can always ask for a list using uniques().
+
+The latest()/earliest() functions are particularly useful when dealing with sparse data, or investigating when something began or stopped reporting (assuming the data is still stored in New Relic).
+
+We also learned that percentage() can determine what percentage of events matched a qualifier compared with the overall result set.
+
+Finally, we saw how to observe percentiles. For example, you can use percentile() as a Key Performance Indictator by setting a goal to ensure 90% of all end user transactions are faster than a provided duration.
+
+We are getting much more advanced at dancing with our data! In the next lesson we will learn how to do basic mathematics with NRQL.
+
+### Use math operators
+
+
+
+NRQL supports basic math operators. You can perform addition (+), subtraction (-), multiplication (*), and division (/) on both numeric attribute values, and results of aggregator functions.
+
+For example, when a transaction event is recorded we capture both the total response time (as duration) and database response time (as databaseDuration). But what if we want to compute the average time spent outside of database processing? Well, let's start by calcuating that value for each event in our data set.
+
+QUERY
+IMAGE
+
+
+
+Great! We just performed some basic math. This is useful if we are listing individual events. But what if we want to know the average duration of transactions without the database time? Conveniently, we can do the arithmetic within the function:
+
+QUERY
+IMAGE
+
+
+Now, what if we wanted to get even more complicated and subtract, divide, and multiply in the same query to figure out the duration without database time, as a percentage of overall time? Well, we can add the math:
+
+QUERY
+IMAGE
+
+
+
+New Relic reports timings as part of your events, so you can use them to calculate interesting data points, or even generate percentage results.
+
+You can further maximize the power of basic mathematics by doing things like timing custom actions or events, or sending custom data. For instance, an e-commerce platform that reports data about order sizes, which transactions were successful orders, and payment methods could calculate values such as the conversion rate of orders vs. unique customer visits.
+
+### Label attributes
+
+
+
+As you start performing more complex NRQL functions, you will notice that the name displayed for queried attributes might not be helpful - especially for others that don't know NRQL! Let's consider an example using what we learned in basic math.
+
+QUERY
+IMAGE
+
+
+
+We can use the AS clause after a function or attribute to give the result a more readable, meaningful name. This makes it easier for you and your team(s) to understand exactly what a chart represents.
+
+QUERY
+IMAGE
+
+This may seem purely aesthetic, but when you're building detailed dashboards, it's important to clearly label your data. This ensures zero ambiguity for anyone viewing your widgets, billboards, line charts or tables.
+
+We'll refer back to this in upcoming examples about grouping, to explore how AS can create clean result sets in more advanced scenarios too.
+
+
+### Compare time windows
+
+
+
+We have learned how to use time ranges with SINCE and UNTIL clauses. But what if we want to compare values fromdifferent time ranges? We can achieve this with the COMPARE WITH clause.
+
+We use SINCE and UNTIL to define our period of interest. Then, we denote the time period we'd like to compare against with a COMPARE WITH [time period] AGO clause containing a relative offset value.
+
+Specifically, in the sample query below we compare the last day against the previous week using a relative offset of 1 week ago.
+
+QUERY
+IMAGE
+
+
+To map the comparison of values over time, add TIMESERIES. This creates a line chart of the comparison, allowing you to visualize how this period compares to recent data, and track it over time.
+
+QUERY
+IMAGE
+
+
+You can also specify many different relative time periods in the same format, similar to UNTIL. For instance, you may specify 4 WEEKS AGO or 6 HOURS AGO.
+
+QUERY
+IMAGE
+
+Comparisons can quickly answer questions about what's happening in your applications. Are different sales, performance, MTTR, or error values up or down compared to last week? And, if you are investigating an issue, comparing aperiod of problemantic performance to a period of normal performance can be very useful.
+
+
+### Use wildcard filters
+
+
+
+We now know how to use a WHERE clause to filter the results in our query. Aside from using standard comparison operators, we can also use LIKE and NOT LIKE if we want to determine whether an attribute contains or does not contain a specified substring. In order to achieve this, we need to use the percent (%) symbol as a wildcard anywhere in the string.
+
+In our sample query, we are getting the number of transactions with the term "Web" anywhere (beginning, middle or end) in the name.
+
+QUERY
+IMAGE
+
+
+If we change our query to use NOT LIKE instead, then we will get the number of transactions that do not contain the word "Web" in its name.
+
+QUERY
+IMAGE
+
+
+We used the wild card % at the beginning and end, which means that we are checking the value of the attribute we chose if it contains "Web" anywhere in the text. Equally, you could use `%Web` OR `Web%` to match something that ends in "Web" or starts with "Web", respectively.
+
+You can also add the wildcard in between strings for a more refined search. This query will check for a transaction name that contains the word "Web" followed by any text, but then also contains the word "index" followed by any number of characters. So, the results will only be transactions with "Web" and "index" in the name.
+
+QUERY
+IMAGE
+
+
+What if you need to be extremely specific and the names don't have a common string you can match using wildcards. The IN and NOT IN operators allow us to specify a set of values that we would like to check against an attribute. Instead of specifying multiple WHERE clauses with AND or OR operators, we can simplify our condition by listing the values in parentheses separated by commas.
+
+In this sample query, we are counting the number of transactions whose subtype is either "SpringController" or "StrutsAction". If you change our query to use NOT IN instead, then we will get the number of transactions whose subtype is neither "SpringController" nor "StrutsAction".
+
+QUERY
+IMAGE
+
+
+
+You can now control your data and manipulate it to do what you need, allowing you to construct powerful, meaningful dashboards and alerts.
+
+### Specify time ranges
+
+
+
+The SINCE and UNTIL clauses are not limited to relative time ranges. You can also provide a specific date and/or time. In the following sample query, we include a SINCE date in YYYY-MM-DD format. This is useful when creating SLA reports for a specified period of time.
+
+QUERY
+IMAGE
+
+
+You can even include specific time with the format YYYY-MM-DD HH:MM. In this query, you can see the data is set at 6pm.
+
+QUERY
+IMAGE
+
+
+Sometimes, as an engineer, you might receive an event time in epoch (unix) time. Conveniently, epoch timestamps are an acceptable value in SINCE and UNTIL clauses, so you don't have to translate these values into another date format.
+
+QUERY
+IMAGE
+
+
+When NRDB shows data over a period of time, it assumes you want to see the data from the perspective of your timezone. But with dispersed international teams, your "today" could be someone else's "tomorrow" or "yesterday" depending on their location.
+
+You can use the WITH TIMEZONE clause to define a timezone to display data from. This affects the interpretation of values in SINCE and UNTIL clauses.
+
+Consider the two example charts below. Each query has a specified timezone using WITH TIMEZONE. The two are 8 hours apart. Notice the pattern of data is the same, but offset 8 hours to align with each respective timezone:
+
+QUERY
+IMAGE
+
+QUERY
+IMAGE
+
+
+
+Before this lesson, all our time control mechanisms were based on relative times from now. With the lessons today, we have learned how to adjust the view depending where someone is in the world. Maybe a customer in East Coast America reports an issue and your engineering team is in West Coast America. They can build a dashboard and translate the view to map to the timezone as a customer would be citing. So if a customer advises an issue at 9am East Coast, we can ensure when we look at 9am we don't have to mentally translate.
+
+Epoch and standard date stamps make sense. So when you need to focus your data to specific dates of an incident for example, and you want to investigate the data without a moving time window being relative to the current time, this knowledge will help you obtain data in a static time window.
+
+### Using time-based cohorting
+
+
+
+Time-based cohorting...what a complicated but fun term! While it sounds complex, "time-based cohorting" is simply a way to organize data into time-based groups like minuteOf, hourOf, weekOf, and more.
+
+When we use the SINCE clause for durations, we retrieve the entire length of time queried for. But that data may not always tell the whole story! We may need to ananlyze performance within a time period more closely. With time-based cohorting, we can further sort the data into logical, time-based groupings.
+
+Using a combination of FACET and one of the many time-based functions (such as hourOf(timestamp)) we can take a week’s worth of data and understand performance based on the specific hour it occurred. This reveals trends and identifies the most critical times for our application. Check it out:
+
+QUERY
+IMAGE
+
+
+After running the query above we can see the slowest response time based on the hour of the day. Super interesting, right?
+
+New Relic provides many different options to facet based on time. The previous example is grouped by the hour, but we can also group by the day of the week to determine which days have the best and worst response times.
+
+QUERY
+IMAGE
+
+
+Now can see when are we slowest on any specific day. This could be logically extended to answer business-critical questions like "When do we sell the most products?", or "When do we have the most sign-ups or logins?".
+
+You can also group results by a specific date. This helps when considering SLA reports, or analyzing performance changes over a given period.
+
+QUERY
+IMAGE
+
+
+
+Digging into your data is always fun! Time-based cohorting is an easy way to expose problems that occur on specific minutes, hours, days, or weeks. No matter what data you send to New Relic, NRQL allows you to slice, dice, organize, and visualize it however you want.
+
+There are also many other options available to group by, including week, month, and year depending on your data retention. To see the full list, head to our [Group Results Across Time documentation Page](/docs/query-data/nrql-new-relic-query-language/nrql-query-examples/group-results-across-time).
+
+
+
+With the knowledge you've gained here, you can create dashboard visualizations, and control the aspects of your data you're most interested in. These techniques are also extremely useful in narrowing focus for more granular, specific alerts!
+
+You are now a Level 2 NRQL hero. You've learned great techniques to control your data and produce much more interesting visualizations. Get ready to advance to level three, where you'll learn more interesting NRQL skills including filters, facet cases, histogram, apdex, filtering to eventTypes, overriding values, and extrapolation.
+
+Revel in your newly attained knowledge. You are now equipped to build really powerful dashboards and alerts that could be critical to your organization.
+
+
+
+### What's next?
+
+Content forthcoming
\ No newline at end of file
diff --git a/src/images/queries-nrql_screenshot-crop-tutorial2-1.webp b/src/images/queries-nrql_screenshot-crop-tutorial2-1.webp
new file mode 100644
index 0000000000000000000000000000000000000000..4df372a4279c1bb302fa3307d174aed1c4617615
GIT binary patch
literal 54372
zcmeFXQyOF%&CqCyI43LHcVf3%UPK(c|U%b=csfq;O>i#u=c6|lDQ
zFKfPJzG~muF9}0vt+wQ60rsD
zKQjTb?*IV*4d_>H*&haVkXO&$zM_Cw9}=IzkK}zzi3<
z(SfwH(9-YQHU;0{HG?C8;=D?hHTo@E@Z;Z0)Z;YWijq%dO~8G%qiUNY2{&<
zR!jJYMQUv9k6gkgY#xgynSUdV*F5QR8Y-0=zhK{jL1AKwiKjcQlCTg$?oJbL6Y3b=
zexys&F7O93{O1qE%d}@@ZI^l8_@DYs
z>I0;RKT~C+$@*FKaohYP!j{@*o$Y7E+ck@`C_6gaB`)u>w_64mi9Hl$v)G_cE3M|o
z-k{|7r4-NCw5m)zxamvs=$FS^Rx!mTe+NyS|g2fB_
z1~q8z<*Kqt3LX;#ZaMlKMG`OS{sVh4l+`cCZ;N%Mb>d}Y7AyIpxROJtjQWb}Iw|?t
zI*nsH$=5oGpWDREcVv&PwtQd}=rSuR+@D~;#BckRrP?F)iQgML1N_dhYSOUkKDd9J
zxf8fs6yRUN5M#$Kg6V)w%D7jK4ey#4U;^oq4u`){Kw$Yh{s5)==g5ZF1mdB&kwQua&<$`&VX1D}>h
zx6|?8wN+ty)_?A5=_oGj=6(RJs->;}RgMSnU5(s?y?z8!=FeWiW$_xom|*m;?HAH-
zmS#`oMPZ!?eUr4~i(WEQwyO6BM4ng3n84oJw2|}ATh>W{u++;9!X8k463MXWPthb4
z?=QW<0;KcN9i?kmT;B;Iu|}UBFr=w-ArIEF&1|I9c
zULN)XFSqd~fhOXX^Y@%yn?(*{G)RGoc3RP`epfHacYO!hk(?VTVz!J6k5|ip4d%lc
z`@4VOjB7g-KTO*_5q8NGQBbK~g8q?P?piPWjIQ|3rydxMFOOMp-1h;?%W6F*as*UC
z7u2mm%lgmIe_o#T8?2Wy-;>I@hVv1O)=X$p%0Oz8o`hj^rr>JD!D#~t=AVu
z5%%>RG1%kd#X+F?Fg%h{19$fvHTc}geY>ZQS%wV=F12Vo&iIw-?-rp{Q}=qUBHH+md<4
z{}Sd}c>qt>xYI^Uxaz@>$gejaRT~v`fp|}4MdwK<{O=-ZbZGGbl=K<)Gqt@Lan1kL
z?{5RZtedj=
z`XU4;X-1G>S@Wl`zI1`UWm#0=|FE}#X#7a{ExKYpmiE*8bjhfCzr{6(LC+15+d)`Pl6SH9W(bYuKRBv{-u#a&uwO_ks2@J
zyLA(ud@bGE8<3W{{X6+c@H$QBZg>u0Oj3RoFZ!mx=0($^Iw~?=N#Dc#1OQXu-iDDs
zs(LWFeMl#~LPzu`;2iuruZYNs)0pqzd9(TFK>DYI`AgT$KHW|zjc6Bn+y<5)q{@+DP4abHi)39))44)pMOK>56GQAr~vJ4U4Q^cw%or(dD|
z2Lm|*A^UUk!dtgf4K`&bKU@=p<7YOokz{NxODEzuVSU7*bX7aDP7A^3D7o`
zv2-c!sY&Tu3lnR4SEeTrVXUO+8B3WWToqnbf}MpLLgT;1b&<@*8DLj$pU|ONZJ~nby9j2W}a}5t~b0v-fvrbl(ab#voy*5va=?e>6_jcyy}b97+pu|F1l`i
zM%3we)@BJ)W%!9JLto(^?g}0O^X}kEYL5yb9f4IAH*5`JsQP!2vZxePh?}Ig&jNgmi_`e;@8^Go4C2QC`y=|QX#!?dYbe|a{$$n
zo@0y6;(+qUN`#bJ$YSGOAHxBWTtwsBKO@_}kf>EizoVG;;_kmW@}K!z9jQRf
zYa$18z2MQ+M3lZ=lSez;73AQ1WGO4G?gtHM+K0-&scXFg!LIC
z;0SS2uFa2#h>J|e@u?Pz)wjm*uW|f;(%b)!ksWgmr;x;?iNA_Vhdqv)3guUuLmG$g
zT_?dr#DL7@II%(ZCp-|FcLyxYNW6zz@Xx8gd)LK+DSc#oyjT9@jGwLk*(nB-^&W(5
zWP)oV=UVp5S$fl+h034gUfW=PrVim<_g`xHpZ&xisaf^sCl`iI@N(M*p7#H+SvE)>
z|8HURkC2kK(X0GF67=tAo3>~+r@f#4+)(n=&G+YkE8$luOwEUNwqLSUXx33%b1yEU
z?4WEr-?*i|;!#*Y)mGkWzIIi4-erGwftY+<&=6;X^`|}&X`1`b1phyNE6+)7FxjQ|
zo0>~DTUT7Nv9j!x&BXW|G`s#j`@E*
z%VyL8eE~#kaeX1;s2)tkp*>=e{n)#sT&w1rt_ztsq^I1YfZP7!&4h>ybe)UEr7WCy
z7s|M)*p*06k0^gx?-C={PA5JnS)AE^p!=B_Tg!IN;#*I@i!(kZ$Q)NEC)M$3`p6|n
zGaf%6AYiQo(1f*nlem*R?D)4^rw5z^0c}MW2aYkcm331pBd)wJ9b>t#TS$q~LE%?)
zwt;DTJ0-X0FfK7Z9KtcX2T(op-QM3Sk)Q}`qE9e0Y!lMCY3pCVa%gO+`E?qgBK5uF_U6pg$zkM5%i
zX3EKQz;#-Cu_Un_^lW|y->R;|=9_IHUlmd*S9qKWS}u+kcUob~f!Po4D$=bm9X%Pc
z`2oK&5Fe4|^69)!L2vK$85h*4EN)2cG9Xh%*`5+QEm??6D#F9XXWu{vxWKkhB1WG#
zuLB`U^faY_jGHNR2$lvSqM2t$41;qew&K4dnoK`FMNmb6mCCtaNWZA1s%YGK$6fSW
zwH^;}=9;;K+H^?AoK``eEqkl6-<3vePwF`=fUUuZ=_VyRg%FU^=x7$-gsU#@!|yfa
zAg875ABHotWdITI!%>x*WdV~GC91fu0TPX{w|xah=epQSS*B1mrib3NRGEF8!ycT%
z4!=Mulqp5H;Wv7#)qw}+{4}p`w6w@Mu*~`JqZVZLZoMBcxXtKZoE0penl~g%c9mS7
z(_5ivK(5UT*$lhG0y=JPUYAIVPM^{umpQI_;r!M3wlX~QMA#!fstm0^*JyG32BvyW;jFq5krTBB199xNIQ0F
zn1$%zGxhLXbr|m6+4wa$ER(5!MxWCTYnOA>Cfw&9s4F*p9Moc0WmJQu-#~HchduoB
zKq^zw`N~V^Oio)gsiE+fKl7O*;gacBQydC*V>n$AUynao@4`mL32B?CKWpKUtc|i&
z7R{i!GbVU@k@VVECsoUXcb{8xTwkah1+J?~QR?#mVKd-__l`)Q@+7Pp%vOrT5Fy@q
zaWV8;ADspUxZD|e5uwI9Cmb*^;@J$V+?zBa0p>V-yr6k_W_hneW4tYC(Pioh-naJk
zZ6>hqEEd2`ro(QAUh@Z;{Ej0=sFw_PmQ4Me*US0t6_2cN0cTGrxE{`HOe=`3a+C$;
z*=!Eesoua*QC0vJ5>7%+of>?bEXXWJHaWQxM!Vhp74io12rAL-t5T2m5;vKedJE_@
zfDlqL7KQH
zkhn9~tXIyLwOzP1>??I&P3cJ@D3+`H9WY33#ZM~^&L@?~1%+=AN6G&8JvZmK_;5IC
zQfYyY+$x1LR-XDtL7mv+JN)>qt%N1x=Y*yD3Bod-InqaYu{q!C`+;dr%KtS*+tYdyH1!wyXaa6TU
zQ(~>QX8q8G8axU{~oe76R{5d{^njo29LJz#Or&yirXs^|*em{xCy
z?a`_F*?b14vB__&6-PPPFGlhyC*1_LFcZ6L2n~uuLW&Y
zZAF8dmFp27`nWRfqp35X>$Ipyf*REj&jec9o8#|SSrqKj$16na?&!RGsq*gj-
z)yeRE{NH-;jUbWeQh>ZNF}t<2SHb5p8GUNGk+^eL1+as14iP^0`-yj%3{YqTWSvgK
z64A`Lk
ztY%~#CL3)}iB$^oxWQSzHxy0`YK%lO9V#d`$m5~E>fPWZ#N%*#GJC@5q#l0C%!z6K
z_W8T*-U=+TT@-@&dmV81_}d=s%KC6?HSzP7G)WobE@|vj|k3ka40yve)A)1+azu~NDv+I<=^P7KMCx{K^jTLD@(ushT*>`
zE~kFj%;$YY%vy|GCSfr(PLnQ~tN+E5LAO=&x|oqZM*?#qMUeUX#&k4~LyT6JD(_Ql
zgUj7rRL>tu=(Og0UE6j!VVR%9eQ*O7Sl9f{jtVp
z%_ZAmVOW+B)XU$Zm*(<8Fn$3N=pP_8sDmREHVj50S>MLuI0DJG=F99Z##^9x!1i_n
z%ElD?W#KORPflFKTu8^i2Y~-d!n9jT_MZ*RhD*tcak`4CFZ^IY#Ki{*7aW2#NwvuBC%VFgLUS
z_X`0#qrvllV%1>H4+wt|8zkdhX2{C*$O|chHL%}ip212JY{#cm!HU7wH_N1@)}IW-
zd%qtRT~r0rl!>m{3z`V^8UsqBf;#u*@8_nOK%FpH%kMRm|86X()qm-f*U1=fQ@JY%
zl&uqTKnYUkvj
z@swmVVu8wT-;dL7mvPb`Ve4*uAi0Sv-cU*+x4X`p%!%R#F@I4$me@wasam0rV+5~<
z{SO^mtcjdRcE9FCwbIKXlI6|#A9jD&HT|S^B)fu{QMF7RNefgD`HS7(bnu01%JG!;
zYL1mEJuE{wT}*%2{at56Fp(9`9nhGpReW1Qw7i%2huz;693szbGMFN-kN|8oK0X4K8nzhwj~2Pyw&QU9}3{}X=ylT-f#6#oP1f2&=s
z5zO%eA))uo@Qy@uWgTr<#xPcw%*fRn={G{=zYxJ&Mf$6aNe#wYmT$=X5Y1&Z`q2kD)#-Nw8`O_p5Q%~a$yWFee?T@-
zR2f9=$M1OYW^Mk;us8P{4JwhBF+!h58|VkAP&bD0CR8Niq|0T~
zM2IC$C#=Qf7pxHhZ^l$d{oB1?3TXx6;fTofHIj3b7Yl#9vNi1r#3QAzdtlLzz5MqZ
zATTAXVJ|yQ=tJN{J;u0DfUW!c-zIR_lyMss#Q*Nt2jsKB2y+r)XyC6((*VvFKO+FP
z<#%zul}}XGHkPAweC7ht2(c5KpYGdij@1x`AZ<3CB3w9TDt?!F|4~y*mJ>d<1%y(0k{C;uW^Zt1Ia&7C8Br?f9qqZ>{}y`YGT6uj5`>
z&jlpx#Lnkj!tjbr(1Pl&%iv(u8Ko;W?b^A6y%VG&S^N0|1(B)`Lsz@CsXDx^Oi6kp
zNDe2*mGR*iPw^E_-KL&J58tG8Mt(i!^jNYJu_~q!%RsjMRYwmp$Drw4$js)LW0F*W
z<~xi_^iNSZlI&t8-q<|OZ(97`p6_gZ%FR?!ZK{15OFhJzv!FWh0z*`oHhkCc$Aotq2oR>Jmkshk>mU6GFO$m!<
z+_%Lvs4C;B`Lm|s!3k%R=i}^UMi6yhLDHAW;>@mE|G1J`Y0s&>W1t+z!qVJMJTa-1`%%46##Aqn#$;Hhu9C4A_@D!%^$PLyi@0cQK17+thl;uSHhDM{8)U1-&)VAyTtl|$2j(8jkv(FSflTAgD+(%4IYy_rWDDSQ4Kbm9Gy6y+sm
z{;r{HtXHo4jMcvkfZq&eBmUGm!J48eAbsS?=C2Q2NKz{8Yi!LdG4F8bLDBYO^V&^UDyhTBR(`~7wlyf(C;)4S5
zzlDOxDl@H)z!h9{8wXAPhwF&4xd~$e`0^JJ(tMyt`cHuYc?q+e2b1aT)9uGp(>?_xM;yiZA~Vu6EMXt
z4RtEXN=%P%`iw8smI(SCs>PNBIxSavPCnxJFbl3jBMV&w-h_H{G3c7Xj(I0;e;XNH
zG`&1zWJlNG>DZTjVksJeZl282sfI^ZZIWs@I_5h&?yp#J&h;99APrav+DSn1IqJHu
z-fnP4KyJGMru82Tgv97s)Y3s9w$}^&S>7O=bv50z)BMLd^sHy1a}!E(#VN1Tnidcx
z-e3$x&WPbWm^ET25A^X`Yd6L`PIR9-$IXKwl%YOg-oPyD(@|c$JHcLiT0k_T4EHmk
z6!b>|&uWaO4M*m5d?@Mnt_CfUC&ZmFWj_@_@BWXH;zD)W>9a+qZ>FQV0&AByAF?>r
zP_9>2FZ`Y{Yp;r0eLwpj!>8x;)y;NbYE^O5{Um@k?#{eCKOJRaNZ%rp*)}QAu7%e5
z<@J-}ffwrP;3h=`PmPs)mdn7f${fD|ayBq)=(->%6f={i_t{+Re@O;KfBe==?NnO9
zeKVENzG|t0ex`(2tTT#=-~LS8<)ZAT_f@sEr!
zliWvZsiD7_LSaROP@$;g*L6Y9jC{66*i>t$F|XTU#pP$j0Oz^GYE25=b5W+*QuH#u
zN}Zk^O60F4;$@^+F!BF^l}a*f3~Ze`Nb(?5TS=>o9PF{T62b;9V9m&jtH{T6pVRD6
zKus!J1OOZJfB%3jiaLwXQFcaic$rbpq-htnU;?_*Jm67vQUyW;p
z9DR4M9(kKZb*Pg>5T_qse6f*uX_(<A^xOev^
z0AGl}R`u=4jiCPh-#*qj&dYYQXWm&y34Sur(3Yia)
zCu0$7kKm~j%XO{HI6bEGXt6K;RJ+_tDb`}69n?lu)ORjS#5HZ6a#_cecJn{2>p!ZDUJ6mGwSGPNO&*yOz9gyU>Y$#6g^!@`>Yvov}pLf?Rgm`8gyv0C3IU&>v&<;($W*|Xbuo4?aWxa*X6a=C_C
zhipB#Y&w*H42WL>D7;+-p|Y_(QE1$=db+^fgJ%6J-P%`A{tf(jHopPal28%WG8Ouy
z+A!MZ?hFf-94CYZJopg{25d8ArIj@JK#z)ss-5rwNbLkQ<^~}$Gr_?6Aut}#@<2eY
zIzRb9kgI@HSc|jdo{!f1gbyN`D{nN?HyV;a^4v{&VG4lu<1%}|5piI059uuYz2oi9
zv(q*$O%V`joD+cz=#L_Miv4a0OV*KgEl`|!r!amD3M{8o*L&xVmt3sZTl`&%MtPcs
zVt<347~`1Q&EBBFvXr$FNG`D1s15Ljp +)+&$wZGys2utHz*2AGg`3Nf;&%LzMngY4!6wi@L$)$ZvkL`PJq`fUp;_=|i#<^lKDVI;m9R(GD@Jl(w9`&hcN
z;qM#vY>F=cG*=fvFfh}dRUWAm`Fjo61MuqfMYEx=l3>(-{U5dRS`ObY86VWvkA&3GPx
z?G~c80%pG+IXHhBGUi=!u+BP7r|Fj!Ttx#{Mb^Cwne*#=TJ?$!62rFXn~-u1<>4QZ
zm60dIl&E^Yu5fWoMJ$?kb2st?Rmc(zA+*F`&al5cJib`RC2IISK8FX#{2q8Z6x5)6N0K5fd=LF$hHmhC997+H&q7f)phnNvFGb)jZ
zNXdpHOi(xc!a_tS?w=65r`-MV0^_}<4w5Ka*JtjwDsmV3-i#T|N>GsVU|ywV^t>
zH7`>kR1*vn3oQoKg$sw@l`>0@CVOu!hK+E7s9jlXPy7N5-|l&`Mu2<55h8W^T2I!P
zQf;4XMr$b=#H<%lTl;IbVAg0k|KJ|3%Fffbt4mDW6e-=1yz02%JaHf)pJ;I1;R@9h
zubhgvG-3ta<659fz3ax9fJs#|(k+cviYR^Gc|E}6-jsNSuzpXA6m=3$PHOHY-9pOt
zZ4;QMF<=IX3=kYD}_}2^48DG`r!5R4TS*$2PqbCfV-8jXVnQE7lYy$$n}&LNcOr5
z2L?xw
zOlWwna#E6GNMc?twPbp&p(-#Pna0o^)C2NMC!)5BI8t?S9KUY#;^`=%BBy<)z&~?6
z{Cz9uKBU5Rn7&{ZS0CI^nakpImY=`;)6w+ug7}kWi`X=HrTO!*7M2HPu19XXJ#%Ie
zDZII8)jsalwC;QB&mBX|HD;0l4`Ej$)n#pwwh7l(PnnE-Tx*bbrkku)$=L)mXoyj*
zO7W1uLbrB;3)UDTe`Il0pvPPs*rsn5L>hiFj#$9ialn?8?c04D5-FFg2UvSaHH8R6
zC6{$-tw9go;QpTb=tt4c#|7V5~58$aq&(WK9dA4BN;27AvA
z9N`Wz{eQX%MN}IvDxj*v%cq^h3iKspVGC{iN;%oD26^rMNRgzxx?(m?MrRFdj>!@(
zt(nolU<`MpRTID%_hO~>?AKmA_0FZU7e3X_ZVZ0#5NGx!B+N?Ag)zG!qwZA6rYAyi|PceqK!?->Sp<_u52uq8ATh7?g|~(FwFhO
z3y2)jcjj!d(iVFBi-K=6H&-e?#NbTTHi^R_BV$5+fI3t
zTRu?oEnn<@1eiE_3NMR5`8+!e{k945_QW9NE7|pVI%D;rS9{dnP~|M>9b|RhSsfFp
zVZxR*$u&mmOf#vXnNoti8WK1D^r##bZ4-uj%;M0ml0pw~CnRFv9`t-V0@tHdYFPiU
zZoVl3l8*#ZucWR>m@wZGD(el*!H);@Tc~a5;j%B;-Mt4GBcQOm;P1dx7y~fY^)QT~
zBK5Dn?vQPG`gq5-FTcJwPuI2-@ePcnR~Mc}HSDd^BRIO1fX`9H|KfuOdU~sG5R^fk
zCe9<&Fso1*t9OZ!FB}XpzBFra#C#k
zBpYm%R7F3Mbr*?%yIaipH88o^wlK9FhVb=U^XJMtbM*a8NT48}o~WMScLUs4<7tD>
z(tTqy-SE}amq#ja=|pu7Yae>g;tIWE{KOycoYbeozAOoZomP!PkF_8$w>ClMjT9hV
zfq9QCwqm%fN^rlE6HXw|EsyoDfNkmvd2cjlT}srZ+lHbVt5H(^a`dMTn_N&QfRpX1
z(r_Cgqi@vHC@;=nOYNkQRJMr6S5nY=yd3(3>_mpfVs%**)G(GJL6ck01X*u*2kWqT
z-Or)-X{iY9+$o^}TEh|7;NYA5rC7)`d+I;4xxYMuxBz|6t?0;%9J*#~(u6?cOD?Y*
zDTcI8%c_3gph@aKu5C5RKgLXOCPmm%O)k_s<49A%(q@9K%BGKj3tr9C9wO=Fyo}x3
zAkcQLij3Bm=#S9y641WEv!A9zXe$Ts{tdBK_JC}`w
z40I2s0;w0sMDQA(noCTcUs(VlWstfE0vzgW>Q94Lz5KZLp6+jqqy-r1X+b#KNyyE<
z3yLx{p=6~NEezsy)lNAZ{XzVGEy)T6{5uglV!v3p&|gkX
z2LrLf&5S{^8&cKzkpm)(j~`o$^^iAHp49N81_(mpY6+pS_QJnoPoymU5P_{4F{g7T
zr=Tsx9R&f=?!r`YCm>VMZmr#g)d+12sqt+&?nEx>*Gl4bVh^d6E(~4~b;bXhCpn3d
zknuz#yloGq8n?{SWnn8=$?QwSGQ2E{7r>z!v&%b~4pT*PJC3Ow^KqDt9YBt-$Gm}f
zKW7KA>LbcHJ~nPTg~n$})nX2Nq-B2Y@W#WOgfKdw4-p0S+rzFaf>w^j*=sAiIt)AZ
zo~rs)Bu~q2v}6zDW2`%fXDD3l9?9pVk8Gvu`L!}x^#bRyBr$B^y`u8t_jZ7_t*IFvR`09}`!8L}RK>nQs8}FpP#(ROHZ0O5(1^;#X^%B3k4CwD
zDKvHwNh)ps^6E&UwGEj@E1_Ai^jbI<*nah!>scC(LxM{kyqEjWjgGDSdUg!?$4|p-
zqOnZb9ymh3FG&0(LC$J|ZTWLXZ|jaQDZgeb$Xs1Wa<`w_j*ZIvhdk#JOKJ|N&!>Dz
z@K)hGp^sC-o?VFbfo@YAI@X45*y$BO%o{)N6Arba&BWhE#dtS?$q?lPX;Piru_3=r
zCQ&Kh>S;@tl3(jhD5j6Yeqz@JXm_Irt2I${pazm5lDYJyyUb$QP)s6LXV`8cHVuW$
zBU{Uoq4n%7Zf_#_t>x}%-ecrEs_BAI);U
z518*8+{bGPwj5s|Z=tLEz_&OBn8Ggx=mD72fQe${qs+p@@>nIX#t!!&$`9%r4sdA^
z=9P9XYYXSH7eic6AHJ_wl^Cs+DuN`=<@hVbJ)q*|usLHmX}}Dnc*CB-1Ok=XrwW~S
zgsmpb{&F1G@0D)S3$dO!07tyo=m~hhL>5iAy&kC#qpjvO0+H}R)Lx-&1#!5a(a{0{
zjeWpH^`Ag&bRrwrz5qV$a&FQ3JqZ!9GU%EHUIzs=TD>m9YA>z}Xk?g(S62Y@X^zKa%aI*6;)+1Eum6Y?E+9Co6EQ
zRg_OK`SNn>2^M|(m8FWI?H=nZgmJ?ZCG0_o4jdGdzpB=*wrRSFZj^j2O!RiQJr4*X
z;E5FOo29pCmmR_Q%ki;Eo@Q&26<Ne_*>XtB9>$=GIsHO?b-K3JUX
zqL}-ENG(->t}smlGL1Mpgs%hrhqn;9jy0E)!>xE509I=aNcA;_?OQ8UVwIa&$3;
zkCrN=!AWDv2
zl5Nt>C}Ym?#CfFm+Zse%g+g4OvVi7Ws1!m#_ZWTb10UiAXf6TMh9&eFxb<3H6nj8Z
z%lcgv@6k}XRv5H_6>>iFTo7>`7D^z4T8rGfU6H`
z0HJ^dXl%F_IhD7pTe3o>kdgH5yHt(7>-)=jT#q3m@65=wg3)#{ret?-dF8NSK-6if
z($H8AdexSQ?~pL_YbD2Fs=P?Q^sT>y{L+S}rc*H3{3Qt&K8wg&DI1dzw}k4$UHsw9
z_c)fCqc~)o98LoXtkZ&nOZ)%>TssI07A0D$OSYhmA|gb*f~zpF!R`euCoBDqif`k9
z!csAh6MZVeJBycOM8RO-1RpvmOHwp_#}C;e0S0iD!w1!9?vlem%uRw*8~SP4^F(AY
z2U&+&}?6Yw}v9^7a^+NdmW|uPuQd{)bg9
znyByG)K&FV+fx2Ls)qf~zeNpCIiYSzlRBnqqFF#B$_eAZ-ozGn_r7t0Fk}z_7uj+
zgTB=PNVQ9)IbMJz%pl#M6vOP@oW8bO&zqmq71md^D}Nto99tr>w-zoQ={}@+L*E;0
z8Z%{kbGA5U$2bb|*E$=#G|rxcMbbXSlT55CVsc%2!!AGw=><-(=1V$$9zH`ZT<@ZY
zypM6rpOOUEy166xB!LY&1%31p1F>8Mo+D@4`A3S
ztn3BJ$Z@L$_s|5RND#UkUPgf1{4$P1f{U1a=8}9gH_TH-jB+0hW
zcWvs&`FSWe4q~Y++|lMnrN*+#el_InMI8==AC#XHZsK9FNn|D6H&vMcT>&Ac6iS>?
zO`Cyo!fUM-!V#f1@A=}hFN5kQj=kooF5=1%mw|oX$Q$UUwe9p8sKx{IR`xkbQ~fmm
zam#aw-KIfwfy*8&BxKM68oRDm&d+UZk0Xp@uR{1Yq%v~mT@Qf5fz#bsMN=NC5F!bJ
zy@VZJ?639luS)3s9Lji$9Q(xjK#nU~=A_kyhoX77@?%I30l(RAM`9K4;5b&G)R-+^
z!A)?23po%~`!oFS7_%+uqX7qVzP4bGhRcsN0=)U|IQz&o`)S|@8=pvSQlh@00fA*1q7WFlaT|(MtTo8bGiECH89*=IF4v=LTBA
z2D$uA1s)OQC|3xSa1mO<
zw%Kn#k84jxb3cA^4a&?6KXly*(we0l@AHy1;@Fgzs3ev{9k6p-IWipZOH!3Y4bX>N
zd7+L2hQfE%s|$fd1TpXHN|O&g^-BDI-fcW54I6TLBQ
zq}gJ_#65wm5Mg+%zLcfQxHSrhnU4OJH-kDMJx|@@TvIl2Kp-DMeOVi3wbz=x7WzU<{~1?+YtTP{lq9GN
z^_4G7-Ug3F68?z-Xn{WA={#$rW{E-i{-p6uT;-c)wx}-C#$nFAB!Df
zKK`pBycYeo>wq#@CK@a2fHEEk_{aFq8-=+q=YRX&$MrXnvc5%HwC(0>0IFD>J6#6_
zycJRA_0k!{$4IIbxzgoL;Kp83F*
z?NiPlB(HWDZZts0W_{GAC$CP?cV|7Gw8aspJ9<3>QXbS-NA1Lk=t~fxNQ>oCOZZ3TEL}GwGjNk=tHsvO8XMaRbe`XYv!s?+
z77W$(qRA&Vn(2}u{HN~XQ5OxIGn^8O$IeT%{r5EeQ6gfR5s#G5c`z7o;uO2_uMhLB
zMew^U*TdP=xzs*gnL@8EvuP$+xnGMBkAdu@07&xMgPfk(>KxK%&M#UaRb3|aaen*|
zqw%^)J!viV<36x4E{mRhf%;If!f&#lnQr#l<&jU!E5K~ez9Us8CD+OIN_soJS<5q#
zcPH`Dh1T;X(Ni6T&%<(+9|)GN>jk6FZHzy%KOZ&EG@xkijt|OswaKp!3nF
zh6u5dr5}*Ix0=i3Y2CsSaecc#fJn3Y-k6!?8p|JhA(kUPLPozItJG}_mZDu}UP{-VF|l=)EMB$8)UlwQ#cjcMcd
zzgRn`B~gGV%a(21wrv}?Y}>Z&x@Ft8ZQHhOSKp55j_B!_$9c*x$jB2X_g<^J_z|VA
z{x4LMea61_nyj1t
zTKbWte__>=g$E>`eOF*&jN1HFlxyy~}gTEuU6
z)_iwo4|4`4CgZQ%!$UM^Z~W(j;rk`P2&wZRBr2imS=J1HtzAg%cVZA6jYELabCao3kR%F?^e9BEQkRO~0T8J&L9zW%nJz
zfooXHN4@}fHK8d9yh*@@{P35Ki-Lljk9Er=1`gw9J@6|QAZ>)-&}4;Z!RBb?VH?9(
ztGP%hj3~4X$r0|gzPfn~I{WnS*o4hh`JEWzzNX6+sKYaY&EDp(r#wvOq0y#IDsyN%
zg#i3&eLwe#QtI=-2GPVJ-)9oD{dHk)v(0nS|8h@~2+xf0CIENpzhv)STA>EZHk^R-
z`xb5>(yc+7SRFIc7&E1}LHZNmGay#pDy0Nzg
z?(B?SWl=KRxbIOHalEjfB6eU*Y;Nn7lG2-A!9x7;rHLQcx@BBik-v4Jk8%MMr#O{><^E~#IUt~tO~!<|3s1;S?j=24rRW*jnitV8gx
zT3AWsUyLQXRpfZY5W<>~sinjpH%EK3r6~evg_ch6<*oCNM0)9Tm*j?XNxEti53EHgq>LHSJfNKVkZpu&n9?ONb+d
zpJKA%?v?~CkXkX^XuAZ_z>Z@F0YUKq75ZSCUyF?_2S`Acbb7ZtOg;Sn}sPbP2n
z*oZV}}Y_)osQZCc;q=(#PI
zm|=~unty-N^3?Gd9?zb)u(dxziU$N)^ZW)>U$yg@Y~oej(L%x$=A
zCh;*;`_K+sEO&pQKl)hHc2p*b7wj>q$nqd~zcEVtJAV-Q*M%*}F2oxQ&S)BmpM)T8
zogiO*oKTQ*ewoH$^aj}Mq@$n5UM=EUnEUK?A7TGiHDjSGnQ
z%7fT~PW^Xb^c1lcW2eZANqd6CPRS)v8tO*Mdo9!kVDP%aqx4l81vGo+{(fl#yHZx)
zIy)~!=b4bclUU&k8^cEI&b(J7heWSB49fnm_oaf*Wo`B7Ep^f((NFi)tiOXq8Tc<+
ziMvh_%=0^@ptwUyR0R&9U0xuTH-Dy*zTgOkHo%G~T&ZPBv+$`&StM`XPfYkX5IuhO
z&4`=nXN-RLF7PwKlo)?liYRM&_22CZaxmJ@;&V4MjF5X)NXF)c+oXo{v`&5V-rW0?6Ist?=(mUn;xN-FLQJ@;r|m0L8chDNYF
z7PXd=f;!l2n#U?<#)$sooIa_C3C4AYkR;8-8LMFZogw7FRL5ldgEr1NWx|UE;2Y`-
z0!?pfbvlwtcF7$X!;_pqF6_dwBjr77d2P6*Ov{w`XtH*Yo>h-xztVLget>jf{^Tfy
zLxEj4C*r6pSQZ9~b;|p__C576b;mNOukT3?cz2iJ`!`b*cK?kbX_y1xH9&812$@Gv
z0HCH*r*qPxBeLnaqu2@EZ|qxp1bIdRJ7!pDS;%?2OIeHf9~XWF5y%;PUieHA>I-!c
z0A+ju2|04g{xDB^*NEdSPS&jkI!XAf?W#`4_;iCBbhjsqB3nfqn!C)q;*y5I-N%+YT3&H$hRYj|TOBul=kjBH%ApP}~5>+;cgSG5gg)dq|Y
zy-ptE(T`GI^MkcAliPIOZ(%R2kdnY?EhJw8J1Mhp22QLXJ4x{t4^q!Q$0uSoSZ3e0
zjr%`Y%||Ja0}AqW{y3un`@__hvMh5(ihr-4(h)`W
zHu9hdZ1)1+vrK7>xDNc8DRHfpUumT>;4ET476~~m(?>Uaf(CQJ*-Grfmb02um=vop
znE|gMpyhd%jJ}P=h)G|E?2&jbUa93Nxg+l__f58-MHdt>xRNk!9b)KZb+T&Bl33AN
zGo9&z!8eFTIU#Kr)2}I{x)>gP=K;?s+IYe?l^k|iNRP|L%p!o6ACdd=Xo@omO~_R_
zFfOi?_Gs;?tu|no@SnR#g!Q2o!I7Rerw~xliVgK@h<`G3Wm70+bRLsYFrfIth6Oar
zo=%`;^`U-{?Np+?x5T$L=bbT8KPaFj>K@BGFqDZiBQi!?SUprd9h
z?BS4YOTO5~&jgvJ^M`hCf{+*JM1>Je9Y#Y-J;9W(Wg6l$Gtm4KOYbW}Txr&kq@_^1Td@!A
zo#SM66Vh!3@hTrVaDloNCmD>V4#N1L;wpKHVgRp7oli{GZf!*>CV_h0ax|fx$SVW#
zy5Iu}R{jA2Vy3^tvA2%pRGN#^*3E*x7>FbF>CyH|&b>xtkM*2b^UvKSQEeo)1cJ6E
z`E=AxoaJ9O`v%h+!d^SszUG|5dleN5)VusPt(cf*YiBAee6jSH!n}6h&!*v~6;4Qg
znI2Od!G}CjBW66g+xhxx>4S!on#YyHvr@|yS!Mp|#jV_pi2Gxm4!;0mI~(kci#3!R|pg6*F@4_jXiQqz(U)y+bV5{Zbhc9aSX(R
z!s|#N=PW8yd|#Amp()YsbSq5m%a(v5Z@Ww0EtR3K1^ghUnmTugJZF~tBf_9g
zPqHhM-Zi?$CJVJzjIm1&`Q!>XyVf&0B#-`?+J><#(pp7O+EnoL(9U38uBd}tz6jcb
zSEMWsFG?=sSkaFpjS2*>0L|Q~8isf=_J#n2#mdY@2@d%;n-5#OW;+)N9E-Iz8IhCM
z4|UB!!ouHjrs~JZ)CN47V*bwdS>H3(I
zH2!|b`oC8Ezi9u?Q&l*D%KNuv;4~97M)mqn!P2O;HqB^4B1^-tVHAwkI3Y}g5@UiP
zF#4M_Z(3Elkg_v}bhg|kInUL*OO*0y2x=GtwXID3ep~S}^BWdG
zr;tSCZidqWU@VY}!HQs&%I_N{9>#eYpsxHD^Hq4V>?^_HrqJ7bd{;VJc43gCB
z=jxgxkPf3LzU~;;(bHo($FESmZ>@t0YUPsE2XCKZV=THT#Z-5gzLBJTqgAwgmwOZ=
z_R{D`0hW&1<2&&e<*=-4p3ZWcap1V(95%mL#7p?taS71~R8Nef%*)$X-l-?un$d}C
zipVm_w4yK(lZ&G>$A~sjZLxZH>tDit>>l7|6s9#Z1=$*(3iYVdUgH#TJ2T|&v~
zWwO#VD$(ma!C~(qBy}g`Qkc
z_4SQB9L)QVjt&sjdH8Qk&EpN~Lnv`(#$|QcfO<^>%9K%T$8HVIcrTPEyYdD!tgx7}
zic?GV=e>vg9JDUx+&rEMA*_mG2+ltgNHlgRrKda}h#+@GJPjs8Rxm_y2UU4B_e$Pm
zo(nqyyz^-e-)tlD%t&hQ!9Jo`nwQ6SFz^JdJ*I@;>J?xbV)%&YTV9M0WnAyTgx4bS
zYfulC$E*4;!OPKzIv~)E$!{95wf!|Yum4JX6A!s_iRd4VW;Mp7yy%nXvi>3&NKjGH
z5vP~ku*(*^5~l<}wtcQHH*@jI+Q#yfPS0FH8bNjgb5p$=O)wfF;G~TwlZ5j}%*4+J
z=S$VqnJovHKoLpxt%mA|B_;n1?{gn!
z2&JIr0qr&`^-#yoB*Oy$Aa=|x=;{`ez1&MMe}}Gv48f~2fObI9iO{d~(!eMi!Sbkr
zZsqs*BWzRf;qdJUtZb|Oby|M+j03Lqb|l!bEbD*pNLq0%Q&p
zby%*iK+HXi0R^qC%G!q_WsyA`G&#W$WjaH*;eS;sLZWubn~v#1(qaeZ_izlXIOecj
zhbFqwnfzi>BS-ZdvJ|<-{;L!oXxM%`h^#8SM_gtM5tES~k6dTsCkl`UtRi3FuT3iD
zq_CxXMWn^2;vazzkzh{KBuU9FKXLal9@Mc)QKjOYYi`-}4Yg~jczs(b;ttfM_Tc=N
z#hs(}zXVsRmRUjDD6q@(-P^6Em
zose&rbIT;JgKIeSY@rC6e{i5EnF3dMcE%klj;Z-|UO-qs%S;XTmtbiUoFpzSiDFS3x
zgAZF`Zg*CAq1)lCq;P|>8=76jyfdQ$VO<1hoi>{KnG2E<ANCZVH?;7u^u7_+
zN1~3mT-$|!Ny{&vRS|w{*7cN?F@1oA*+;40EP#accO-uZy*x#7j~Vz;f{BmGjI9RH
zj?l?b1A3Q?XPiSH11NNy2TX$(6P*D^|DaXbzECJ32h*@hTlXQPJ)l67DnhwYv5;hBLLYIxS8%XB&g_!$U`c>QWmP66kLnU6EANQG$Cxh#$XEkTE5BnU78$R5uX=78Jzy&g0aL6=dn=uGeDxMT<1(yzcx#vmobvi5eJoL7T+*6`Si^ovK=I
zQ{Qx7eFh!*mrL4d-5ucu%AP_9FgI^Gn!Ns8Vzk@FkDN-6;^-2R;3T1iWj6T%oX#*(
z?6wEIxDV)ouNF%Fpz
z^wieEW##w!L87m<9JAQk(CW$O4jX*i37!KKq}FQ_Gpw3uG@gqSX35H`0aBI*7j0}(
zn`lr&fZ3y3;`;f;rg6{>yHmIEyqF>3Gw5b1o8#TArcs1Sf;!F4Laz3VFImCgM}bZ<
zaYax@;Pw*~NB@mD7p3k<<|Dj9V~3Bixf_fP(dwV*9rUN$3n*=7B_
z^X(zdNaACVje`z&TSTlpA<~+&;<8yn333*swli2oswtUW;utky-VkI?z@t&6pY!Ud
zHtSV2L-YgCaYs2#sua&QpUzr7=iTo=!ushTCN=)XSGh(Y(NG#In8>q_reJC13Vu-5vRe%*0nV*Pj(dHT+^>Jb6pPt5;9{X7xmcMZe23>
z&n)b#(JqUhSv842ME8|bH=SP3&O~9Abd9$D3{N@h?4)-f$8>FMHz!llO#Myl)`MM=
zo5DYI1Fy&qrDO?tEy?(7mo3e@0IRn-=Q@U!1SPP~9mIItVVvw%@500FE>nC5QNfdf
zRt_CfX3lwgLtZWi8M~BDTI*kAjyW(P?>w>_GaMNkQit>^~)?$?oIftiAjI=WupbD|Q(cqI<
z+u1gE-~wNFaC{l5UD|vW;3qGd#%QBV+Wd(Mhzo!lA}tA$W9g*L;1>NrZbq5yn|Dm0
zb^)3nGAf;|&+aS?QihFza?4!eV1JVO`i-f_*~b;1W_od;WX@bAJL=&zNT-M3-t!z+
zhn2@$=_huJgSA_Azt8TfKwBm?5^HJ~XnRF@+Ie{Nkk|(UV3d-LJSEb#(aJMP^W)NL
zeYoiSTVm_9N^^N5Qg$mTssso8HU>q0N&eOiV2Ei(gsu&@Qx-ouk|bUH(<@CTB!N@T5vP$
z?kNwY@MbjZ?qq+N-7~pZgOj6X8|RSW>j2NfjHTl
zk$=>VZe4R8}LooA4I={vOXWWrqL*YafReb(qv
zy+g+;-03*Key9&{it^1z_1TpQlU6H14Z$z|V~OZ~Hf!Nuz*graAu-YLD{0*av2U6y
zMqDpf%CH6GnJ$z7RX@??mqc1~`A65y7_V`SVmgkQGNQ?M|nFVtiwYFysuRN0Mqd*v&Yu8yRSN#4n*eNA98)*?fW-r
zp43$LsR;s%Lm~Y}ckIHs!A+u=Q=)$p(J{)LTf>uV>it04iFJpeEGV;-ThveyJ9Yu*
ziE^w_K*-3?#B0{!AgJ@A$r-&z{SV$~fMSUd*))q3W_4Agj8_P{iDR+B#JOFjaMqV=
z@@fZjd2MvIFO#M9KzYwZiTyj-J%nXcI`1%fFM!1Ozh@*aLi9S_4E3+2G(tSHot*6S
zL8HS9b?PpX;g@^olNGi(*9gN^+(QQMns29}gxw&vYHF~=Dg-h+{gzr*w7KDZA8e%~
zSY-{%y&1;x#r?FGuC}j>WwsiGh%w>Qo7|a5`-Ju?Tm*AN@Hq%axB|EBC%bAbEwg`|
ze_@)s|7tfGWSS5pCddoyE&iT`E?f0ZvVW7E0)Cx}jsR|FUBr&PSZ!@7Tvn*<%8yGJvz8pskh83}a@8~FVsAoL)PAb(m3k7QI
zIc>4YrA;e;gVDSuCW|ycs`~++`q(Krn{e|W1!VF+HDsM|=r2-8`vbZ4LWiJ7?HxA1
zcDQ?kSZ1cQNO|)!zRpfQr7-I>Tv@xFGAo9u)k|g$uo0GE)Njtb>^_6dKn8Qv)E>+D
z14h+FGLGL_jxuarOH2=&vHD^7VrYKVlXfV2SY9~udJ%GpAt@`oZ|~4!k^$E7)}OEq
zNJ|Dg!BRwQhdc{ghufe~_E$p7U9@@r7<~`7i4$&7zg(*g$GQT;OBiP(8xsD8GB|V9
zxaZ6~r|RQK&m@q8?GHT(-wHqPpadBqIJU@Lp{r6Q%Dj_IZmBAG_U@AP>+z6F7wVNf
za*ke8`X-e(MO0VGQ-FvUku#REq!0ZCrMSy6Xy&Ue3hMb=G&QZcCOO7++=t|oJ|955
z*&BepwLwP~&3YX!|JW){Id7e4zNDQnQIScB)Q-GMrXz_nJ%L=7OtL5n+iQy;^_GG`
zZ!c2;X?BJBDOxmE(`Tpgzwbqebh$`3tVmE_nPn29Ot(T5{MC246I(9W&V|+C(BWGd
zcbytD7s;&((w>WO=^o5W#I1z!VZ=DJz?YCLx2eBMsn%A-bSs?`Eznr}M>!(mnhbuD
zWG=YtS&^#zX}7X3b0*}_zaS^?$kS1D*dCroOoiV$X%M;_RtJv>vs5uA>e=?t&^+*Tfy%i*3TIUARI1j&HHC7
z((mRIJW>db^^{j62o+3#9SU;bSfhk51+5_@+JyqKFCL8?m9Di<;Kr%fC1JF3a+p*g
z_zvOHy#}{c^u%uAWabD@MUTlzE8trPUHTWCDb+ufCaDz|+79-d=@g*%7dePZ0|oD1
zfowR4JzPYBS4T22$=s*+-sMK!4ltjeu_U}w3zV0_UevhKcSmR)L}OIo`l()WPi)RA
zB;gGw65IbE{e%5hnf~{`UDiKen3ra2p6vv*)M6A>@vd&dn)xs7Ih>W(!5oeG`wX-
zFt(ivR&>IpX5Bm(*>95ESgZiw#X2ipnQWi++`#E3AbaId0=qy9Qxr?OW0uQG~l>XQuTs%Q4tkx!2$u5@fzw8?u9D7zW8&Jy>GK!CEbH
zyF{l}O6&HUf*XZ|V_WqyHhmCEF6=L;IEb%}4*sQz(B%80q%wqMSjby_l{oBDr!S5y
zN@+YdRL(O{_lNIMxH-nZK_rRx9s|5CA!GP?=lsrk=D|z-=j9&+H
zxQUVM_%Ix}v-(A{{MYAKTr%HU$?D+YM$vHUA*eM&h5E*iekEhs6FaYQCPh*x&JY%@
z&FdaCadQC>H>h$Y^a&LXLO6a~eFdOuIzDm?Y4qOwH#5Y{Q~1z-7(^MHZuzxKxcq5i
zW!8|St_S$BD9tLa0{+5GJw1d
z*pIm+{F}5n*BxDKnqz4uZV5ejENBSUA;;K5mxM4`v|#%q83VkmPBAy7BkQncQ6XiT
zqx*oLyzvhgUrr%^lFuUwR*fDJ?_>s`6^|U
z9K&9Ea44!|Vgr7Q<_{pR@M|ZkR1Zg9t7pF#3NUIu*vQ6DvTnDf72^fn61aUSOQD43
zJs?LW|D%VH(;rPr7M5(PZyV8pc0ZRf;S6oXhCJ{R;8Ax#{Oh9Vwl0?Qc>_a9Pi*L0KAZ9dk=fxm+En0asSueWApn_OuoRB3U|}#mO{0TW;KaG
zh1YhUT;+0CZAR`Fn|!6*F2%-`Qz{qXRvixjNkp1J#gKSG$B
zwC^BYP%NcwQVW(n_%SQK!yM*cRZF7Xv7c6){CFS`5@MY{liT20JAHOSb_BIop`3XN
zi$sL37>F@qIsrylbul`$(db4T=G_%PQ&spOli=TpfugbU_(ng42Y!Xsat7GWmPBM|
z4Kh`%3)<5YCNnEhASe~j-Zj!X*k>A;!f(v8*7i}#fJW|YrPtct_}q1{yIae+S)a%Yg0ucoW1YLYC(^>)3AQi8&6o%7hZZvKRbzTif%R!;Apj6$|s
zHtDL8uO>3>{GR6+kt{0vd0*sc3=?xYW_gg}a%$D9ztlh`i|Y1a`#udJPvC*4?=~Tt
z3t+*aY5fu;7dNk)VJE!`-;Io<%afs~595uVA6E;{x6tGq0@_MlXRlNWqDP6)+RS+P
zxM8s$Aq}0Y#Dd_XHC~xv>%ZdovPFQkLW~x2wD2xli7KWr%f`9bcp;?v8TUCCjMcC0
z?eHy?3c)(a%uD6EeeZFtmBk_#{fiwXP2>R$Ax%#D^uc7tClMV<+egujwSrv`jqR*$
z0e{hCJQ)-k7;Wss3;(IM6ypA3D~x?Mp}35uF~|+Dr1hj4-{{+22;^u5Y8PWJ;|oKc
z(s5(^O67iag)ECd^0Z;SjdwX-qt^nJMs&3l^O>8n{FW}C6brwj`6J(f)+ZZy0@xof
z1lw3qtay*N2(TJ+~%;vuE;GKQ;(C_(K9ZWidTG(h%z%jK=;Z{9UA)4x;$qbUN8
zTfa9$+@IP*ADGeRIbs
z=mr9xp=PkCYTibYY9V}F-K}Zp2s)6pQwB)GW-iv5HIPKR6=si{_pG^Xt_;Ao5MO_M
z)#Tw^C1Otj%#d$8bjwD{&q}lBiK}P$)83{28b4JU@YZdI)Qy^bs*0M$e|(7#r$buo
z_7v5=>$IB(N2(G4PVEGj$2CT(lD{jO_^#1(L^$xd&{m0go#3<(>mr}(=A0pk9>3c&
zMga|~6~=>C+}-VcM<
zNowi+YwR~OGQx0ql~B^f;M%)?dK>0{y(#_<83!_|EbVzp{d%2|?K~$ar4nOKd|KNQEq&F=ou5JFY^>n>ZxakwT|en|$YtCe`{Dg%z*+WMnC#OzUl
z^KeK(Qf#VAEWdC_H<$9tA4e=}D=Z5G2WK;yxT_Ee^jfO+OVvHSBTQ4ii%q@N(-!*c
z9k|q{W;A|8n9e3MTBwFkF*~&S+NKRR1+03l=9dVy$_{Pw0BV6>=JUX$bTC=^3qsR|
z(+~^!JB)JGZ&>JX{3y@%{qA!ow>U06hiK?6>5FG^;jDY6pWlb@L58lW8q!EnfZ_A*
zx%}8MQ3KbmsrUL&d`wkWPJW$wU|&uzXw7_wPjFc``80}$_*O*rXo
zNEA6e>~Ih?Y26eBD>yYHW6V2_d)r|RZ*dFJGoggNtzp_LLaT=eCyH0#0`1`sX*1SD
zk?x+U0t0gM3=C}4`K`m)K-==+t*qSg9g0w%l|*O^j6&Riq_EE0y4lOMi+(t4f5L|0
z*NU)NeiV=#nS9nF{zj$dtNzAqwouPnmO2_D4U&koBS+{3n2WsaDqWEhep|z}e0Tlv
zvcpTFPJR@JV*~(cve_8A$}QoH0f)DH#&aIKYSwDs<@Ey6w<8KQp>b_-L;*Uu$H53d
z+kYc
z?@yzJMf~d%C;Lw3JTHbz*t1J9ELx{OOc^KJueCjkkmsQ(=yi^sciIkk&_ey;K?!-6
z$DYOd55~wG-Jzyuje?13oqoiJ)Y>KD49+?IA}zH1EH3ORpil?TKzj
zGUDzy?4=%2S{`Juald3b1GDN&`6knd?4_H6m
z0ZM2pWvdS@2m~C`yC153oGE{sAp8Kv>w*IO?iIYg~jhI%azgDL3eKX^uV-X8eVlMLhjag$!KFPVzm<@oGc?*zr${JST>9(
zi*WO2RRIe4(rn0smseNew!xHQY*-UU*hyv8sIisdSyTh!>3B)0XA)^7!F+qz
z<=)jBPz)fr2=#~YF6C08=12fb_Pts(3b%8-n6p0g_bzcam^Q_#a#k6t
zBCiva9-=aa=us(pd#-`MJ4^A3l=-KcHP2#b=G44)^WjoL_UHPJ?S+hjC$e1+C5~$^
z5&Ocr70Z9;?_F%YD#ymT!dELHQ4uKMJX42eI-u99&;OwHS6ZAV50JKVIqd
zPE9qGK{=Abh0#jxiC4|Owq-Yb@|C)%s)Uk(BT@b4$mHqO?pd}i;~e8P==o^qa=H!i
zVo-A3aMRRXsTY9I>o>wD&740>in>o&pBJH^?<4%a$rDVTl$rKVklHvljav#$}~#y
z+qU-`^W}|gOEn)<+$b-Y5c>d*o7&#UuY6$^f~vEwHnqbAXT^&iAemE9_*runKhY4=
zhu(ox5A3`*vq@FfqT;^=i|DaajDFUF9W^j>!C#1_K!=u{K~j$x6Z;-@q!uFc%3~oCnwXm?`q>5V1Jh|d8X}DaM&#|ay;@ZTD*V_GP$rQH4*v
zngC9wmggLxRYNp38+6ap(OdE}E9Fd#&v-ZlAZ>GBx!Ic2v5q#>SZ;uoCor{C`<{#y
zf{!@~t)-z4EiENW-NX?`BqZD84VWKx(0H+QJ`Q3ud5OO|YQPi@g(SIz;=bk)1v7a)
z07Hm^?ecy2%vZo1(3jF}`?rxoKvp9-_egh;>q?4%2*Vm4`x^3VTWp|)Y`d2)g0{h)
z2fv)WwHlm?lM0adMyh*&U$6FR{WL(D=iPgx>P|jqQA6p&=en2e@z1ImHknNY1R7^C
zBA%gMt5U!8`sAI9OE}pJkYXqjk-|j$Il0Q*8%jix1XRUwsIuCRW+@TjAxC>G@q}i)
z$-%rHwh-@LMw>1x&@|v%B4CQZDWDk@m5~eR!f9CPR{?}Eu9O&%tEZzg7v^AG_TGgO
zXcwQ-4ke+mf>?7f4m%1#Klys560x?4Mj1+O3X)NuDRfIQt%tne;69WV{7E=arLHMrLxV$`n3WOX*kPZoOE;Ok@A;**ze{DzN>JR|&BV
z=X634!ONu<+!P$L+CIU1Fv+N>Z;U0RbBH0)rqMjaJEYQ~rvqI>vM-|vZH2RD$OTv6
zTr7c1LxpLrizO{$;*0lrMiJ7H=MDTCJae#o9t|eK|s4c@E*i6Qv+
zq^L@bS6s{n)=yZ6u*gq|B#AVUMk|p1pnG#WFmHR_0vptD-9Z-nEN?dMJ`+B|fIaB`LSKA+@1y
zyxx3&IKHGAZ`&_i#EWq)eAgsMDA1O80$eNFxwu(_mdx`Z=B2ca5X
ziQ7VJ@9zj8k$sO|E>W|PeYo``ntGGP{>lmq=GG?mnm?
zSW*)Us)9ceZx?lxky;iQ7>w>Hj?Wc2jpYvZbar&X=?OOl&jtE``myPEz}ePP`q@Y}
zLS0w&LXCWQ%gH~kvz?*%q^@bfcBR}^sn1NVZZ!kYtdleyJzxSyn^R}`}_RifqSS&mSt0T{&n6Eg;)B7fB8)+tILlkO%>
z_!t%D
zF!Tz#jk;pm*9J;gGMd6@d$-x#j2l(GR)>})bUb>TS8ogGk;mrBTO0v6&l$HF8PR+N
z0%Z~W9L~tPyfJoCr`P?6^&2)1U>koN51?FJi}O+uE3dmLMBdK8kd3{shj(o}n0^r@
zq3SVY(eb9fUzG*)E93(deuvrh+KR@hvG#k&Y|An4A=Q=+IHp?3yw1$b_uK*PTXX&S
zg2~zIg?ELn##IsP%YaG{iH-^Y&0KL;^1ycRY`6t4LYnP>JvZC9`%005@`OVp1YX4|
zLIwmzlroy7khYJ)(H&e231etJS(#=F<_prUt&RFk$))&aBoBVJ1V;H>Y9Y8eR9@HE
zb}V~-HGZC5NpknV^2EWYr@_6qgUc)bb@CdlOkw_QM7VRF@@IrR1iHf<$7>w)K~6F`
z@Y`Q;VHl@Sf@$8x?;3DB5QqdhCk4#9zIXruVK5sKIQ!i+?6X%cAwdy3)4fPJTxp;V
zF}bHQ*EOAZM$
zXvxTn6gfki9F$n-B}i-*fFh)M18Un&Jit_KPMayJB%8h=kcW5fEV<~AK2iK{>JNX*
z<{J-+F3wvw^JjqE(_A9UXQCft&2SJ*=`JD)y`VF+ig@h-*i_(O(YH3?u-Uu<8GFTH
zD%&gwq#;E&-hEzecT^-+OvgNti<6{he(sr+f&fQ0zEJkL1h$+l5uPb3HV^{VOBL)F
zFN_GL8AuyO4At}F*=FjIW_g102~%xt?aihMycdQF%Q{$2IBw`J9Gh<}dxG$y
z@H&i36%{`BmWy5_^o*VAkcD`Ncn`SC0Sv!+IMCXAqyz1>wjW))Ak}i|Q>>hjG0hU$
zDon1`IcWCUC=}PXI}SpADp#d~T9*>iIhK7tMF8Gp;)rtWxJvHQAQsAM3E(HLti4~-
zcPL>)z*x3c5)57Bz^i$1scpv#9JUiDobHW*W?DC?9NWt@=|v}x+tdN4D84)uDEgMZ
zA~7G%I#>-&F{7(
z;1aiGRlT+=)w%RM)1*^NROKuJDn(uYAab`tcnVnA6*1eGu4VokwZJS!k*G&fISmuS3R*S7JL3bg5>4?
zQSt%MxDvdd%oK_9f00>mA6=13std_A$E(vA|8-skGDg@581?mUa!Dt)wPb1EyzBBT
zFVh(Lg7wmHHsisNa4b`vaPqVbp7m%FHwb;iR+zLH-Or1(yg((i{|cu%c>Um>83RW7
zQ@l$9Xpn*-y>T71_ku1Jm)NaX#=Pglb=bxk`@>xLUAj(oYd9|L#6!?0i
z-R9Ex`y2O~JEzBC0<@RM#|Y;NsLX^m=VkzkI5T4LjOaf{b;%y$5*_DLab4-B^1Df&
z+TXbso4D0-B@ffnB#An%7(y@RPy(L?@|bZe{3uepce;DGg3)mKWI>a4_=ByLE9Ne^
z<#e{%fq}zNVkZ4Dv5@NxzBd1$?2jQCSS8P&20Ai`)i1)iNNJlX^_AXkfV)fb%64m|
zysk+)kz|EJlE}cZ^qgHE((!$Fe(Ul-kuVIg*|!lY<3Cic0tgFa84$`aha0n=;x)e9pwXrV^{<~4WDg9CqUf4WEm+)%3;}jel@Vvr54-lZC
z_2Anr^O{XurO<^p9mw
zR0fvtP?hScPGrWd8^y5M#&L7Q;K&yQiQ`I6gj=~{mIB9I@;
zp3@J%@|8#pWM?v)6TMpCG)iNiKW6q!`b}qoqYQo9P(lE{1&NqwqTe?9q;Q}bY|-lx
zQ&psPDCUDvnXtRo2XNt6pb9MB_hcmT_^ePKk`>%#{*mzgIZtZf9o%ArbG<-}DGGgH
z)q{^;57@A0(5KzrjSRkeyv7dV2PMY5x8Uxd?u>js5S%;(xm%!Z(MB^YOsX)n=}p
z3kt_p1T3ql#V?~gID!CxfKbux>a_IrT@6J^4`%0|tqOxNNS);S5PyBqjOqi=bk?aK
zGyl&`luH@*Vxb+=EjH>y^>Z}ZH76ByCKZ#Hci3Oa3@~%m2^{JH7!VzmH}pr+AIqAb
zvm*ZW&&;{(@O{b^P#)<`>0ttP1@(8s1x%O|@Kk>TPc5!|1MU~4<&&@-uJ*M#Z)14E
zDar%p&~$B>Nh-4ojGXj(`!Zvb)kshg)OOG3kNJk_-Iq;emFV{Vqn>Di%7DJ?W^vrM
zWF~hKvNwCVKveC2DDf@(>@*Gmh~uTd26*?e6QZ)%f+?nZPQT*`3jD%V4C5L~6>LEi
z=$(vSSd;K}!kDNP+vH*{3!qwE*+G7M@A4U(j0rG-9dP1t!nMAr^W+;0inv2Z+;&lx
zPJLN25qGw8(sd!Sj@HGU3e)9H@Mc9C&e_`;l*mtlj&}5NodK|QGjyZ5cs{COvXGV?
zxg|hT*cJ@V1NoNQ;a8h8VR~k%BY^cjQq6z00i{`yvMBo!5^-f@I8Zuj@QBy0Du+Fc
zlB%Gq-kEvNy((Avhlt;pJ`nl$1(W*{Y;;A3sQT7RqTzdEa2WV=7ukzN!5VGgvZRa6
z5VK1nCS7iqWs)lmg@z~|KgxKP4XF|+HV~~Ch`%tFKen?!Q_z;$4;*(OzbqQh<}BA=
z(qC5rEUT^7N73c_nBw=&t^`4c6()>u>3|mZ9A4wa#8mf7C-7izNClxNXmO7NJ{W!w
zx3Jp9#Zymz6w;P%2{N~0e(6SH6zmG;-*ZMt&to8IiE60H2R{-FI?NdV1q%~`EVR&6
zST{qY^B|H?Q$Jw)qzDeqN&ryxVC}dy?Ez9EroNNmlIL$7<$?yHu7*kd_Yi^#crKbW
z{+1g4z3VjY`FgrC&L04tE?cu7Ix&xcGFqsEC{3L}Md~U*zQ`#akX@gf>(yL>ijIjq
zsr@5cu&Tdx-`rI1S`&<>;D5@HoiB`UqyBl4|2rgFBWK~Qo5#%+7aee#owcePNKh4Y_eh3GNUpdQ-+<>e;hOZndVq}=OCi-XZxy4x@o{ZUMH^miLV6VaP_
ziJ5R;5>D|CNs>3#S$P^y9@Ma<*NnB%=K|_)!NIOh{PF@4D5*2jh%JxB+CM3t*2my5
zDl9X(LvGXgW5yekzWAg=BmlVnV!(Zkb}myo=`D$zUDn-N%PwT?5FeuGZV>RmNSy*`4^!i<@kk%KA>SPUM#>+eNx9r@SD?y`6p
ztCpCeu#r3Yf0Ai36vyqw%LNROg5$mVTN$v>+04s&BSOK^K@4|%`;KVjQ52Af5}P^6
z(87v6-1&xRfQTvjTdEB#$Gk1`#xIQbhj_`cOR3`1J0u7HWKCK*>Yv@S=~mF$F&~QV
zESWgM#O-A7Hop2BroDr!be`;{mHzx}iEdeW5w%r__}S6=Fe3Kn9k
zm|alv$Z5>rRpjOGa;J3>Viw*iQXj-`@Rgme0w4{t$cMjOUQ(KcRoRz`8S)A$*)#8&
zy3cTazIc3sMaeF)j6}AcbP=0i4!iHtGo#FZhg>Q-oh_b;14DV1%Mp)upT2@~5KT2W
zPa!hlqUtnw8QP|o%@$p?YZo=OKbW7Pi}0r+bf7PU=log$ES~?BiD^P
zkGUQxx3CRcS09lgXykIVlApuv>27O3V8yO4_B05&u8+6IJ8rScE+Gi%<2F=Zh9lwJ
z{6~B56yI6bWefkqO2xKq+eXE9Qn78@wr$(CSxLpV?NstrKkwUp-lzLJSEu`&o7y+)
zZvN)ld#$n7o^#AGbAzaMHijoZjPmvwNF;fbv5cLh^2jQ&D(qASM&+2A3tV&6ImKBo
z7Jxd7gZy!@@G|haj_F7Id}r4j6jdp@fsYtRw();{jW!lfnF@(amqLCaukga!pB9oq
zP4q`ez!=X-mv!jw-EEOKG{||Ah+3FX#
zCL4{7@``lF9pJby6ufW4zR2Vu#B2fHIMPy`Ji1Kcx8PxfTR71mIur{#_gQac)N7tn
zq-1TGeid0YWAs1;lFdNX961B7Qi~-L7ThL8L9SC>kRtA4Z(~*`A>K)W0yRA!z16Y;
z;{A#u#+6EwD)y;ffBA;C0;Cia3t^$UHvZe@&)=1)lXhG`6z0UvxJYKN`o@DzawW<&FrrUd%w9ao5cS{tbhO0%M9Hq9A-tWnDFR|<0(UL#oQ
z*8)t*c}W#0>faK7fdj0ds?;WG5;Bq?%7xM)$5|1}`2)DCnYSlR8ApCG$nbyLsopRS
z>l#wNk$^Yxg5glr`rjT*LXWfg4|{E_C%=cO)}vEK2c)vLyLIi2B@B9GNMv3U}M3nO`k528CP*E
zJ%M>XvW}=FPqeW{ttsy#{g&Fq*xAu&9Vb7{)2%%~q#F5Yc+fZU3M~Lhf+y;m9lL27
zmv|C0Nu_`?SHu}=bC24~JFNTw5yD1cK%a1%>FPPYfLfB+p|;m4X^H>6Y?it!3rj@I
z+!9lHaxo*!p-&oHlQea{b0NH+cdv2&Jya+?1pLr`6>&rQnunJAx|1=-T1V0~F$L%f
zCJmA9REPTpj7%w}^|4SF(L^wbd-ASV;Wdp#?~|yGmM1)VhT}E=nxEU<&?+|z><%Wj
zcBJTD;(=Gl<8#^Gp{rcH!eU)I`R%09o?E!0n(A(K`4e7MP{w5+t)M?=1a#f+%iV~a4+37TX^!WfNr&;-PGTDPf}&^MBY
z`+Fb`T2Fl(iO~BUsXL(bNC-9@5P=|pmB{r_-x{WCem53x;O>8mvPS;ljB8ABgv{zy
z83k|rU;-Rq)+TgD!Df%81GB1oO3>J_xJKj$>-zl67{FckB}-i6VkvNNbXtk`i>Vn2
zVlMfYVIToL5q>c6d<~hGcq6#cpu<+!*@t3o{`SgKl(;+DVzD@Sor_;J>yRp6<+GOAT^gUip$Sk0r*k*hp@cYD!bWFC%cOJ(
zir~~#nA{NU5O<8MOW5HZlBMaN!$cLhv};y0A1%M#b5u5W4i2}qa<QwDaNALM-n1=|t$EDKKl=22faB9RAvZ
zu|AC31@irG6m|B;rbfv2S8UnBVMalGR#(m5CZgpC959irLnxJc5PBryr(acEOG4LGUM=>Q4j~0GLHZ`g^ip
zOC%0vl!DvD*Dy?IU2&mrX1#(--2q~k3W`UIAKTn(^aL%@RQoMp5MnB$Y8T1WZzu#X
z6#_#Q28yvC@IslXn<4UDaeTr*J*bBp#IEl+^AN!yP`>90Jz^OaaKW2
zyixM2f@9RnKW=A7vDBBY8x3yPX7q^0d$V1g9!3$+;-4}|I`f{T%{>SGrW$1kLypOd
zUbXHeEqs>M)slprzx*I(ou03$DK7f%pFu8o8=2!Ay-X+TGZr^_mF`hvhdFI0gB{dD
z4l*mjwvW<%LZQs}CUAa4<0U}MPto*h;6*fVv0XH`nnjj@B)DcO*lFfbbad35k&J1s
z2CYmr6BJSz6C$2BU6LP;hoI@3a$Yf^4&a0*4n9Uur?R6GSX$lXBTA1D4I+ue{}_Z}
zrzz;Z2JPl-q^(eKIEE75JKoVI24$Wm0D
z;L?3vthxXcBd)e_(Xkl_T$?FW%PF-e4TNt-Buk%=#tj<5ne#ceps3eE$1ZA*f7{a-
zDEp8>1wry1D_G_Zdf_S>BLVjaJ3{pZ#X!pY
ztJhK-2M|7cnreJPybeKd)0to}(K$cNZun@@P5$)e#0Akg?OW@MoUiF^@je^wEO?xI
zd|(QWoL^dO4VP4ioi%HV#SPd#P0!5>33k{=MeOX+)%m5hUbo9OFqrq7`}20}x+-E$
z0raM1rvhz!&$Yn{_^xws0SYoc{5?JfdS-rBs7h%{R6y5YX*c?ewe3XwbEUqmEOrY4
zAp_1_tJIL&ipb@)BHjv(k%dnQ8A;fQ>Vtyx7xPLdU>jj%OQc~^Gk0`UH>&85*y-t+
zIC_e?9eY%~C;iWRX>rhR*nuQa_O?F^6jlkDkxqWv?n8IwO|PyEkxf>UH@wXUVb_|!
zbh}>UPWz~|JsjvBATut!fETn5GmYTIXdBms-CT1W4V^nEJs-h+aJ&lG1&dq#Ch}^X
zkEH7a$9E!=t_D>nk0A;zVDl^2w-Jx3J)?Q#MBX|z#^yI;8x%Uep5jS09JpDdg
z(q@H38EPjK6EX>Z@H-mn|FHB)9Nr8L5EE!BiAk>g8J&`+b~tkb`J51!pd0JrV^d2e
z?#Lo3&#Vd+IpI*%6MONpQ?bMGmAXJ7lmT+pwY05@n5GX+kgZHiDD^=ZkMUUc+X_Ld
zEU|+_=?)5dnlZ?QB}!AaIOsEfqwP}z5*@6uE^|n2$LCh
zufE5h3hWZtx)@TP28U2n3_B$pI-)OXC&+PGHE`nQMb{(Azy!HEVT}8cUNd|7!&@uc
zVeioV5V3q5!uIS^+;CY16HSG4=d~wTltfrvYy&OgQISMO-rZkH_{hyxi=$gF9Dg2H
z8;{B1`b2?Nw(EW^YsxP*Vv}PyWZ6UcAMA5*ttEB|L*--P=%}@cV^z28LtNBBMYnsKt1FZ<;#t66Tbg*xPAO$IrALkTm{_%PG0t
zyM+552>~y?+FFD6qE7bG=mZxmU+Au(Q#FA{fYixLy^h*uS^hThk@mk)Lv{LLtIu8`
zY3Bj;gABhgijHzS-@f!zawnh+Ms8y7r#65D3sBF@QLNfKf3t(39mo#1M(%LBeYK`v
ze%VXq3A+x%3CkACN4fK+Ho0O-i$v^RLSq1q!z44v%V{P$xUihca{&OzL1~eX906X0
zYVj6zRMQ0ggozF2)FLLru!0D==s{UwXND+nYcD8Ps9bg!s~1_)+Tgopimf9iW#nnK
zTOwLp&HSMU=^P_t6Iuv@y-YT}Ynntns+z%tY)l6dKq2gdXD|huLqR;A(uA*1Ntse6
z5mokS9zh4M_sGDZ4HxdfeEuZw>{s%~AzWOZ9D+h(837=NECrQ%96XlNCx*{LJ)hU$
zkBWPzF6O6YvE5D%G>LJP%l%oIw;d5R{q9?CSYOo@tOG{b?rbpg9l
z9H`G))mA#(c~+3nQN39{z_TQ=iu3DWgG%(@e$aTVIb}F}2by!jvWb$ma(+A2-2X^t
zC8uUJRHC16<6vgkN?AJEIhCLX-YlWqoT@1viF^z!9&I}*dnDIGV2!^MaOi}?42m&M
z>bnoS-0L=)Xro=X^gUHh8gm)Tyi}xTiODDBHZJY0Qh;!2i`>P4=Xv!#0yamJDX2O
zaVGLvJs|=OnV<0DB_|D1ZFZ2&W1n|b5JJs>kD^J^)G0q9*a#@8!zW>&)i%f4avk9C
zF4cpk9EO$8Hb-+7Ze)Zk)1>Q;?|v58OJN;rlvPpNuO6pM_V*)E@+~7xG4)ugs3ohV
z;xQW{{JNYtIPK3T5%T5+eWnhyGBTlEBM_3+W~~c%DXF32!D4i|V|EECk_#E)xtY0m
zmezcMwwxa9?2#2FJDniyZlwq)mwvb9QT!sMZ0vwiZFP7
z)PWnZeu>CFkBChyEr3+l@|8h^#k+}wmx(b{G-JG?A1$JR>)
z^quB08_JYY$>aR|T>%GI0TQ0}^4)ae?1Vd`x5ixAi`zkkw*#`AH9_4{`?CCCPydt
zItfECK1|BZsJ(Y%t%<>Smg^FR;gzba(jf1Ac?=P+xzQlc^e?{#vW&0`$@a1(yzr8l
zv2$5k5p>-8vgEyGx(hIw`xtn^HNrk^(^zzzid>jP)D;XyJU9C|`=n&;6QRt&+p@Sc
z=vkMke_&E@6bWU?MfEdWdf--{;s?8cLe0lHqV`|H6Xgo5m{R*5m#V|?hhYAnbU=yhE=
zNk|LC-n()3
z#%_piRoLy;CE|WkdVK|^u?*|tvY}gGqcvOe9j+P{GEM^H7jb$Fx$Bu>pGU^Kh7W{3O3s_yS
zemLRxKa(2yX1*i@j}vsHRIHWA!)~Vy`Q0_EYp9iER%#*=J=ez++!V+(`c3XCT*2Pc
z%Hi&eN&fnVYoeYU&&)v+?CbuLYUV2op@YwLEVvjxjmRfsfxnb;_p(hGmW6K0TE%5O
zQb`df3d8P!pI>pa##f^=%9?q*C;N07j}BBO;ShgobY&afR5E3;gi_Fs_=;}wRHej4
zcbQ3XQU>3ggY=u3{EPqxhX`TaQj{hu|I8WE+9~~Mccf+62&Uh+I^-j+4tD5u82hVV
z(icZF7AD_Vu)Fz=u`z^Bf-Rj7%|ifyAVVwje0c;wej{l~M^sc=cgxtyhXH`$>Z5$d
zI2>2pqRp_EZ-UM>GeM;q;HMfAD5(LLLfHacTEIH2%j|ben$&!FyBp&K-yR%}yZU13
zmGW(ys_kL1oD`QTmP~Kj^V0sL{n>V2zW|o|TE>*FrM)%Syux8}W#Cf*!k?QdjweSv
z4a=MVH7p8_vBzhxU&I@|II63ty(#|CKGjaT3q_!iHl88E@_H)uD;{vJN9UU{=-T%c
zI9?7i2r3<1=8h&lqx2tCUJ&qR=9wSF#*vm+AO}hGwYo`Mwb~$Gm?JCuE@CnPQW?n>
zW!0WhQ1eP~EtOR07l|ZeQnEtIitmnGAb<=M()RCA6$JusxEAGdCIZRK*&J7WUCyGw
z?PLzmvqUeAJe%fX{7Bj=+9wMfGO4Bxz0Ta4VG5w!t1f|Nu3e=VFAhq8$we3P-O7_6
z4=9Z~xDc$vZi~9mUG8|+xb?-IeTQYK)gS&*Zou|5dR;w~))*-}r_ID18J&dTQTta!
z+hEr4Od@QRx%l6aG@a}b0!ZV{0;0)eW)}jT+>wufsR8Y|80w-vUtjo(A|xp(Y
zi+PJmOJD~26{i-HL=+m(!?S&}Cj|}j)xyso`rc0lY4oPc^fnHB=W-r8K_n(V%CnKG
zrii;}Y36@A0Ez@6MUCs_b->c
z_A0Iu?0tarGVFBM>R)<62ejTLau)Pf&l(shZz_Hfl!B
ztw>1M{THPm=6$G0*Z6CmA_Ii6-cNz?%4Mx1vyl;BB|7rf+qCDD@
zsIwt>l0Xk07)zo&b!P)&{<_D`*ih1++-_rKjstmi>xCNO2Oh28P+BC)J7sM!b{g8a
z=q46i2+`9z&EyHle?FrNKfHPNMgDP|YDzi_78_3eAJ@
z85CkCEdD!{q}p+7GIly%uPSo>0*L_e)ZD}bC`gkZmnQGV_)f!((;zOmn{+=LvG-zt
z0T8zr=gm|7f)e|*)@xB+(PtB&78v!eb@{N{{4MLQ`ALK5%0
z^6^v+x7vI`+!I0>@I*q2+t-H^)zw1InN+H%EC0T2k?4F8$d^;zxHK98%o!MXV+W0
zY_kI4DoFE9MTDIuM50iC@J&4m*Ay>Vze2ezwYxb4=~TbdiZa#6t}wFh<3z76pl%~R
zt$j4Avq4Ml;{_IYKHrNI&IpsWn5sgzw68Zr3w4gP9t3ZecC2_B*34=^pfXQ(*_BFR
z{SI>m*Pf_{P;u`D$!befEEDvYMRS^_^(Ug$%cz#l6Um`lqg&oST8CCZC;l+1u;8mn
z$J)!sjlxm%9v%8ZN^;|S(ntsWHdggu6S*NpQwZSEzs~)OF
z#GCACSA~e)`3G|^2&RS*Gz7m@x>n(J4@FrV{yYAe3&v|(XpYBXwoKu(0(gR&Sr|#XK1zNEratJ*P)82iIB4$!LP}^pB+}W#|
zLr&RCZ#4jD`EbR}N&A=!!`{jY*u{O$A#rvUopzAaRHgbM$T2tNaZ8lvoTa~B5Mh=P
z60@Stcj*QNl%sTUkVE6^6q0Tl%mQE~ex(*BOZ~!AzlFh1EAACiC%?K{uJ`@*1H}EV
z{=oP97++^}?XaMO#BF+-d$fZ2J(P66fZtqWxBv&#TrK1gkc!PQ+d(!0h$gc_KFb%D
zoZd1ZdSb~}%-R*8Yo;L`Z&BWKPu)JaiARJ+O5d>rCVwU?8vL8U$a1T#+M_Y?3DER_
zn;5v?9Rz#Ws_KoYc%@1!4sD)@>3=o{d(4sxX~wW-DP&{#L^HfnDw&@ap6>1U>x5EV
zfL)!GIA#j(%aYJi=6iAl(%Q3v#if)(&0~7+eozf(
zQ}fy%*Okylio*gMffHX=+=jd#sJ*_TxtJDMJHTRIo)IFP{}s%eu}c*fsF#;#%saYI
zQ~7PZSiNE0_P5*(+X`a!b;a4Rba>(`;G^j3zS$lB9?R;2pq-EhQ5E2R`2|ejWlu^<&LM4(%8&a5Glij!JziN?nlEPn56~
zSJFeJ(NfM>!dbYW@iz=TNy=<=0-S2t*
zB9Ev(UGZ47VS<`2Q?r%vn2AgkS+t+~{ZvpWLglR>YR2>gM^Dt>WR~rH#O$1FtC;^0
zm^l!nQ^S0_nkKg&bKR}+^f!a~-7cmlE)-ow&GU`kEa(Pg^9q8=QARr6HP(^~=(pw!
zB@d>dqNpO47FwXw6f@~MASZwIqB-HJS<@3Kde)Q&F_~XDlL3_pyaW%yJ_p%K2tDBF
zCyv_04){Y~k2yhky0;s7iS1zR;JJk03>MGv!26fTfHs=&;$0V?vB$D)cZKtfwg|ye
zO*p+?AG~Mc3*kUN1y8rTC2YKq-wM71N(S@7KQ7P7I?-W8FPcqQ3r75AazxB4A58u11#QP
z@iyp7E=eKg;E+!;f)_MmcTTzEnQUV7f5JyQaEzD1k6a6rMlFt@xtDv2(bVFa$qXJ$
zO<4_m6GuobxSQF+Ll$^mm{=OmdVvb8wJABYL|@`&qsU9p3*`uno6OF`x3vx#FbVMg
zRlf&Ec_CixgV~F2UDrPhyrZ6`(_+L>{W%s>{;Pu};#CU@J#%7EcL$i^cw!tALa78H
z79@Zld_CKe&SSL<=Aq6;-VB#(+=({j9eV9Daf4I^1zQhgoMs6pcR|
zpCoX62=9kbAWf)|erC(JmQY1Nrf7+fGg-A@eR|UOJRj1T-Vp?D>b4Jh8ZC+0WtXAx
z&d~wk7ZL&2-452zZfd8O#q)`W+!?W2h;$gWBy;@9Df%0h?K?W`WUq}HfO+g8+DVz@
zupGd}tHlx)LVs}zj*15Ockp@Ik)B|itz|jM%NqzReOGoXn4v+;JVFNYkl0th>5ZZ>
z4Ef0(Ztt9gJhtnnN)XZ{SWgsY7mDtploR%1x*)XC3nVT4K^kb8thciiu0dc2q1!Jq
z4%r(AQe0x+koe+9_r0AhAqM&BZ9m~>ZR;i+_gPkaC{RFI(JE>zmRWU&VGX{l)g`d=
zNYMvK4!E-5O)>HC_3G%}NY{Dhwfc8EuOp*7%6x*Fv8$5mUn1R75y}OU;CI89kki%kFw>O{dz4p{m2g)@Gsn
zl>a({U%+6`-i*+?uRu1E7n{F}B8IlaGw6mKw6Wf9-)kU2E>j;(b!Je{+&tvtd9b(W
zAvsUY9m%#CIX1&*VO%l&An#v07#A5=QFkeVwMQB+7ByrN*kaS8G+$+OD*nwC25)nY
zELp|r_YMBLq?};
zS3o1-JQ*7<%tXGn1V_^_MBc5~we3%abQ_{b#zZsryJ^v8>cz
zq?y+5|N7Pcvxwxc7g4B0fh!TdGFyFdd976Wt?&VVub%67W$>f
zdpnbeXb}
zPGSG%Mh+waMKsf+gHb?EbMu4y?x@c=*mo@a3^*-Rh0y-uNY_R>XE|7<>Re5JP`<)R
zQ>=w-?~d93${zkd64JB>Ap9+j`OmlS!t+-QOfF5LjuRK*t>0;ha9TFZ?sB53}%7=tX-T{x*h3YTP(4unX<%PD&FSIN+8cnp_
z6sL;hXJniYMtW#s(yNB)1}&b`P$lIjCSR5+JCWkJnZAt+FVfz$0;tK+!BG^H
zc#vt^{TDiA(p<7q5J2IW9?9P>{qSqcsCuLXX|;?dOssp7x>;9>E9s>#c*LDqMl$tK
z2Z22N-PpGIhQy$hRC0b^Eg4GMyQNq*b42V^6pf5?RATU8803s>`rYFjFDHpT|eEwQ+=OY_|7>yGd%8H#U-v&+JN9lK0*0w8c5}
zvB>DgSGyC>Z276UPr`*LEt7s?k@lzNgIYvKCZ|XhopMf0AVP{89|?#C>#0uKLZc^4
zyGtQsmTLLRYrzQV(5G~Le7Q$RYBn_Xo;^j7w?b|!co;ivQho#uxdxOCbbP=kpSN;4
z8QJc)2RM!L1+1|!*67PR{#8peRy;TFV|7-vPs3f$7ZmPM{ZmJzsCtkzUt7UaP>Zdw
z4ukni#AxJr`AgzT2mlexUy`%`I)k3*4W->dun
z%l6@?`Hs5g-W389#iiF5E;*MdPYc6i7Mi?d1F}45O_=FR&?SLl$>|-Be1sMpK{kqX
zHL*~0-J-(Q*j;>V3x3zAT~<(3xMj4k3+=A0VMx*b@r6pZfMGG}DoZ~Q$Jeum6)PHy
z&kgRONs@UhLgYr8-85ASsnjfE{>S)7*BW#sVjA3q9aR9(EUIaPEFBC{^3!va(G#4h
zip~d4$KJTSL!H#w?BS~X=#4p9+&oQ;i_Zo=7-JPJWOd#T$jb8F_nG7+DM#_-i~4oR
z#e@FS2;l1+^2?8lBJyUBUQiW&05*N|OxrqAl*Z!wUPesp@#fot)qoUBM5Mrt1I396
ztk1RCqZ|@y0H<-N(Uq=-mC8gjM1ch|R{*ER`vM}fKV|6R4>q&8i$_{18Z(5n2nwDB
zSH~_;;pTAEFb{%t#+t7#SLr4&=)2_y`I*!iza_SuvwX$TQih_{6ZM1mfzba!md~la
zXHzcm`oBgjO~NqdmY6$F*HpR^sTu$jBNCyt9hFsa$P<~C|q{e0)1eBL~o?aOdaUMi8N
zv@`RglllOm+UZ#r-cLwMP*(esQ6lH#?fb3fi
z*f|>|KNts)toP8{qY!)z$Ua&t@yP(1Yskf9?ym%O=IPzNB<>9Bf#|O(LC{ITA&d$@
z@zMOn&Uu;4Kr8rIM@YpPO3jM19+qeL`VuvXWs1+k6CuMjYNoW5eX3`>Kr9#lx=TNg^u#dqC;F
zn-h@j+UHj#d{w1CT)L)#2mLOZ%2jfa*!uECigfT9s71X|e_SbkF@YO<+~
z*3WQjKywb~nh8dLeu8o?x9jjo<4m)J$(Wbt#){}gn385qw+u}-D)33Uydt#3rh$Np
z-jex9=V^B1u16l$p=uWc-SZFZ?A-aN>A3@s2=<#DZhK(p@e^<&k!maN+!ao@m*0@#
zcJ&IG^R?!_M&kL(6ZPqU?d$n=b&`nKzdrVG`+Msu|M;!a412$|)UX!-q=FRUGb%)I
z|9aew!!l1#mfY6ocKkF7fH3(V3!MW94UL8VOBLZi?{EA7VB&u$_y7IBp~gATKYx$V
zO){r!0do?|8#UPIV=-5M9G`;f#thX71
zW4PH7UqN|PVe*R0Q7Q8@3uSRN+?QB~bDOC_QY<`kv+-e>3m~=Qyc5<1h4x!YCw#_o
zo?6m+(