From 8bf696948f208cbfef56624b7eacf95641edfa1a Mon Sep 17 00:00:00 2001
From: Josh Suereth <joshuasuereth@google.com>
Date: Mon, 8 Mar 2021 09:32:57 -0500
Subject: [PATCH 01/14] Initial cut at migrating Josh MacD's datamodel document
 into the specification.

---
 specification/metrics/datamodel.md | 184 +++++++++++++++++++++++++++++
 specification/overview.md          |  10 +-
 2 files changed, 192 insertions(+), 2 deletions(-)
 create mode 100644 specification/metrics/datamodel.md

diff --git a/specification/metrics/datamodel.md b/specification/metrics/datamodel.md
new file mode 100644
index 00000000000..2f9f1255fd2
--- /dev/null
+++ b/specification/metrics/datamodel.md
@@ -0,0 +1,184 @@
+# Introduction
+
+TODO(jsuereth): Write this
+
+## Events → Data → Timeseries
+
+The OTLP Metrics protocol is designed as a standard for transporting metric
+data. To describe the intended use of this data and the associated semantic
+meaning, OpenTelemetry metric data types will be linked into a framework
+containing a higher-level model, about Metrics APIs and discrete input values,
+and a lower-level model, defining the Timeseries and discrete output values.
+The relationship between models is displayed in the diagram below.
+
+TODO - Diagram
+
+This protocol was designed to meet the requirements of the OpenCensus Metrics
+system, particularly to meet its concept of Metrics Views. Views are
+accomplished in the OpenTelemetry Metrics data model through support for data
+transformation on the collection path.
+
+OpenTelemetry has identified three kinds of semantics-preserving Metric data
+transformation that are useful in building metrics collection systems as ways of
+controlling cost, reliability, and resource allocation. The OpenTelemetry
+Metrics data model is designed to support these transformations both inside an
+SDK as the data originates, or as a reprocessing stage inside the OpenTelemetry
+collector. These transformations are:
+
+1. Temporal reaggregation: Metrics that are collected at a high-frequency can be
+   re-aggregated into longer intervals, allowing low-resolution timeseries to be
+   pre-calculated or used in place of the original metric data.
+2. Spatial reaggregation: Metrics that are produced with unwanted dimensions can
+   be re-aggregated into metrics having fewer dimensions.
+3. Delta-to-Cumulative: Metrics that are input and output with Delta temporality
+   unburden the client from keeping high-cardinality state. The use of deltas
+   allows downstream services to bear the cost of conversion into cumulative
+   timeseries, or to forego the cost and calculate rates directly.
+
+OpenTelemetry Metrics data points are designed so that these transformations can
+be applied automatically to points of the same type, subject to conditions
+outlined below. Every OTLP data point has an intrinsic
+[decomposable aggregate function](https://en.wikipedia.org/wiki/Aggregate_function#Decomposable_aggregate_functions)
+making it semantically well-defined to merge data points across both temporal
+and spatial dimensions. Every OTLP data point also has two meaningful timestamps
+which, combined with intrinsic aggregation, make it possible to carry out the
+standard metric data transformations for each of the model’s basic points while
+ensuring that the result carries the intended meaning.
+
+As in OpenCensus Metrics, metrics data can be transformed into one or more
+Views, just by selecting the aggregation interval and the desired dimensions.
+One stream of OTLP data can be transformed into multiple timeseries outputs by
+configuring different Views, and the required Views processing may be applied
+inside the SDK or by an external collector.
+
+### Example Use-cases
+
+TODO(jsuereth): Flesh this out
+
+1. OTel SDK exports 10 second resolution to a single OTel collector, using
+  cumulative temporality for a stateful client, stateless server:
+  - Collector passes-through original data to an OTLP destination
+  - Collector re-aggregates into longer intervals without changing dimensions
+  - Collector re-aggregates into several distinct views, each with a subset of
+    the available dimensions, outputs to the same destination
+2. OTel SDK exports 10 second resolution to a single OTel collector, using delta
+  temporality for a stateless client, stateful server:
+  - Collector re-aggregates into 60 second resolution
+  - Collector converts delta to cumulative temporality
+3. A number of OTel SDKs running locally each exports 10 second resolution, each
+  reports to a single (local) OTel collector.
+  - Collector re-aggregates into 60 second resolution
+  - Collector re-aggregates to eliminate the identity of individual SDKs (e.g.,
+    distinct `service.instance.id` values)
+  - Collector outputs to an OTLP destination
+4. Pool of OTel collectors receive OTLP and export Prometheus Remote Write
+  - Collector joins service discovery with metric resources
+  - Collector computes “up”, staleness marker
+  - Collector applies a distinct external label
+5. OTel collector receives Statsd and exports OTLP
+  - With delta temporality: stateless collector
+  - With cumulative temporality: stateful collector
+
+These are considered the "core" use-cases used to analyze tradeoffs and design
+decisions within the metrics data model.
+
+# Model Details
+
+TODO(jsuereth): Recap split of 3 pieces
+
+## Event Model
+
+This specification uses as its foundation a
+[Metrics API consisting of 6 model instruments](api.md), each having distinct
+semantics, that were prototyped in several OpenTelemetry SDKs between July 2019
+and June 2020. The model instruments and their specific use-cases are meant to
+anchor our understanding of the OpenTelemetry data model and are divided into
+three categories:
+
+- Synchronous vs. Asynchronous. The act of calling a Metrics API in a
+  synchronous context means the application calls the SDK, typically having
+  associated trace context and baggage; an Asynchronous instrument is called at
+  collection time, through a callback, and lacks context.
+- Adding vs. Grouping. Whereas adding instruments express a sum, grouping
+  instruments characterize a group of measurements. The numbers passed to adding
+  instruments define division, in the algebraic sense, while the numbers passed
+  to grouping instruments are generally not. Adding instrument values are always
+  parts of a sum, while grouping instrument values are individual measurements.
+- Monotonic vs. Non-Monotonic. The adding instruments are categorized by whether
+  the  derivative of the quantity they express is non-negative. Monotonic
+  instruments are primarily useful for monitoring a rate value, whereas
+  non-monotonic instruments are primarily useful for monitoring a total value.
+
+In this model, the primary data are (instrument, number) points, originally
+observed in real time or on demand (for the synchronous and asynchronous cases,
+respectively). The instruments and model use-cases will be described in greater
+detail as we link this model with the other two.
+
+
+## Timeseries Model
+
+In this low-level metrics data model, a Timeseries is defined by an entity
+consisting of several metadata properties:
+
+- Metric name and description
+- Label set
+- Kind of point
+- Unit of measurement
+
+The primary data of each timeseries are ordered (timestamp, value) points, for
+three value types:
+
+1. Counter (Monotonic, cumulative)
+2. Gauge
+3. Histogram
+
+This model may be viewed as an idealization of
+[Prometheus Remote Write](https://docs.google.com/document/d/1LPhVRSFkGNSuU1fBd81ulhsCPR4hkSZyyBj1SZ8fWOM/edit#heading=h.3p42p5s8n0ui).
+Like that protocol, we are additionally concerned with knowing when a point
+value is defined, as compared with being implicitly or explicitly absent. There
+is no delta temporality in the timeseries model. To precisely define presence
+and absence of data requires further development of the correspondence between
+these models.
+
+## OpenTelemetry data model
+
+The OpenTelemetry data model for metrics includes four basic point kinds, all of
+which satisfy the requirements above, meaning they define a decomposable
+aggregate function (also known as a “natural merge” function) for points of the
+same kind. The basic point kinds are:
+
+1. Monotonic Sum
+2. Non-Monotonic Sum
+3. Gauge
+4. Histogram
+
+Comparing the OpenTelemetry and Timeseries data models, OTLP carries an
+additional kind of point. Whereas an OTLP Monotonic Sum point translates into a
+Timeseries Counter point, and an OTLP Histogram point translates into a
+Timeseries Histogram point, there are two OTLP data points that become Gauges
+in the Timeseries model: the OTLP Non-Monotonic Sum point and OTLP Gauge point. 
+
+The two points that become Gauges in the Timeseries model are distinguished by
+their built in aggregate function, meaning they define re-aggregation
+differently. Sum points combine using addition, while Gauge points combine into
+histograms.
+
+# Single-Writer
+
+Pending
+
+# Temporarily
+
+Pending
+
+# Resources
+
+Pending
+
+# Temporal Alignment
+
+Pending
+
+# External Labels
+
+Pending
\ No newline at end of file
diff --git a/specification/overview.md b/specification/overview.md
index a542c2b04b9..01c91193172 100644
--- a/specification/overview.md
+++ b/specification/overview.md
@@ -264,9 +264,12 @@ supports both - push and pull model of setting the `Metric` value.
 
 ### Metrics data model and SDK
 
-Metrics data model is defined in SDK and is based on
+Metrics data model is [specified here](metrics/datamodel.md) and is based on
 [metrics.proto](https://github.com/open-telemetry/opentelemetry-proto/blob/master/opentelemetry/proto/metrics/v1/metrics.proto).
-This data model is used by all the OpenTelemetry exporters as an input.
+This data model defines three semantics: An Event model used by the API, an
+in-flight data model used by the SDK and OTLP, and a TimeSeries model which
+denotes how exporters should interpret the in-flight model.
+
 Different exporters have different capabilities (e.g. which data types are
 supported) and different constraints (e.g. which characters are allowed in label
 keys). Metrics is intended to be a superset of what's possible, not a lowest
@@ -279,6 +282,9 @@ validation and sanitization of the Metrics data. Instead, pass the data to the
 backend, rely on the backend to perform validation, and pass back any errors
 from the backend.
 
+See [Metrics Data Model Specification](metrics/datamodel.md) for more
+information.
+
 ## Log Signal
 
 ### Data model

From d38c7b94d308de875dd68eaede8bec94610cd055 Mon Sep 17 00:00:00 2001
From: Josh Suereth <joshuasuereth@google.com>
Date: Mon, 8 Mar 2021 09:38:24 -0500
Subject: [PATCH 02/14] Fix style nits.

---
 specification/metrics/datamodel.md | 59 +++++++++++++++---------------
 1 file changed, 29 insertions(+), 30 deletions(-)

diff --git a/specification/metrics/datamodel.md b/specification/metrics/datamodel.md
index 2f9f1255fd2..67f1a708b6f 100644
--- a/specification/metrics/datamodel.md
+++ b/specification/metrics/datamodel.md
@@ -1,6 +1,6 @@
-# Introduction
+# Metrics Data Model
 
-TODO(jsuereth): Write this
+TODO(jsuereth): Write introduction
 
 ## Events → Data → Timeseries
 
@@ -11,7 +11,7 @@ containing a higher-level model, about Metrics APIs and discrete input values,
 and a lower-level model, defining the Timeseries and discrete output values.
 The relationship between models is displayed in the diagram below.
 
-TODO - Diagram
+TODO(jsuereth) - Diagram
 
 This protocol was designed to meet the requirements of the OpenCensus Metrics
 system, particularly to meet its concept of Metrics Views. Views are
@@ -57,36 +57,36 @@ TODO(jsuereth): Flesh this out
 
 1. OTel SDK exports 10 second resolution to a single OTel collector, using
   cumulative temporality for a stateful client, stateless server:
-  - Collector passes-through original data to an OTLP destination
-  - Collector re-aggregates into longer intervals without changing dimensions
-  - Collector re-aggregates into several distinct views, each with a subset of
-    the available dimensions, outputs to the same destination
+    - Collector passes-through original data to an OTLP destination
+    - Collector re-aggregates into longer intervals without changing dimensions
+    - Collector re-aggregates into several distinct views, each with a subset of
+      the available dimensions, outputs to the same destination
 2. OTel SDK exports 10 second resolution to a single OTel collector, using delta
   temporality for a stateless client, stateful server:
-  - Collector re-aggregates into 60 second resolution
-  - Collector converts delta to cumulative temporality
+    - Collector re-aggregates into 60 second resolution
+    - Collector converts delta to cumulative temporality
 3. A number of OTel SDKs running locally each exports 10 second resolution, each
   reports to a single (local) OTel collector.
-  - Collector re-aggregates into 60 second resolution
-  - Collector re-aggregates to eliminate the identity of individual SDKs (e.g.,
-    distinct `service.instance.id` values)
-  - Collector outputs to an OTLP destination
+    - Collector re-aggregates into 60 second resolution
+    - Collector re-aggregates to eliminate the identity of individual SDKs (e.g.,
+      distinct `service.instance.id` values)
+    - Collector outputs to an OTLP destination
 4. Pool of OTel collectors receive OTLP and export Prometheus Remote Write
-  - Collector joins service discovery with metric resources
-  - Collector computes “up”, staleness marker
-  - Collector applies a distinct external label
+    - Collector joins service discovery with metric resources
+    - Collector computes “up”, staleness marker
+    - Collector applies a distinct external label
 5. OTel collector receives Statsd and exports OTLP
-  - With delta temporality: stateless collector
-  - With cumulative temporality: stateful collector
+    - With delta temporality: stateless collector
+    - With cumulative temporality: stateful collector
 
 These are considered the "core" use-cases used to analyze tradeoffs and design
 decisions within the metrics data model.
 
-# Model Details
+## Model Details
 
 TODO(jsuereth): Recap split of 3 pieces
 
-## Event Model
+### Event Model
 
 This specification uses as its foundation a
 [Metrics API consisting of 6 model instruments](api.md), each having distinct
@@ -114,8 +114,7 @@ observed in real time or on demand (for the synchronous and asynchronous cases,
 respectively). The instruments and model use-cases will be described in greater
 detail as we link this model with the other two.
 
-
-## Timeseries Model
+### Timeseries Model
 
 In this low-level metrics data model, a Timeseries is defined by an entity
 consisting of several metadata properties:
@@ -140,7 +139,7 @@ is no delta temporality in the timeseries model. To precisely define presence
 and absence of data requires further development of the correspondence between
 these models.
 
-## OpenTelemetry data model
+### OpenTelemetry data model
 
 The OpenTelemetry data model for metrics includes four basic point kinds, all of
 which satisfy the requirements above, meaning they define a decomposable
@@ -156,29 +155,29 @@ Comparing the OpenTelemetry and Timeseries data models, OTLP carries an
 additional kind of point. Whereas an OTLP Monotonic Sum point translates into a
 Timeseries Counter point, and an OTLP Histogram point translates into a
 Timeseries Histogram point, there are two OTLP data points that become Gauges
-in the Timeseries model: the OTLP Non-Monotonic Sum point and OTLP Gauge point. 
+in the Timeseries model: the OTLP Non-Monotonic Sum point and OTLP Gauge point.
 
 The two points that become Gauges in the Timeseries model are distinguished by
 their built in aggregate function, meaning they define re-aggregation
 differently. Sum points combine using addition, while Gauge points combine into
 histograms.
 
-# Single-Writer
+## Single-Writer
 
 Pending
 
-# Temporarily
+## Temporarily
 
 Pending
 
-# Resources
+## Resources
 
 Pending
 
-# Temporal Alignment
+## Temporal Alignment
 
 Pending
 
-# External Labels
+## External Labels
 
-Pending
\ No newline at end of file
+Pending

From 12f973405cdcb2f1bdd05022297aafdc1d2a3088 Mon Sep 17 00:00:00 2001
From: Josh Suereth <joshuasuereth@google.com>
Date: Mon, 8 Mar 2021 09:58:15 -0500
Subject: [PATCH 03/14] Fix footnote import.

---
 specification/metrics/datamodel.md | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/specification/metrics/datamodel.md b/specification/metrics/datamodel.md
index 67f1a708b6f..71f82d0372e 100644
--- a/specification/metrics/datamodel.md
+++ b/specification/metrics/datamodel.md
@@ -144,7 +144,9 @@ these models.
 The OpenTelemetry data model for metrics includes four basic point kinds, all of
 which satisfy the requirements above, meaning they define a decomposable
 aggregate function (also known as a “natural merge” function) for points of the
-same kind. The basic point kinds are:
+same kind. [^otlppointkindcaveat]
+
+The basic point kinds are:
 
 1. Monotonic Sum
 2. Non-Monotonic Sum
@@ -181,3 +183,5 @@ Pending
 ## External Labels
 
 Pending
+
+[^otlppointkindcaveat]: OTLP supports data point kinds that do not satisfy these conditions; they are well-defined but do not support standard metric data transformations.

From d73341ba4e708e70466c0e11dce2fb96e568de3d Mon Sep 17 00:00:00 2001
From: Josh Suereth <joshuasuereth@google.com>
Date: Mon, 8 Mar 2021 12:18:41 -0500
Subject: [PATCH 04/14] Fix footnotes.

---
 specification/metrics/datamodel.md | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/specification/metrics/datamodel.md b/specification/metrics/datamodel.md
index 71f82d0372e..4d4c419d4f0 100644
--- a/specification/metrics/datamodel.md
+++ b/specification/metrics/datamodel.md
@@ -144,7 +144,7 @@ these models.
 The OpenTelemetry data model for metrics includes four basic point kinds, all of
 which satisfy the requirements above, meaning they define a decomposable
 aggregate function (also known as a “natural merge” function) for points of the
-same kind. [^otlppointkindcaveat]
+same kind. <sup>[1](#otlpdatapointfn)</sup>
 
 The basic point kinds are:
 
@@ -184,4 +184,8 @@ Pending
 
 Pending
 
-[^otlppointkindcaveat]: OTLP supports data point kinds that do not satisfy these conditions; they are well-defined but do not support standard metric data transformations.
+## Footnotes
+
+<a name="otlpdatapointfn">1</a>: OTLP supports data point kinds that do not
+satisfy these conditions; they are well-defined but do not support standard
+metric data transformations.

From c0e132629e307b112d5927632e14c12c9c473985 Mon Sep 17 00:00:00 2001
From: Josh Suereth <joshuasuereth@google.com>
Date: Mon, 8 Mar 2021 12:52:49 -0500
Subject: [PATCH 05/14] Fixes from merge w/ @jmacd

---
 specification/metrics/datamodel.md | 41 +++++++++++++++++++++++++++---
 1 file changed, 38 insertions(+), 3 deletions(-)

diff --git a/specification/metrics/datamodel.md b/specification/metrics/datamodel.md
index 4d4c419d4f0..a4f97079347 100644
--- a/specification/metrics/datamodel.md
+++ b/specification/metrics/datamodel.md
@@ -1,6 +1,38 @@
 # Metrics Data Model
 
-TODO(jsuereth): Write introduction
+**Status**: [Experimental](../document-status.md)
+
+<!-- Re-generate TOC with `markdown-toc --no-first-h1 -i` -->
+
+<!-- toc -->
+
+<!-- tocstop -->
+
+## Overview
+
+The OpenTelemetry data model for metrics consists of a protocol specification
+and semantic conventions for delivery of pre-aggregated metric timeseries data.
+The data model is designed for importing data from existing systems and
+exporting data into existing systems, as well as to support internal
+OpenTelemetry use-cases for generating Metrics from streams of Spans or Logs.
+
+Popular existing metrics data formats can be unambiguously translated into the
+OpenTelemetry data model for metrics, without loss of semantics or fidelity.
+Translation from the Prometheus and Statsd exposition formats is explicitly
+specified.
+
+The data model specifies a number of semantics-preserving data transformations
+for use on the collection path, supporting flexible system configuration. The
+model supports reliability and statelessness controls, through the choice of
+cumulative and delta transport. The model supports cost controls, through
+spatial and temporal reaggregation.
+
+The OpenTelemetry collector is designed to accept metrics data in a number of
+formats, transport data using the OpenTelemetry data model, and then export into
+existing systems. The data model can be unambiguously translated into the
+Prometheus Remote Write protocol without loss of features or semantics, through
+well-defined translations of the data, including the ability to automatically
+remove attributes and lower histogram resolution.
 
 ## Events → Data → Timeseries
 
@@ -53,7 +85,9 @@ inside the SDK or by an external collector.
 
 ### Example Use-cases
 
-TODO(jsuereth): Flesh this out
+The metric data model is designed around a series of "core" use cases.  While
+this list is not exhaustive, it is meant to be representative of the scope and
+breadth of OTel metrics usage.
 
 1. OTel SDK exports 10 second resolution to a single OTel collector, using
   cumulative temporality for a stateful client, stateless server:
@@ -78,6 +112,7 @@ TODO(jsuereth): Flesh this out
 5. OTel collector receives Statsd and exports OTLP
     - With delta temporality: stateless collector
     - With cumulative temporality: stateful collector
+6. OTel SDK exports directly to 3P backend
 
 These are considered the "core" use-cases used to analyze tradeoffs and design
 decisions within the metrics data model.
@@ -186,6 +221,6 @@ Pending
 
 ## Footnotes
 
-<a name="otlpdatapointfn">1</a>: OTLP supports data point kinds that do not
+<a name="otlpdatapointfn">[1]</a>: OTLP supports data point kinds that do not
 satisfy these conditions; they are well-defined but do not support standard
 metric data transformations.

From f3587726a3279922c9445dd545e9806f53abd8f5 Mon Sep 17 00:00:00 2001
From: Josh Suereth <joshuasuereth@google.com>
Date: Mon, 8 Mar 2021 13:28:20 -0500
Subject: [PATCH 06/14] Add diagram.

---
 specification/metrics/datamodel.md         |   2 +-
 specification/metrics/img/model-layers.png | Bin 0 -> 24407 bytes
 2 files changed, 1 insertion(+), 1 deletion(-)
 create mode 100644 specification/metrics/img/model-layers.png

diff --git a/specification/metrics/datamodel.md b/specification/metrics/datamodel.md
index a4f97079347..70de49dd800 100644
--- a/specification/metrics/datamodel.md
+++ b/specification/metrics/datamodel.md
@@ -43,7 +43,7 @@ containing a higher-level model, about Metrics APIs and discrete input values,
 and a lower-level model, defining the Timeseries and discrete output values.
 The relationship between models is displayed in the diagram below.
 
-TODO(jsuereth) - Diagram
+![Events  → Data → Timeseries Diagram](img/model-layers.png)
 
 This protocol was designed to meet the requirements of the OpenCensus Metrics
 system, particularly to meet its concept of Metrics Views. Views are
diff --git a/specification/metrics/img/model-layers.png b/specification/metrics/img/model-layers.png
new file mode 100644
index 0000000000000000000000000000000000000000..1ca9ea0b30a8edd5df752ae37f130694e43e0a2c
GIT binary patch
literal 24407
zcmeFZbyU>R`z|V7A}FP#3Jjr?bPXlo2t$Y>-CaX>4T2~V$|&6+CEc9@BGS#!0z(cx
zG~CblJ->6#{rBE`*1hYl^AF2~pZUz*?|%2*&-*;j%SR10WpYw_(p$G~kwacQe|75?
zF7R)h6JkQ(H#6*B6t`{}Gee%sz4kWRP9;cSQk;1IBs)<Rl=!hx?nQ_s&OPGxxe$CZ
z-u6^QJbbT7%7=B0jVi&56(N#+=!(lvXv_laq&P9HP?8p%V->p_khYtMG#m$e3~W@0
zC=ovjrIi1F;m3#im|Zi)ST=~`^C`AR&lw*3-Xx2%Lt<w_%OE9R?#YLCJ;?|DH!~FK
z8ugX(@lzwYZ)^}!c86H|oKQ;BebPcU$SY1$gaIpiDDkJa8o(#zNs?FSxgI^&X$s+n
z#8SVUVw-$H@~C5(<~1ZXN23E&uk@1mE^TSX%UHIsdm_EUujD_Jc@w;eW#jt#zrXT-
zd0<}^DjT)PB)MGB3`?ueF*jd~Zy7bPLBi<%`$~8iO%%l#MSxh!^v8<*qJ(um7TOnP
z8=g#a`R;VmH9OLe$bX8CSYr`LuWjKn4D3|w6;N+or{*ixXgJxB+%z;+m^m40ZgNyu
zA0tzXoIiEXtp)!O<OP4nG*L*o%J9u``|WU~xi!Uu8zsy(Gkd|XfOSH?Vy@$z1r9yW
zu<y;shgq(JO2otO_eeLJ3h5StBwWz52=+K#DT&RT&g{e7bQ-3BgQeknDtvK8=`gtI
za3xF{?U=0oqo~}qYWa7c-&XQSq-3u06%pnWZ_~K}FtmyR*S)g#VUWJlR0gkImzzxq
zh^0>7wFF9FJ1q-Gt>;dncZ1!7Uclv%TIgxwM>CX0a;a~HosPfVWA&ISC-Sph|3Gy|
zWhCCv<pE!Db1xxjvQu?={KQ06zZ#W3HRn@)JMV*&OuF6UgNGs6fq3WK8aCO>%oBPX
zkDe#DJ;^`JDrrS+{**0l+@CSh)6PM}tl`8hoE;pmuN>wook+V`yCXYAmHb$w#3gQi
zS0aw-`o;Jf77N{)0t{RZzYhln%}5q72QAh&X6G3N2~GF7?Mfx)AT7^dlwEEv7`=#N
z2|8|HpRP^+5G}Q@-s|G*T5{-rIUXL{1Qq`gM@VXI?CJYXm~l)Z#jh*zc%uB~U5t1J
zM#}nn^~`r;?6-5;VC%(e*MV?VasJ*<l|fsTZ|x5aedof456t6L29BpjqoYfjj$L<8
zwM8_hPRg3I?l4I_Ct@rwjSsheo|bwt`v{h?H$YbzyR~G15w{q$nTeuP-VL~VU(52m
zfABhdmx{@K?7Fwoj#FdtEY<azq!--lH%V3Nd+{nmg9B|^>~T-Niliw-g3oTm2?-+H
z+(hR3*y<Y6NOB)&@(dnqrT<Zl6Gd8$=9~8%tcV7=L^`n$oXZSR%cY^E8x*B|J!4+J
zRIz(U5khlNA3*IbXyp?wUGu5Iy*T1`Umv2iDEs4MBz1_O^_3c;_GCaB8<Q<{9Z_yx
zt`c&qBIhFmyh+&Pj+oD@<@_HF?$Lx4Ow`}vzJUV3+sBV2KDu7cVdli^jZ?WL6~CoE
zL3m`-GAGVHIpAp-%bIXu%M$OA7Jb6Nx##tToi>`*;3q6_vE}ytF&W!ek6|Jj-lyL$
zvP?U}L0f*u8{ec@0^RQ5+lVYAr#HXSzxe5KygttA12XR}{#?{YNGduj$vUmBsj2D8
zx4tExw&S%>GVjBsk*|f;CHqfW(pCRCP+=#GsFe}rc_J~kgs;pi^jNnBjvH+f-|Q8M
z_TFAwuh<%$Ar#dpYyb6CE2hZMYk9hd9a+&*APe_DN;bhKT#kxA7SnHRy-06<;^X#x
zqC@(m$RVTbXq_D)dhy=k@DQXPm}FdWk?G0mhi*EqY?|ZMpPaa@p%8M|Q_ymBE!Zo;
zz`1BL*l<*3!F0x<HZ$9(%;V_0$R|vFIUpjVAt;N$JNcx?GUAvaOFp!B_mHFg_mMvD
zkGc15td#W9p7k>xm=-NJQ+tMW2ix>)a_aTNSEI!s+fDEFFuzz2DbX~FvuMeygqZ+t
zjeM$;ZeT}J13Qu~46)chX_&5_`a?oF;QUxE5Gjjey|@O%4gY?V@F8e(O+RSVV^bXY
z9yhG~Z_gMqT2wz}%FaIK1T+0LNpKUV$NKOuWkIu!d#~c=PWbAh!*N4NSpO`Umlzjp
zV*7CKF9jsdG2?UbM_HlLvb4pvlH^ob_}cB-v!#Z{(w8Fz=U#XOA(C;2QE^kAsNz_m
zBp#hLaHNhi4&Q<VL-DtjN;qzSp-!6TD7%pnZhz(>MS-DqhDP4(O*&;3`4Y{TkLD+y
zdYjh<1}oxq{@hQVs4zZmb<8N62W@$uRlqy?DGcIQB;9|@8MPn=Qj>CoX*e~^mz-4I
z)G){jDT`?6*p3T)93A)}nLCN>ZXuC5^*{Giy&;V&7R-GRVyX0@PMAt-FeO&Z$uHdj
zh&%*ye>#`3(a95+AWl5S;jnW;QcV?CKClrU(p-aNuJ1*bl0VL0%}5Min3nC{46TfY
zjsR(}=KCWLnd<q~OU8kGqAw_rxA%u^rrInWYp9KuH-U!{^_@sR(^M0MET&}tC#51H
zx;4>J+-{;ks+SGO=+8M=rfO@!aR>g)#TD+z+~X9gjlo@ZOxJraO)(>_YnAY*od!zw
zbacBNydl%B?*(?OIwOP_0r_N5*^{vrb{s}b?SPU;vp~s?Xfo=9g@8Qnmyc;a$f#XN
zX08)~WY@IMi<ZpwTVI4v$(q&+^np%#nu&BZ59N#qNIjOB>47u11AFEU)L%yalob|k
zr-TizRKhx{;T;%#Zy#Wd23V(q1~e@d_hj|-DX1w~Df6M3Tp?=kX>%%mA$LB5{*SWo
z!$)OHiNif1^yEnEc5-Trgm;X^8Akbvo<Payv*#Z!<*^A)4auc7$*T?oLWqGg(%Al6
zj_vx+pz+~B!@pxK^R9dq7xqviCF)`@Qp}53M{~dqNAAM@9I2c9?06Cc4Zjc;)yT+Z
z?#DmM@H&h#b~C7m!BBEKt2u;^u9WV4u<5da%PA0oV0+WY26)j3wV<biA7ybgy9V8+
zTt}>WUl2FWlT-87V?rzv)jp#GA7skTke?VLp1K7HnbviJ9F<iWI4RvL93*S46f*=&
zx1e^(4b+@tAN{OCDIG~mGX{QSRWc|XH@x<O%zWD+CT66*L!S4CQZ6N8F1H?zYj#`O
z$Ho8X1IN^_fqOb1WHY~KGt=*6Q}%f-RxbP6>OC}B2G@^rBkAcFGntNMC8ZFr_myB5
z(p794Wp2#;?#HVXX~pp}Km6gim3A|kNz)&zW@X`LRO<%6gr|dVX(r32Q(J9T!QrXi
zpB}L8NGQGh9CHkun64A75TXhAJEE~@Us+F!7oG!mfTjKYjR=6K*rlzoiRm_VHvTMp
z{XrDs)sY^DPue7Xk8QDT8|-(o^X-pj8!Q^y0_0wmJ9Aj>R~Wa&#ah$&9erQhF}5KJ
zxO)n3>T{R4s&N+oqwDH3y?dPeE^l_up4Mb|sba@^$l~mf^w7bWbcK-g1EkVM%~U#@
zemjM9?72nls<@McOEz27Qe33T;{6{b)h{b}C1Y^mgmR=fAe57Js6`cVdscrQuw&}2
zc|vUzuseosU|=AvuFI1uHjv=9%im%_V<X@41@EBEl}bAsD<uhn<@_lQ`1=5|Jdxm|
z_f}RG(S%;c)<&*&%`P|nsfk}mGH26H&wdKmAziO?_9vm0%ct(w<~fApXPB9y?LU#z
zqK+239GzU1Tcyogg*i&ysStr5%GTh|4t)#TzGz{pr<Rk`7b|Lqe?DuS({GpGlt5Yu
zLTcVCbvQ@6Z?gqwuXWAGJKPG^0bfU&$NI?`p$m{g3M>yKk*Fr#-y#16_QxdjOurBJ
zR&}F&XV8h>#2l-+kdY?P|ELDWB+M$<lb)^LwDC)=JzC^JShuOZcM60g(JA^k|5pK_
zr>x$(Z}W_X|712|$qk>R5c)aib$!L7f3bRFXDGct(+RN#E9p*|#>KLI|KBCU|3?cW
zoPzb&`v?eGE>HtHdu-?(H{SifS;DT}i|C0qcJZx6&joHu*6nhNYGP~RL(aaEy6ve)
zhve5zM^md=*kXiU%%+WFfAg_uI<{_J<M|!;pm_{BE<Apd>bv%J{R%c8Ie&2;Wiw3^
zcfF9WsFb>4MGHLmQ>r6Yg|}XN0Vm+P`ds3N(qPk>&OJ6AX^BI4zKTcuf}QW=(`_dX
zpc?;BpZq(F&iHhmx!L#fWY*StjEu5E0#lWpeT{b!K#St@^m_R3vCntO24ph*Vw_~0
zDpubQ_pwNA5fJGP<Cg=q$PMiD*AlAX0vnm1vx@zftZRNvup4lzF#zk0G=@H>dG9Rc
z+e0T83jMVCme3m5`MTk}3n!l~Z^+#WszG`r+e|3)`LxSwN$jWm&7KeN3%a{oI5t;{
zo%al10MVTXd&@&-;g^AbuO@wQWL2t0?98tv9Zt(W&EjF-mc=wu8?<mG`PSpwXYie8
z%>4QGo5B6s*O3wVB?Z^TDkr>sbv^fhmp#8Q)k$w@UeNO{CkSf1=tW9}-zdAEhjA!N
z`<+~JgOw8fG_W*96<$O{1UR{BOGKviaCIv!n=e#b-|sSrh=R#I(a+%DBNQ0)c;00C
z!d0OMZkY8#U<l}g@sozvk)9M!u`#Pr57&v=Z5UMBtygQ&kU$*>W&t=ph>9gI+t2ym
z3j*HL($EBWEsT0yB0UbmMrIgF&K7{cL%sg<228tCv+v2JzWwnV@1^FvET=j9z#W}&
zmm)ut|9svi#M}8|uJ)o0`{}T^=?rhYeq8f{)i_QTcM=B5wIhUiCqA4*Nc>OCK4iC4
z7l7x?$=%<s<v-iNNT(b-CSHS@MODPN!hrcI5Yp1o1$p)St*z3d`bkJNan%fKZJDei
zatf#Jl8~O@E6<8F`t&Wer6C64bgW-=VaD?+UU(^1$5?X9#JQ-oYdKs04EJ6J#V=|E
zA+;m?jro<`$W}lXgi4-xSZx}cIxMYMXZC3bzR=UM*G)IZ5}s(Wd1+_Kzl%DpJu{*c
z89IjM#rL<yb=+r*239NK9n+wI${MrA7_+O<F$&rP0#cC|r~DKA96WurAn7=1PaCU*
zqbB>v+(TL*k0bnQrfrvv4LFQBu|=Kc<R)*BsBLwC-)fhRSK|}tc}+~^yg;o3UYNAK
z+t@H4E!8XOBwZ9Cse&r~!^`O$9EiCPEoWIthi>>GvsseU?+o#ryvAn=FvLxQS5nmk
z8w<Y7Sxxy6b$7E{CRa~0h3b0OHGNj*AuvCI<Rczl-fPT6IP7#7yA(hTe&n|cZicy2
z%tm{QO%Q>Cgl|Sb#iqgCxCCNS6E~`ctzixs{6BN#IwTX1-cOC?2tQ!l`m58}leCb7
zR2Q+O>QyhNjTnD2!BcN}z-lL7A<Q5|Q#Lo6D5aiwKH!(GvtT`v9wq!NUWw|@P1ku0
zWYDyK))$S}{$+0ME+Q>2thZk6<FPAm>MFTev&o44xw6w0s$Ba+bx;wD+ZH&YI-wxD
z96R=XqUi?x7^ifa<wlCLqu<8qF}}we$0|NPz4T}}l0+zV(ZAq9BD1|D;-<I;sI9I(
zx2QQyW5c#-5+q_9-j<u;X~l;?+iM~SxyjbzHYR0f<Ul+6?fTaj<)?RuV1N4(h3bAy
zIQZMY;!vj|i)OTrQdqmZ6|NdwSqX`$E`z*_vUc5Y|HYQz`zL2oo=E%PdNK#qE`kpl
z>^f(wy*(yR*)|ko$$=0VRTZ*LkJc@Z6W%y(X8IIez8wTX7?GW986if-&PX1Ga^9+S
z6t_L3Fa(kS+PYeiuQJI!PuU3MFvnt<O?WM=p4hs^2&4`G_iUUO3lg=Ftk<=Wh3Ez0
z0&{8B_z7yK(Rbw3cZ&PJxLOpAr+Qc?3F(wka?XC$C5-c_3@3*+?5=LbB4#4SpY@3a
zY}LWn5L?gs_}g;i+UDymYcj2JazvErIh~7K5<QexCdhk>vfVa}-CC2~0)eCF`eE~@
zPoG-qH?Hk(2C-Q}i*&^J4I0hCJSb)d7I~P$V-(h)OsMGWkX+m_7Z7xTvsN{u?b-MW
zz6L%?(*q=Zc()CwM%nna?mBh-H;6K1)?uX9t?-ZspESp^ekun_DvKbCPQMye<sZ)x
zrqZo|<4(%Y8*awhXL_3WqD7HXX%IXHPKyf&r4FTuh4{o6fn>k93~-bXm{cY?8MgZW
z9ZkgUPos;yu6*^@d71clI_rfFq=eYb4)i(@&;5RmI?kKrq1PE@rI68ZR2&KQwdJOS
zZ(NV>TuyZIKupxIMn;<yg|3>2?Vg;50#ExRP{1fqP0r6$8dz-N60|)Sg$0|mVW{f*
za^O@Jcm(H)pIe|#;(e-GZ$}p=<b%nnD}>eYNrkQ~0Qza@KZ{B!>d-K*#L@^HqDd}2
zNBM4VtlyJ7i!70rO_o**9|%0F>mPeFEK1ACzquHt1GAc#29xUA80XEUiSBf#7`WjQ
zV7er5zdnucK7U*}`#Ph|!}3JMI9w*teZkyyt3oY(m{h2MF<!rB=%?%`In(`24Y<79
zfI4tjz}Gbm435S<yo|3{CVS~#BDbr?cjx@Z7TIRV#2gtJp@g?rD;>;6?tBPejXNYD
z6}w<prPmlz{o<(oi;lA`5HvR^;3Q9|1pXZE83g=hY5-AVi~l~UN(<Bu{|;;>QIEoF
zx_lLptz`FwPpCM<JoTVEAzvK&f5Ab`L^Zl?bA(nL8a4y}a{D2JCjdW=i(-s1DgKXs
z240Y_Uufy-?so&*loS6JU*gz4KmINZ<cP99^VOPsFi1#x7V2hI^*ru9E<wm$rMvvS
z2;^6j%t~f8I*vi4TZ=3__qH9s80v@loXN;igqRKE6IT2TU{s*Ex2iL$@`nUhjQ80E
zq~q_DjOLxe&E3*Vgf}?JO2RcMX}uT{Qe`v6Q2>U>YBOkF3OxpUr21#NcQy-V{T}#s
zACn8ZW%oowkbfZOjz<I}RnU#oqTLkwJ9`zX$As?iO`<%I8)~nKBImO7vlG-7C>Z4k
z+VvHG72cQ(vFES}S+^g7eEVg;fT<PQ$d7)yom}fmhKixGmQhYV&%PHtL+14r2U!NI
z$%KVUmddp_=I<}gR#ZwxM^rmnJ;>)u_^j)+c?n#?)Coz;vCDj=EslCWgypu&dOTal
z-GA8Hh+wl`zRG2yVDr^6Pc{v3Z4e7In6+hG-*EBr&d*Ah{X-r^qZ4yB#e>-MWbR<`
zSX0rmi=w;q866f#NM@+GK=!RiotjoSu*uPgt~`VpHhB8{`mYcF@F>~5^lF)2lsJ?B
z08_oqEOL4i@#)hcbchv;4?Nox?TZ>W<4X{HU!8)Ea`UsYSfP|hzyI_0Y&Too4GubE
zw?OQ`eHPKJp1|Ac{7eKUJDf)_&!n`0F$n6^SGcXp7+o*w4))T?!-B5$gd9$Xg6Qbz
zcGsD8#yX_%f>z!sGb9{Q^<ysX^~1$UNrlB}5wZew1PE<KaRCx)<9Ah5vT*6a?w@S0
z+KS@+3mXK`qSu&|&N+~@w#t%%5mDol<luVl1mLcCv}OdLq5-DX?z&R!MLT!3;wU7y
zml_O5mou2eLS%A;O80`wG0tymg5N+T*gZHzGw<eh`HmQhGq^_{e}%MKx+Fnlg8Fk*
zv{WS&MT~>15VAaGBfW^C(=@_SOY6yLFwbDJJE*UFF=$0vE6C5ndiIan{1{^b%>llb
zCF@X5w?l$`BMETj)(#g`6&-Z~O!-yi)(YA801D*Aq*|(^wBu&b+jatF(dN3G({Q5z
zU&~f~{@Tu}{vbAu1g3<LdP;c50}%N=ZD%Ug^W1*bREjzh@c{u_UvLrwC%<JCnJ{CJ
z@CH~<UIi~q4n+QfniH5diT3*I>5X*0rjxIz6|zjP7Jq3yld(M(AR4Dr>R{T)(fJ3`
zR`p)d(pUL6JQb%MZdppEbcn#miQHQ#)pI#cGPOT}cLm7Cmo8QU4;2-`5kKkJrioz3
zO;G)F)C7<H2#n@)IE;NS6>rh+9smH84&x~N=MIz*Em4)^3vTznsPu*FXFHce@|4$h
zl0=;f|4w_=8b}xi=hy~|2G%g7=nuwo?fAGlU`D45?tcRnlP@vcBX_BTq;gf0@D|f>
z-a6S<@v<ar(RWFN?2-@0B;6L(SPZ1pxa!I|38kmJaHIgwcHP{b?Dm=3vAGrp*G4<e
zM|pTUH3Pc@W(M$(Q~TyEF4!%8Q|YD1BpEk~FPH2e2Vwutej^6d3$f<AO{hD=BiZpS
z9lcsM{A*1l>H-^HG^cU5#(yDPH^kgVDjm1QBpKgby`UPPQ|?jz(%mI7nscG^B;HBJ
zx}T0tU1US_A+eB7k0{;g1bz>zM&0k6oz0?7h>Xn$On%SuH=TKa*=Ep_N1rZ+zsEz*
zH96PH7ZdyuIN?sn%X%Y50Xh606unD?6hm|$!!zRwe`Jo@NFZj$G6fQod=>kh4Cju|
z?4X%~JprkEKHDvu{O-`}2L=ps45)$Xd-+#(n~dOo_7tQvC?L%~>FDokgFYL?{1%>T
z-&o<)tuL49-7_;eC*fd<?RXP8peT`M>igYy?Yj0ouNuJ1a>^`w_IYYFO^L(!!v#g=
z)56UcnCpw?!P$M8Wa~fNPn0^}Q*@qH(>6h`%{J`_pSrjF<)OCjSB|rhS>Qqh7D7_R
zPy_MIE_+Urryk{2s{rY{^s`|lhZdJ3+O6ConR$DX`)n_kz*@3vj6$zsg^XHEJ=(IK
ze*%?gS-XM5@uBh$A?w%&%BM_UbmIxEf4RiFQau|Ahf5t9J99nHSj1?GaZN~qn)Ad?
z{56pB=VgB+z)oF<RdC$)F;-BqaJhzgr)Q1QFGt^psj}^k5p{DpOzdHlW+f&Ma*YEq
zI~-EL35iwGhJS@AUH-_=UyoG#mI^lT`dfXo6L{XBJc#yhsH<mUu?~-d+AS8)qM~LZ
zYQYmozQ@Z_EPYoX@6)FfU+Pz-p-Em91+5fzgwA4O9t@mC@yPCwcYWHE)&S>BiAxSH
zP(2pUR!m9?e{7tw7=8Z~0g-_^y5JGuvf;ARrLb4uKAZZJ(_`2T{S}<euHI5}Bacry
z_SHY(uJFU!gt1{!d?+{meO@7|Iuyi_l=T%c?||?|WmY}OtZch{nIbv)jrC>gka39h
zW1s{jDyNchIr|oOF@S|*3Vibn$J^t7QHVjw%o^JE2=G0ZTXJt4?d5rJ2-60>#h~t1
z69nV62~buQrJa#0cRkK<sjbekY9k~)di3ic$yQ9`mYtxt;2L@Q4ur!d<E1pq87{%h
zLM5Z!2$+OcRO2Ck?^`$7bOO8JgYpDOj6HJaQ0^x3{Q>o3jbD>TL)4SWmy<`41!UBP
zs`SpzySA~JXW{hh*P;w$)cN0+lZa^-p4ql_bKWK(#nw9&2L}fyYT`FPjSZCNXBy6H
zY=_(0@obEppzCD8aF*4gp1amR<_j9~XmNUEp>h}`jr;GP59)SK-1V}xZ&?A9tVjnM
z&)Px~-|El;NRd6wBC=<Fw2T--^zYmFg12L;P+P9`kYs1y1`?1#hOBXH%-lCnGc5z>
z&};eNGFL;_ky^O5S~|5eUf6fHjh>`)IVK#UR4Z)Kn2`WUv)11uWj&vS6+njpb!0Gl
zA4Vb!Dz5O9OjzR7P4~Dy;%nPq9;&zktqB&q%gZ*QoyjifwtlYQ1^~uaj*Tbx5l)C=
ziJpx`hdFay00<?RZq_Cnn~i*-^t8Z#YC2-?e&K478AI3u?#r1>vE<Wgom*fB@ZF(v
zI$1wiQ6=j4=Yh9du~ZJ}k)5u6M@uG&ZB#XbJfnF(<%1G-RnPwbLq%uh=iB^#*2j0;
zy!A(43s5{=#<IO)uIAiq2^^x@qCh)eV9Km}l6}sIwa(rV0X{NVwAjc7Xzos)q8mOd
zc}6|FBOJ72^MXMPw)Pqx<oGI2anQFh()@SNHa=-&$7z4g6@R{NCJR1nnk7N#p(*j|
z0(2%aw%D?_2f&APlkLqJ6t@)ah;t0EA`70W0HsL)w^+*)B*euNk03)S%$x-_V#TJE
zyEx<s%xvY-M-5??YCST1dm*Wzh4E5mXd039!CX}$DO?s|6=kUEJGhZ9i+X978w#vL
zc2!f;WpRNuvt&=wLnU6J5K-&*>HQ?yA{&&-DHPG!nMMaqJNl?yd1b}<foUSOXOjDr
zFqGJ5hh5((o{5IU)m7-GhR+Y8B))Zj?@~yz<a+6l#^+m}-JJ)s&(tE8;&t;2+FW$=
z-?T5e42&_}3SiAms5*Q+zM7J!uQEj0q)6rO+>#<Hd^;-YeQ>qv9I~G5>P20yzn)Zl
zxCx)6L=O9qxW}Z9S3(s_cb$%8a)W6!zQfZ6$B_$xuAGIkMemg4Fe@T-CON-NVA%~t
zd{R{mArpf*Ea@wDcIJ11EL*M)m(DY?5qO+cR1_(y#rXXnaKlqnR1D`U|FeAKy}KnU
zg)({}eR?U4B?VsJkOI&zi%HW!Q&V6oK`Z8R+yfmpRzEuM!S`fG(XZ*>=O3^>;g+~Q
zUH?NQMP$sm<1>WToJ=Z-eDe>R`DFvgUxs3w&bo2&IDYxaselZ9&`XPdh>|Ko!lTf8
zb*~`2<|WnhK_lseMJt?>3;;WY+Q=|!&J~|4PxjtuHe;4PX+ms!`9MrHpnqUcDKDCa
zsdQ(wbnxT^91xCLzsYxMx|sn=*g}7nO)RzcoWZ*Ncj51ti8+fMmN`CY+Lt}4fk!vX
zvY{XCfL9O?8?6s#Tb%&#TPD?i{`Fuz-7Bh^nU(bJbg(X#BXj(!rhsC&+{5x{GxSjM
z659L1mke4r3NQYR?$c9Y--6Y1%<feTv}Tsyt~i(Q?dt04jlhP%`M(GHzJ7}jXeSOG
zHjlq>-&Md|UN74Z#H6Jhe?ZNWvA3{lYHQ~fBm-R12Tyt3qAoK>s;#4ugOTu;eJ@*M
z7S4m{%NKDM99wpBWmLtFf32?(m8iTIZisIE-mo%5c1=eFiV;KoBO5&l(;uw~HeG1H
ze*I@%Z*Fz4{M}9U+jJwDlB&(A8&$A#?RwLOwIbjs$fLV0l<VCH`x{`lX$4HtjMwsY
z5w!5f5A|J{jD^GTxxyPGZEeg2_Rf+3grO|fTxNPhkFC|=PrEWZ1SB>`39I0U+J%Yf
zJnXWsnMw5M)u^|E!y^rB;EhD&5B4+xMa#I+qx5X|_m8M3&lD*xiJcBamLy9ndm-BB
z>Ki0k0vmgQ7(tVtA%`$Q-)+}XI$i(FTEaW@d!?}>vAfV2K5Y1~@Be@PhZ*LKSDOtI
zJEr|wPy75Gpv67NI%)OMie)S4CvIG!=YI5DtVtg0QF=ASKB)nCl$K~-0Uo8#9lZ7L
z0guw1($83r618Wq@N4-G&fa0K0gqBnnMz3}V2kQ}k^|VH9)%gYl7FLo%xLroYm4IA
zj<wGZrCi!0EdVI-7pCzB3V<yt{Ot>@EsFGif93z`z)Iw~?-;&S#JXsF0UtqARov!4
z4-A4CXOuoZ&z{F*dQ2bFv7<2Ed4gNTL8!jT<|dwTR=x!mtT_d14oa}e>qRQRT@dN-
zJGpFB_c<wigsp9RMU+$By(?f|#wAXED<Bxl1#`_?_X8iT>`FdKX<tlnPqU*z^LZ^@
zQvjp?ADmt=Z_C={`nOi_uA`tN_fP5CB+tKH)u{6hKXsoCZ7J5|-FmOIqY>-3D5%Vn
z>Ac1kF6^1t>-l!pzaU<OBlN!~RjQW6;)ejLcfknvQN2PfKCxl?c#y{--lPvQe4<7Y
z_ALbos13<aTaTxI43}JNQ!{z51RBk+k%rfHG9wkc*t>dL{EoJnkI~2<81U2j?R&%K
zrBXV{H}iSBd+0X`Nv@k;dPNjaEcZML(92B7@iD<X?~7lBIaN83aHU|BQ~VnrAD^np
zcE9Jlfx>>#X%C|wZkCx-U`DLoB6nL-y#^0#YXiJW`n^sNwLTk_KaM$y7(CX8+VvN{
z63>d!p~I6!IKDR0$$t<DdHVwOmGlAUpY!$aKQ}DL%gud+{8ujLeN)OB18nMrNt;Yv
zVHjz5K0cUR^^(;V$HjfgKvj3ZBQ#Nh>i0arX=KNFo*vkoZ9VV*RQCphD)l+B8)m12
zVD^EVXtNl#ljf*L&$$>{d12a$GIFsJvia4<b`|ntV85msa8_{J7a71<PNV|-+`GH-
z*anl1c0A|0ku~BCyNQyB9SU&627T<>Z$i>bK3-8PUA{TH#YG%V7nd)%VJTvsW~sm(
zfm`}{fQ#mt#;?4Hi0EO!1$Qt}OLR3XyL9<vP_4$;?Mj&5zZ1MMTL4L`7Bv`EsnoMN
z`yMMrzLWhO-ob@RU;LiAg>K=77zd=jX%+(fLPeFsdPek*y&r(hu73*5U#~-re6{TQ
zWlj5!wbgOqfNavy=QfHr3G#_!X-vTmM5G%+JS?(Dx{>m#Mb>mVr#~C(slZZyHXFk#
zN`N;BYd%v%?o@cZvR_1J3kcJ4E;3TcTz|G{mDyneM7nP)otmFL`%~X4{V0|OZz5aO
zT?3nToSM4c4^x#&8#GEg9sy#T>OpSHVVWcxER-gdI}7W*O9Ct{$I-{9KPVI#99!hZ
zvfsw!_%d48^jtS8Re}TSMF#J3Fve}i)GR%yHNJ~AM)j5x-QG+u3c_mnV027B%1BGQ
zucXZVmHB6Bl>c&ri;3q=$r^x3E7=3e+3Ai)ijqjFhXAD}d9pp#r?+5ued_MvZQzE?
zKBK)KC+6bEV*9DKnnOwjUGA}YH208iFe(Y2<Rsi1P;b-iGrw;LIQataU-nO4^4r!r
zX16O4u|uIN09QJDVy-6s*hh^qIhWNgf;n&l43mjB@-jWRnXE5#a}98a=hN`MLL@?{
zu=1QrlK-!*YZ-10N_mU@{}yP#)}<B+Qe$RS_K;GG@fXwgVu>m!m(5%VZutoRX=`%i
z^2`1GUjye92d|%F3fWWx5+Wq`00Jbo(6Ldv{TAvcn2>@Y<-X53o=J{COVtTtw|~$G
zf*S=;FzigbYEr1v6^SX&xv0sKIpdlb+&3Gq(<@aygijW`PSVv5T!I|HCrL%l_y%tk
zCevu(_b#$M+o1|>+!}RD^A8~BdJIKN-*tsKG}@;3i=59r=N@U@QLWL(T7|Ne{%RX@
z>DB`NA~ql5)0%k~vkIrMZ%P~|{%gi*>KIXrmUc+nZ^7KZY_?MBjQrI&XmJZmJOYXB
z3YOY4_(jBF9@j67)-S*<2WwEPh(44?ADQ1HGB(so#jl<A$CpnT#La%sfztsF-%gZ@
zFXT<#A|MrN<f)#Gg;3V7=JW%Gz?+()qNlp*hg4@+zNo@G%w4S@Qx@*g1fTY}uazPM
zdrdB0<Zs8Vu_%T{%D@hq1RA8*MN{iEV{i7~jt&HgXxs<T(XkTuP+!XN(YkeF(%m>i
zWz{_ayQMGHSM0L`k>tYCL+QbuXWi&dpioVAKqe?uSHK#?xJAinBs82aI3EtcWqEv<
zMUA!{?Z;DOaXfa1c{cpoVT>$l-`d)n-vcr)r%R5fxbZ07K%z&UvA1ip<Ad@}U&dO$
z%c2Un{0QGqTL5IU;D<Y9IG>TzvG-BcE@aJvS-jEcDvaGmSFB~pr!Drjfr*CUOZlNB
za_W3V^K|^8>42GBZo3*C?Tm+^zm{o6#J>Z4^ZYzknD+G%UGS8To%C3YD!{Vc;$G4_
zvrZn;C^K>}VXB9ShI2UKGQN(w7D=Bc8_t_eu`sWegjFUgz%OJI?Ja4gAHpEoPcuHg
z^8M^5R3YpRHT^x|htQ}!8>fu6^=Lp1wTmx^a=9SX7*!z0i$<38(Q)o=lQeB7*U)sj
z^di5q_7jqxNO|}jNFpy~weGcpTIXa<D;-6kafiZW({10X^jR7noi*PqU3j0Z{{m)2
z%v&ewbSD3mC{47L=6_yDQC88?tryAJxwr&{dUMCb;qc+_dP$<1xOCA`{Fs$;IPRO;
zX~%*RE9vnmPeacSj2Brqc*3{Eo+S4)V=X%TchVhuN~<H{*M582<mmoqHg%Y7DdJ0t
z6v>@PtN!ev6j}JAOnj!4x1Q2ahQ{4mm63waNsNUv({%mRBI+%&ICrH+@Vk<)BA)OW
z!#-?8BRG6epWldRm?Q?`90e_k%CT0zoyCirlZAt6#@^RmJP(pUFS7An*$w~>#ud?Q
z<yn8RbHFiINpO;P(-xRk$06eVe_FK3U#yc~M3kWfr1#Cdq^XGa<E9HLv_{BM|BjTX
zk>m7u;f0OzN)^mw09qq-KMw&CCiA?{nEKa&{G(q0K(ETW-&*Z&DCm8+t#98#lAw`^
z7xw;~Q%Jd=gnAq4AwmJ$DeB}Bth$F%dn37X4SdZQNTYbr7x(sg|B!zNzIu_pZHDXu
z4zO0ExN0envT^`(R)CxHHX3~Q%KDc|&*Uhd1p8UPaQzu^G@xw-*btLewCictyjO*X
zP(J>Iy?kf%GWJF;2H9Y59R7;9=)8kPd8^<Oq&)XT6Bc$jj_-f8W{EFyE8UGDBo!mB
z@w!m5I6h*C0pZ*|$;t$M!O(I}fw<=@bZb)C9;_}KvBFIzKCY2buNy#^=VvRC2SKt%
z*L<^XzkZWbqk}Ld7Kv-&(ApWX4Jz})26zV`nyW<j-+XUeDE6FsHV0$;HD#YOnwuPR
z_ujO9j?z2S&YFTy{qWtm1ybrzO5G-TiM+81+%QSzMse}?NT1t(Z8`AGxXmVj^e=-2
z(isLgHTz?tjsRn_THM@L_9$E5n7l=;eF?GBebN_NJ#T7_(BB1D2JaB9M2$gf?fBrh
zj_6sHl*f+OBAkof4uY#y*_*4@zqoH$a;BTy?^%c=C)9axnTY1LYT@?>`=#THW@T}r
z3~Zji(jThMwRc3x0ah^8$6XbUNoC8^BZI*=khlWjW!=%GN%9vyBGa-sJ%ix6kHzbL
z3eX;)9D>IEk3CXMK#Grl-=xqw@m^B&hpG!#wWDa%>D#pwCxFdieAZ2`ptgharmLz~
zFz24KDeBm(G-m(m^(R64U~W@MKQp)Y%6gjQgU5@U>kP7$q|I~Uq5~hVglIkpacE@F
zb_XhV&>Yf#F8vJ0Jz?{^J#*>UijzcAp_aICo4~>ee}EwTMeDsJt8Ky62VdxVem7^u
z+&)P;#?<qF{j7;Ux!~r3wXh?S?qgK6LjBToSIMY}pX-F1*9_7CDTbr^?5n^)ZQr~k
z=tdG{5U4a_(N$i#VEQnBarVIs1VK32u=sdmmJGOhZ$!~F?G(Zs+k(w30SIxh!^0w3
zIpQx}PjcK|MxNw0Q=3~(4&w}ZEPgPxS?RjlXiNT4{EiL6bCMO5YZ}?mm}Y0t-RNda
zxicfCWyc`aQat~iJIS)vFd-UNl<sdb=iG^8ZYlS~We;8It{--rU+eV>DC!)lj3Upi
zlW}V44sFj%njs`4<`A@+B_+qG{EFcAi;gdims-}XPhGC4fc=R)v=%oGZ(NJ;4Dh<q
zZp!={d`O@nrg?5ZwZ$RR$^78=Z<w8zgVEZ{Hq+Rcr2DuccKpOMIz7&v`}WZ)b}Xk0
zS%VOT6G8#iICK8nV<qMlY!mfvN<*1nkGGK}e@~CfAM%%@25)$AKKVUGmos7d1aqh1
z!QpM($uTq3xXQ>(DxRvTklyVC5em6`s8|?VSBeI@2>_P(K&4smt}CNq>*mAyNpPpl
z#%)B{J;xzN8Oa$lqkDsu;@M9QNWxR48vksExxEtN6FND?f9=GQ>~D5Wfl1`6Bjx$u
zmFNHO>g&^ioUm5_Spm%l#5^LbEc$<uCHWw)VetC3y9vOlz1`nEn|+QYTmCdYePv_>
zw*S&acO94~U3=49QF!p(+mJ{U@sIA|`EAC8EH^)Ejgh>{Pd~f8-)P*-%v1L(KZx!d
z?au0d%ydni>mwB-6iP9M7$mkXd;Frz7u%x()JHPiKJV)o0OJY3Bj-Mwm<C!Wt`4{R
z)GoLf9$I)Tez~}ss4qNpLw93mmEfm~l_W>sOfuiB{x|<zGhi(n{-8;AsiUXY^rMmL
z-=Gc0ekK`>f8X4VUvPGLS4~`k_Br=N;tRCRW&RuW9S}7gT|o8H8iT*cR(5=oNe0)B
z+j?4S7f!bpS`Faeh=*G?7`AGX(89*^&V!LRbvUwa$3Wu>wm;+A*w%Lc4jKm=>0$76
zeO=x9E)wW|=z#@b3Q?E(h5wi>!1D-05VBy0ZIgF<YbdkR9II!Ed7}H#pN1HrV3XQ~
z25ql^G3gBKkY-2#x;5Rgraw1bz^vWlDCqSVpms~F-~9d^QR^A=1o*NKp0bRb+pR-c
zWPvyHlQ8e-7RrX>Oopt8-w`a!3dgQ7JlKb7)HM|qyZ*4MOFd1Dt{1NAGm^ZXsLtKG
zLc{;f3w*!5mVw!CRqmhK9f)};WTJ6%1B64}`9F?1PgzaP;ml9G^}kO3$?8}J6;Q}q
zmlZs`{_86Q(xYmOLmkMWGmgm%b`5VAR+)9MS-0~p4DizdmE1oI%#DTB7)*8{9T4nA
z|L2!$s_-=aT);X44Bb%QG*N{EH<1)?c9`|CuW(F{1yq?@b*YO}ZTpZZyKtVst@%o1
zF&d7#Cy`hm#A#D7tlSEuOPW<VHh-iJ*UbMJ;)Y(Af|~=mxwv%u07slvvd-?$2v;-o
zhbLyoy(=3JwmPMf9&9~F1eObuZM|?Si1bK@#MFi&CiJMgW_TvHa?$h~DZYV4yovt(
zrVBcJmT?P08!cqLQ#$WvzX1zG+Ed<FT(|9ojY&zv@j-r}%J@m{s#-2j#dio3>PTm+
zo1V@-UB|e~_ldSHFN+-B89K^CRVlvboNL0^?08@`OvNl9RhKXXzvWM_mr(b(x}7lb
zy?mT=o{0zL+M*lf6r#3evizgNR!)Qlhw+J)7#C-&vzUqm$nIgRQwXrM9h$U<^R{$p
zvvyEQs+#$gT4w3SxkctWrUMH#-lp;8GLb$Wq^#-S45P<Wnk|{~lKvt9IWVvd^_YkK
zIiGlcxY|_fOl77mY&#?8B+pvqnNe1x36va^ZlT{LJvL8EQS&|{ZU<pQPl5u#o70V`
zL3Va_v1H62w{oM%@pkBkL~8plv%6UT-CPcA$Iz{BS~k<ID9=2A&VFOIA@@GrmMKC6
z)?__$RmsxVTaJnLu;d-Qn^^U9H~EO3A!G*Ou=vclXiL>_^mn9=!m|wkzh0R^KOMhC
zI0d~af+xD`@K5xEpn0G*dBl9JB2=%lJs?ng0$zW)3X<j(mU8H9{uAA@^dv8)M-Q3P
z{{)tBbRSq6daRhRBJiW=v>0n+uoX38#M+F#5G_2lZ;1(=RjV3DucIq}D`S-*kr&21
zfQ1nV%cU9pd@b3<`CN`#+|_XSa;45-QkCH2A^d|V2-6d}pFd~MpC0{KOvx47VlZ<q
zf<b*5lT0T*9fv@YP0vvWfYotG9Z&gM@2S<nPZt^5EpshV-AN4u2FHgpzQw0#(5+8s
zj?sR^ZB48XRZ}lWdt=?0Bor*c0rbKV0b*JT-N_ZbmP|fisU5zxYPo6|Bz{pLs8>v4
zmfbIUTCBXw+~iSXUGr=M$vr^0za}F+ahWIuL`+ELvlElZolk*`g9&4au{FxP#azx+
zVgpb663#CGsP`bd{GsC`gj40G<e^)131ib%sZU|&_!OwMI6wuBjyR^!3Oap?>cKb=
zG5)1OOx)9%5Va-8g-0Qe_(8SX0}C$SA$a9xT6CN#sr^85Wcfs*|2a2WZmO`(jrf~O
zgRy4i{y`>5Pev4weIcFGy*UBo*L)K(xg`aF`>9=`3hLH}Px|yw1nP0U1KK&7dR;94
zrs$??W)8}~fs`0OlFitAZurV5{{Y39&t-e(^|bL$iIA<<&mwGkg(*WNlFVaSm6W7n
z`#J+l{#*@&GjKA}ZlQG8tHodC?>uCI-WJgqI(Y#c1D>4R+!Aizgtf1`u?n=J;J_!~
zSk0tw>X7zi2%o?1OS~K(6PxB!)!QKlq9>Uo6tF-042}!${yghSfk5vXj(YRldULtW
zVT@9+qtDkS!X@x?B8R(KhK7;z9tg~jr5-ASiYL4>X%}Jj3sD55rAbwZ%llFdLh&e8
z#Lay`l@{yG5x&PO!q4Eii5n)oxr)6Q9Jm2WN~Q_u&BQJ#z}Z|Ed!?tjo7t=HdB(gm
z{a^uh67q0+Bt*lxCjynhE$nD#I=*F}R8?crPzfe$Crf=y;XE-q9c!}vXVw)k;MGm%
zpt>wx@g8|)JP39pI=7vD6>mVcaGFtyi_Js>vy*ROEo8LlJW0LxYf;Ur^eKR>8VG7;
zL7yQ_i}6W-B6l57Xg*9^d<W#Mn~Vai;h7T3AUtOmSdgEefKa%(=2Qd)bmpM~YGiOD
z`E7|BDqZ`?0ZSgHpuWn1B@}3Q8JbX^FP;m%{d#oNPX-ySUM+!~{#XLIel*PGdtM|_
zY^PAl%DR1loIq~j^(0o&T9ox$j95{<4~!)Q0*5c!np6b~4GSbX8ea^&C5-=6tJ5sA
zKHLrH0t%2Oz8{=R5b{LHTgTf8pDk#B=LtZq0hk1UB-iv?#h7aB-gYY$69hyf=(yWN
zd$mWtCkmRXN%AMJ4SD^p-g0Ji2U$i=4d%CfbFKy25e%m4`XZ31i8b%xeyWAe90t`P
ztV~~k2EG>huZf8b=3Fr(@QjQA6&$gN_Y>6g(EiseN=BD?Wvc@Z0ox#4yN9_6?)bR<
z?4=0}Yq7wwR}z&(B=PPZC4T|c$)~GnO{zK7&BX;#+Z#Mwf%A7MDJr`RTdklSKGSgJ
z*zPl9tC5R#w=9U|cY_980>D~J`HkI@J>AIjO0l&LaANA2&y5Fejn>k^3U^3SCa*yO
zTOi^LJAZ4f8}gLE+hUNnO&bl>cV(>`--Dj%P30h+mO&OKq~?rla%2N@JNZ2?qTVN|
z9<&Rne>x6^lzs^$>Kp6((&4duYU`BoAjF5ZB0#-6QG)1>*aeMXA3m9bTQVg1u^RvN
ziQs*jc4CqcCJCR0D}g#~ToH!gw)cPn?-Nv{1-<f9<L|l3!6*Qo2rL#-qkMv<u|Vh_
zfxI{MMvvo1Mn?SAUi`bMQ?s(NR3@jL`~NxFQeF1BG=Ywcgy$R=O1%H@ihO)*MeJbS
zf~)S2sF3@uYjq2N<I<dC<DV2OX%5XbUaK4iF4WM|KOTUtEF?E2@E@I2otlOQx?oay
z-0F_?OJOAx;I0d@yVO1Yhq~_M^YU_CyLK~;>N`^ZoO+{u*?iO1f*z8?qQ5GtiPDmd
z^BtfW_z19h9<3B9pLVR2T3TA(9<;xM0FH}i7^8m^fCCe-=WuB5rNiFpxoP6;o8yff
z2&(CgsKeocii(5t2sN74qn(BD-uKU$rW5ZVasR;E3~oo+vLXg9y<a#bV~E|LjGHCY
z@*jM@paC7lc)@vBxOf6z01z=-Deeok4YZ&tOeQxA)hE--(CR&Dp6)_s?F6T5!7s$S
zR&E02UwDdcguc-J6Z=awX4>ONcSC0mYQe~KgcDQ%Tnu+BZGnik1Rhqt#hQntF~8r{
zkeFFuGVlnGk&SL|Q0W_|FIupD)%4ZBejQf&tht=>ztJb7Hpu5O!@K~Z;U>r=;a=yn
zd?z^w4R|Ow611WLk$$T`nS<qc1FNj4-(~<Go%p4<WQz;}wAjr;VIg;6F^bK<@YD>U
zohW$HnXgK}Kq1zhv+%HNFZ1!lK9sPf?H3F;BO5BQ`YhdJt?#Su={f5t@1d}At|oYR
zYyqW#K!?X@3V1wgA6e0XB_2@?=6DRQ&OM!YQ_d{1!I7N2dDEVcfPhkQ6YW<@B?E*e
ze~PjvJbsY35WA7dKuVoJ==W|GmYcIie|n#Kav)&YF*Qa>pi6M{nM|C1&m>#rQ^yg%
z-}P?Bb-TKR$?lDf=5rP!F{<^fj@lnR>LKrVTiOy+(TQB&cLY}?N`yAR(-d0D`lqd{
z=0=Qv!3Z#{re7XPnm)EBmQs1e48Y<)G)F*-_cbPts1)`M^x>o>kQFP+1hPDH*z>p|
z#9U1Rs^9(o^V3ydp9YW8rLA}Viq2}~QIJ*x?;j9k_8I0Di?5z|e6--rM9%nnTkUP<
z`N8C|2m5D#IVY6M(E<euH5b1EA&0(*z{?M^p|JaV9xuT3mi9W%7Te*ax)m_+i>?-+
z<#Cs@3<l19ejzjPZPmf~ue@_dJijgBVlu4(S&!L`?B|ZRtY<BF7deh;ejQ$nNeT+^
zj!)&=c|AP|4O^^y@OBw9X}T?GL|{HC(SY=_KM%g5KV<}k$-e)Ki|Mg|TcpbU%7v#I
z)TjY)9c@W_2S?G_t%+*T91)zJrLrQug4t|yAnIbk7>wKPmXOwmmm+$Sv>55vS-JFA
z+y4pvgz!cHtHz=M@#$&XIDNf4P?0@Jn98<3;ak8q3k}oBQ?$ApV==!)G=b<&*hyLh
zu-lgL*Qno=U)#RQ`xX|Xu=Mn4)0^cNr`~s?nTbT%;;rxqz~s<kco@ZFX~acy9OSUb
zI6lSSW`hYK!BYkgLFVp?NbimRF4D?X!t04a{O&P95VARu<Mx2OalonllU$XlP?u2h
zV=fn4NXF+m4wiVcxspT=|A8bhMVs`p%2)L5bbK0o^;sC2>ta0~vLFyo<eMy1=KPM?
zPA#qfirH)=zygmN&@M0!qJKc4G8xAOkatuM`H&W-b9V+LfckC0eSK(Y0Mh6*3LtOt
z4d9OsC<a`dq#_hpCq)YO^CJ8WlzLVs4s#usc)kqZZ%TAZ(uLuLeN6P=@6CZBUT3E>
zeS#U+?M*<)Nz+$r;PM!i$VEcZNVc|pO@f%i<<owGo`zt^JHn*kZR^6fA{*r-PR@21
z#8CgL!sw-}se@s~v{<dv3#&QadJmIb5-9)QXXjc@($$>)g~{|p(TW^>4hY<#*Rn<-
zO)d`vYKaD8=|4WImqI3;1N{L>mc)A9;#nji8Dzx%Q>by!6)q@!2ixxx+X9?e3v>JH
z{lz2r!=IxQd~JlBHG&xbri6&9Gd@?)YvHwDR&~B?>a|*E0=nHjf=%90g2PqGyZVk1
zR7F(TNk|FBb50AuS~zd!q9?@uIlXYtUiEx!mQt)Z5hNW{*nI|FG{Ym90!)m#=oy!$
z6fL*YAt4!!Ul_hJVPS*Tut1z5Jij3NSZNtMn)_VllcQ{H!ku!7$7esUm}4*c1^Z&j
zscn>;YpW|dea>3NL?*23kMa5aUnVcA0$vjd(DCNn;UQM>5I>q2$)yrKhxz2W&NHK^
zlat3XC_JeSg0-$HAwRledW&k|A^OKLh)P%yUGfq6eAqSNN`a6EJY?%N;A?zkF`||L
zAQr=PFzwkx3Y-r1@;H9m9RexaO(N1bpatUx;6^N7*eQySp@1IZg*o)}<>C@Z$R8ld
z)96F4*~UnRcH#E9HE=ybw2L|HL_yUzML6{s;5%Dq5o_|Y&)s_7f_7rCj!72mfgFVD
zMFy=&0D}f}vA+q<4sH=}0DI;B79(T2o(C|p<f2%wGCAHrs;Hu0nutaQ*F&<Mxf)vd
z9fa7AuFY7gG=5u@IMB*c*PhdS#dGABp<qO7_jEhZ;Ew1<U?r4V7_^fthf-<F1kBtg
z<z;Fqn^(Cu#j<ccfa8tMS5bSIkTVgyH8qfv`lvEph~r^Jbo$@@WYD*#TB>?vax5{A
zayNk1_Q$6>`MlzAlRiv%MkO0mL6>;twDn)=;9yf{;cCpWcpiTr{kBvYxL(Cfwo*{k
zM=TlrEa4IZIdXV?HIq_MqOr$QyKnYDH(w$4bYgEAIGxcw3GOUM5!Ta~KRq9t?`@eQ
z?tC%m&*>*vSrQeZ<}|Q*m8bSi<j%L~fD{Vpv!2AGs6={#0U4I91ou`0j&|}ZTw;?c
zj!7Ibmz0NUE^iRptF81~!U!XnR+V<fE%!h+9kH35$wt6zr@tNO19t!U)AaB9usNvs
zIxm-SU9e4>^%kS`NFqDQW%S^83IQk)X{2YSRj~c{8k;6ac)sg`1IL??7a04&Z~kMP
zQ?fvXx%Sc%jcnb`+x0(d0GwI{MCq*@g+SXwl;58CE!ePM@)*}Lem*1vk+~BdoZUh4
zW+e{Jz0(7oAsGSw<OE5TawP>Le>59eDm}@=`)bO}zH_As5EM^Bz@1-3c(_DK%*3kB
zNBB%yP(AX2KrtBO>l%@^F!$tidXo2yBWDLVS|_0B(Ohk~ln`R*^{`HB_*1~>3MYk4
z&Z=Z^r)@Fl#i!s?6jgST!Q0+@=wkS7GrH=VZAa+e$Eu#|AfhE6G2IO%@5?!SS`fB0
zJ>#d8VY&MNPTFH2QtR+Y=g-8vbJ`I_`QYa@h5=DB`0Z-R9JfJ|J>q9rQtrD#_?;MR
zCwTD;z6-hv+kk0jKs3x;21q4;fzbSv>ZMG2hz>%R0CY`)f16Wl{I_rWxwG>cSY|=~
zrz$gl-A&ICyK}1lqpFUV*$>RD!mFw|^0jl5rc;NoKr$h3o$~PKd^q~70Kl1<K!fL1
z-zn4CKcJ_^S%B6K(2Yr(3-V08WEr<-X}M1mPc1l-AuhfBkI(A9Z#Y_fj-YvKbUg`l
zvM-9C%M>!~wu%%MZU3zE{)Y!y8?+~18v~E6<pD;VIk&yAk@7sIGkX`8efN_^U=E=+
zjSfb`1<M!xwW)i3e_Ew>#ZTnzI{PQBU(t7`lL0TuiW02K%33Z5{>Z|mEFySoOkd%U
zoaHxBL_*@fR$RwKfVATK6cLg25g`%-Sc*f}udR0v82vKeXlr*<{~zs~`8Sk*|Hoyi
zC`+lRv8E^`B?eimv4=_#*(yteLCH2VmM9Z7h(4CFjD0H<8QTcuL&Dg%88gV9ZL*9Q
zlka=_oO6Hwf$#U6``qWAALhsFI@k3+*L!)r9?wUU)`-uaO)n+;!A>K~${Jm$*C|om
z#aYUj9cGOJ9wS7u{9Dwus6%$j6U152&S#{u7JftyN<;%+&6R7^KWSLb?mE}b^c1XE
zMZ;g=7=hr*tgHz|un=uJr(p-81OUC^)b|;JC#{8rPR|0ckjWl9!J0jXz$3z#E>4x6
zR6lX0(XqQI#*d$HIDF3Qivek;7ph_EaU=fkbExGek;CS`{(tuF)%)_4B<ybA_I(Io
z8p*|;iXR$o5((D9bCMAEvO8{Tagw4zasc9gmLxYP;m!h3rA6(=!ta~b*25u4aQTT}
z70oQil(R+dHRuB~*6q|kuW40Q<^f{xgGGx;QO&I&aI}LNG%Nj6NoEf>d$j&f|H#L4
zopW`XLXHDokC2sW2kb!y{tNg0M<a}6OJ86y(ql`#nxiEra-G+^_HDpo>bGKnUyN-D
z8@9S?^LY4VR-ItF*dG`Vv9Hj-9;sFHW1y{W-4y<%V_<`%j!%G;dpx1YsGh7huce>x
z((wuRM2?Xsw@AdvTZ+XMO*_x)TILaD(TVAYJ44aJa)_nejuKE;Z*;;Nq<SYM4!mIu
zg$z1TtR?Fgaw`u1FXTxnI%4C|$j<UNy%=H8vG$j0zE7KEg9gbgxwd4brwqB2CKu0I
zgLEm9o!a>4;Vj;HeQ@40?-6lu>;B`7At=7Fw0<x(DY~?k_px_!G27Y7HqR1#%=pyh
zvCQxv#iNd&k=^_oB*UJ(RAf#izWVaTrKEJ*Q8u>+FV0O438Y;rFUa|Lf>)EO$+peB
z?WhDcAJkm=nXr3|>P~4C5^fq9`W)fFr0Xdb7oYO){HOm&(4($56-WM51BnD9sq=sb
zCThO0*|P%r?RP*qgQ&d5wtFeBqy`Po*&{RANxEUaL$+xlQ^T#*zvrsL>y!)W0JF=G
z=7R9w^c(d`EC8u?zC$*rNQ$m_a1V~sSb`p+avpjEBOso|ZWh`27T?YSmHx;Nr0K?h
zCE{Ro<ZJbnCLKN15r1r@<Xf;0Q)VT!#c2}jYT+>?v{d0YolRn4_hJlfGFK)H`_?NZ
zVN#uyWswDxcJ=2L+Xl1Q=-E1?OBEN=gl=7ybJdE=xl!D%G&8$YE)7UH;>zGmJWGq!
zoZYIVa^kEGb6}$TF23#<dv2Z-gJ->`&?C@*4us{~d*5FJs2h?AyH$@@WxaEvu#3-i
zo5N@9C7z64^w-Z%4Y?Ax+{`S+0+@$iQSf`sqE<)x%flT;eG+u$3glcAn=F}~($PuW
zmh{%z_Nt<cY&ZDMvFhoQvtasX^c)y`)OdU5#Aa1`qE5g-+XO%l*^JfAGb*5eZ7%_n
z`T%@2P|;KLYgGc#9Th)A#-~5DOdO409#33ySH2!V-$c1BZ`7HOj_0xqY99}Q7pyYg
z?pD(eGdWuxP=*F;`oXK;37uvm!Ta1D4*~|XjgolVhx0ibPiwvEz*8$A_cBvKUf#H;
zz;All_-%BdNR^P`{$wxjI7>^*gzb}E!}n`>BeR<fT{~5((+e@OlqFxfM5|KcbCelB
z;NKb%$bjF}{5WTtAH3*m`WLX$h}wZF7`yXRF<+%G2KU3Ot(~)b>(;wzQoB(sNWGr2
zC%=*r;XYump>la!qJsVvX_gks=lX+Yv6UsXcL)ufa$a$+qgOYIv?|&7h*vKc)4{F!
z<osf3RgM}NF$w@Pc{aSaSV5kF2^n|XVj%7Mck>AkVEt%57^hLs)h`Ny_V~%_rKPuV
z^<baw<EVl1wB*K?_{hb6)sFf8J<G-w4}@br4V&-NnH}msQS3$59`v*j+U)dD0^qa=
zc|PaNxxyzp)ntKKjZ0r8Meu46y_+c)INbaxbH?JqQ_kHYEo~RPB?LA7{Sn71_IXs!
zT&|UC776eC;C*gl&?XluX_g6mA~{%`^UC+Wp#;?c!UR7~mBb@*^snvd4m?rKdCQp2
zSWLfr6_=*F3-e{2NT*pCEq(@Esqug>$nA4KQ1ixld*57DIRDNsgjhH6-s%qg2_pYS
zDupdnqm+dV{IqnBaZkJ0C`j=H4aN#)d4+Dg;E+P~#-!Fp)WvL2<m#^Kw<_u|$Oz?{
z6a|rpYh@u@MO9BZ#bF}$=*$8O=X?MjN^amAUZ63OA%<}=^V!y+9B)t!HEhjbwIB#N
zRLrlO1`a7P=NPXF02QBnDypK80}^Q^&wZ6M>;9<V+>CzV=X*>25uxk^<jGw}W?-hA
z^Lr<YYI2@lg<c;Ei9n%%L{Y9~a*NG6o)8^=@?F2;4-RM(Wo1ag1(YugeIU{xOut%>
zG*RKZhY|ff+I-kTkeuuOdob0Zmj8B-S}edYRvDN^7YIbOt2*}BbC26fPxtF@1H`+~
zPE`)jRR4Gw>(DK!k6~r$l}igqF);zMeH$o$;kgApdppHrsHq<H%lA+x+XgO-uJ$di
zub8SY7Lb|Wv;zphWJa);XbK)mXO(%R%{R7)U01||3b;vESKS>Tri>p8b^J`jcyxt%
zZ;!lX!?_fiJml2-YYmBrV8<Obw7w{KvjVB)q}ThfhASXQA0oOUacL^UDX^tAz6bT|
z2@4g_XM~ZP2U5IMq03jGQt!!Zi5y4QZW5b%QUI(B!bnv0l^CLNczxjXjbfbLXb)fm
zopDEmIyheyWqdo559`qLgV25rxum4Y%6j!%|4cAGMoJ(88r!Z+^T}`N$nu8dDTvkq
zrRy0HGAW${BXreT+Q73hy0?d#AI+PJ3N5fc3bv%Jerugl4H+&xuI=Dh$@`FOns|bl
zH`hoNF*V8Q9(_gx<Ay9_dUNp<tyxcpG*V~uQ@mnxy1l}P%Hi+Z_rAWGCpOEY!ruzW
zz>=~K!NHM0=eX!_AT(gsYiu@9iL6gQ=h~fg*1p^e1az+{r+})Yiu%Hr-F(VCA5ynV
z)d$;mpm2gk38<jMOn&mg@Ad%6CFK~Fm}LWHP7n65XBS7UMy_TuA~D}535B*=Q<PF9
zJ!J$J#UYOM|4pQJ0F;~@6tVx%tP~E&;DT+RtXHoOSZI-uC6P9VH*j_bU_*#xN~rAd
zl;)sZ9Sj0fDNu638si3z`4&Qtxt?Pyhq@@3Gh*t2yf9v9;psB?6SGOI7HYqhXNIoA
zfi^<mtKUk5=4SpwKz_O6v^g8UDjB|Wbjym40&uud@0%1%Mtsna?$cv?H;&m2yWa#S
zqlc&Mr4eVc5Xtg(I1cZgl7d}h+Z#G)JQ?x(VMgdfk?D>k=+Zljvi{Esb54f7to04s
zG-Cg_+E6+HBw4|z4alUFsz2rQBK94uQc*XZ15#1bXmeq%+mm%$YsL%%2W|#6wKW64
zilx67kN}5^jtv%B(mr;u5z-;e{a6dRZb--%e3sG)Ks=!e17#h0JDfI#wq6ENOs#vH
z8QF{th6$Gc^2_4CGMmGL1w6I&-QEjAP7S%pxIG*pmark+KVM~30d#6xc9Bq=7*Q~t
zTP|I<h`8$|&B@cWm%Ar9vwJCpBh1$`1)AlmA<lJF-YA|&eAVodzgwLqk!OYQ?QWLc
zAPr~-0cQK+Aemgg>20svHgxphqmP>r2{s0LAp#ohorv9+x$k&*VT_?5aec|!B_O}n
zoc60QGlTaTt7|+qS-d=O!N>DqiNG7h>lxu^#P`7OODEl*+&2$EPA05_FCJ0NoU;9D
zGRmExz6VDe$-uaXuMF!pM!y^K#itR2dYtqZ6ZnyA6z`u$O3F)GZBxyoxfSyd7_Qhp
zUgT*lEB@wUn4F@@-fx6)kuj|gnBp0^28zkb{^z-J><mhMk=w!_hi-5Xc^C`Sa@u)a
zWM40D*C&PJmp)$D=QJ?a6Jd%q@}oV^Q<ekZv~bz0i{;rzf4Pgt76-TRmZR^vgp3SM
zwVbmm^^q6f|LA@D-;S%@O_wp(msDh~+!+M5VU9TFHIC%8Zz%^3zRGc|_v_=WfIK0!
z(?UpztP5eXsML4B@3Us$DBh_B*#=PFeT`^SY?NI~fynQfDi=!N>xQT3Z9K)NsZsI6
z+Wl5-5QbduYPza$2TH$}zB|kUSGO?B_phZ*>XmH&&XSO!6(KxRT)$<1w4bK7xG>dz
z$`W)@Uw~bXs96g`v?b6&rU5bm!k>qW35;K4f!O_e<&^Jh=YOqv&WQR|o%k+>(59qE
zPvjg4^Yw9)cs>l6XPe4$O92U}8a!f5v8_GDaap}(-2SJ2`-VW084j(9_mJe!xTcIU
zf9qZpznIkadD`6*d=mO&I~b;Mi*Cf@m9y}j%Is4@8>t9h`PuU(y}LGNUAu?;wZZd$
zHOocJH0c7oHWeBqc&C$ojdK?X>$0cvXtTiPl>OsUUf58C@9_BR?fSUpo#Tz7**@|b
zR<gT12=&;1O)*0KIUnbR_g2BJ$KQzRx17*qU|&`SMVR+Eb;n-FF~zY|RkSXu2v0sp
z0JDxxusYlJ75{1XyeH95Nl^C_ZN-+WuIrI(U-a1Q$ncRz`umfe9NPK5UZPk~1u8tE
zXaDYmZn|VU>0jwd;&|aq!cQ$gG_%5`=c%MHZh?3HuKYPiFx`CULy;tu=Zrb5!Qi^%
z@?j9${?vPK`Scg&kfe;r$b7zf<f^wE^!`Jsc`8n>1Y+Q}tP{k!ASP(tT^;+0H&xP9
zO4*@W#@=>(TRktx#j^cT!`6b&^rtDNL-y$?oDAm$((Al^6x;ErnVR)z;dsZF-AA__
z%FX=)&hOfvt+&9M-iet`^N>h~P@ulp|3=vU-Gh2tJAOY3xE-ZlWID3kwGNyd87&Dt
zbcxISN}L1EfM@O9YvY^c+dZ1{8Lrq{ak#0(=^MMW+;igk&eki(@XcDw{FMVbUV*oC
qI|G+aot!%}l?Zh9{~y2K3p}r$39p&1zc|dkypf(6tmKMg*#7`b6r@W4

literal 0
HcmV?d00001


From 7bff96b8c8a085c9f946d5a90335ab862ef6974f Mon Sep 17 00:00:00 2001
From: Josh Suereth <joshuasuereth@google.com>
Date: Mon, 8 Mar 2021 13:36:21 -0500
Subject: [PATCH 07/14] Add transitions, changelog.

---
 CHANGELOG.md                       |  1 +
 specification/metrics/datamodel.md | 10 ++++++++--
 2 files changed, 9 insertions(+), 2 deletions(-)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index 67be2664bfd..faceb8d7633 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -17,6 +17,7 @@ release.
 - Rename `gcp_gke` to `gcp_kubernetes_engine` to have consistency with other
 Google products under `cloud.infrastructure_service` ([#1496](https://github.com/open-telemetry/opentelemetry-specification/pull/1496))
 - `http.url` MUST NOT contain credentials ([#1502](https://github.com/open-telemetry/opentelemetry-specification/pull/1502))
+- Adds new metric data model specification ([#1512](https://github.com/open-telemetry/opentelemetry-specification/pull/1512))
 
 ## v1.0.1 (2021-02-11)
 
diff --git a/specification/metrics/datamodel.md b/specification/metrics/datamodel.md
index 70de49dd800..200fe522e83 100644
--- a/specification/metrics/datamodel.md
+++ b/specification/metrics/datamodel.md
@@ -119,7 +119,13 @@ decisions within the metrics data model.
 
 ## Model Details
 
-TODO(jsuereth): Recap split of 3 pieces
+OpenTelemetry fragments metrics into three interacting models:
+
+- An Event model, representing how instrumentation reports metric data.
+- A TimeSeries model, representing how backends store metric data.
+- The *O*pen*T*e*L*emetry *P*rotocol (OTLP) data model representing how metrics
+  are manipulated and transmitted between the Event model and the TimeSeries
+  storage.
 
 ### Event Model
 
@@ -174,7 +180,7 @@ is no delta temporality in the timeseries model. To precisely define presence
 and absence of data requires further development of the correspondence between
 these models.
 
-### OpenTelemetry data model
+### OpenTelemetry Protocol data model
 
 The OpenTelemetry data model for metrics includes four basic point kinds, all of
 which satisfy the requirements above, meaning they define a decomposable

From 8d0db7a36a6b9f91652a11211d4146b38e36952b Mon Sep 17 00:00:00 2001
From: Josh Suereth <joshuasuereth@google.com>
Date: Wed, 10 Mar 2021 12:20:12 -0500
Subject: [PATCH 08/14] Add out-of-scope use cases.

---
 specification/metrics/datamodel.md | 15 +++++++++++++++
 1 file changed, 15 insertions(+)

diff --git a/specification/metrics/datamodel.md b/specification/metrics/datamodel.md
index 200fe522e83..5631eddcf7c 100644
--- a/specification/metrics/datamodel.md
+++ b/specification/metrics/datamodel.md
@@ -117,6 +117,21 @@ breadth of OTel metrics usage.
 These are considered the "core" use-cases used to analyze tradeoffs and design
 decisions within the metrics data model.
 
+
+### Out of Scope Use-cases
+
+The metrics data model is NOT designed to be a perfect rosetta stone of metrics.
+Here are a set of use cases that, while won't be outright unsupported, are not
+in scope for key design decisions:
+
+- Using OTLP as an intermediary format between two non-compatible formats
+  - Importing [statsd](https://github.com/statsd/statsd) => Prometheus PRW
+  - Importing [collectd](https://collectd.org/wiki/index.php/Binary_protocol#:~:text=The%20binary%20protocol%20is%20the,some%20documentation%20to%20reimplement%20it)
+    => Prometheus PRW
+  - Importing Prometheus endpoint scrape => [statsd push | collectd | opencensus]
+  - Importing OpenCensus "oca" => any non OC or OTel format
+- TODO: define others.
+
 ## Model Details
 
 OpenTelemetry fragments metrics into three interacting models:

From f157ed915ad7bbba104aa035532926d21b68ea93 Mon Sep 17 00:00:00 2001
From: Josh Suereth <Joshua.Suereth@gmail.com>
Date: Wed, 10 Mar 2021 17:21:16 +0000
Subject: [PATCH 09/14] Update specification/metrics/datamodel.md

Co-authored-by: Aaron Abbott <aaronabbott@google.com>
---
 specification/metrics/datamodel.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/specification/metrics/datamodel.md b/specification/metrics/datamodel.md
index 5631eddcf7c..7b4216b54ba 100644
--- a/specification/metrics/datamodel.md
+++ b/specification/metrics/datamodel.md
@@ -177,7 +177,7 @@ consisting of several metadata properties:
 
 - Metric name and description
 - Label set
-- Kind of point
+- Kind of point (integer, floating point, etc)
 - Unit of measurement
 
 The primary data of each timeseries are ordered (timestamp, value) points, for

From 7fd0ad2021f96885f040b0e15d673eb750c49c49 Mon Sep 17 00:00:00 2001
From: Josh Suereth <joshuasuereth@google.com>
Date: Wed, 10 Mar 2021 12:27:32 -0500
Subject: [PATCH 10/14] Updates from review.

---
 specification/metrics/datamodel.md | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/specification/metrics/datamodel.md b/specification/metrics/datamodel.md
index 7b4216b54ba..3e90273bf9e 100644
--- a/specification/metrics/datamodel.md
+++ b/specification/metrics/datamodel.md
@@ -165,10 +165,10 @@ three categories:
   instruments are primarily useful for monitoring a rate value, whereas
   non-monotonic instruments are primarily useful for monitoring a total value.
 
-In this model, the primary data are (instrument, number) points, originally
+In the Event model, the primary data are (instrument, number) points, originally
 observed in real time or on demand (for the synchronous and asynchronous cases,
 respectively). The instruments and model use-cases will be described in greater
-detail as we link this model with the other two.
+detail as we link the event model with the other two.
 
 ### Timeseries Model
 

From 8895928c60a2dcd9e995ec1e8db82c33f6e39d4f Mon Sep 17 00:00:00 2001
From: Josh Suereth <joshuasuereth@google.com>
Date: Wed, 10 Mar 2021 12:29:12 -0500
Subject: [PATCH 11/14] Fix lint issue.

---
 specification/metrics/datamodel.md | 1 -
 1 file changed, 1 deletion(-)

diff --git a/specification/metrics/datamodel.md b/specification/metrics/datamodel.md
index 3e90273bf9e..2968f03ee69 100644
--- a/specification/metrics/datamodel.md
+++ b/specification/metrics/datamodel.md
@@ -117,7 +117,6 @@ breadth of OTel metrics usage.
 These are considered the "core" use-cases used to analyze tradeoffs and design
 decisions within the metrics data model.
 
-
 ### Out of Scope Use-cases
 
 The metrics data model is NOT designed to be a perfect rosetta stone of metrics.

From 517b6544253ba9003b1864e73ca8b8e000a4cc9e Mon Sep 17 00:00:00 2001
From: Josh Suereth <Joshua.Suereth@gmail.com>
Date: Fri, 12 Mar 2021 16:49:14 +0000
Subject: [PATCH 12/14] Update specification/metrics/datamodel.md

Co-authored-by: Reiley Yang <reyang@microsoft.com>
---
 specification/metrics/datamodel.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/specification/metrics/datamodel.md b/specification/metrics/datamodel.md
index 2968f03ee69..f507ab4ea27 100644
--- a/specification/metrics/datamodel.md
+++ b/specification/metrics/datamodel.md
@@ -151,7 +151,7 @@ anchor our understanding of the OpenTelemetry data model and are divided into
 three categories:
 
 - Synchronous vs. Asynchronous. The act of calling a Metrics API in a
-  synchronous context means the application calls the SDK, typically having
+  synchronous context means the application/library calls the SDK, typically having
   associated trace context and baggage; an Asynchronous instrument is called at
   collection time, through a callback, and lacks context.
 - Adding vs. Grouping. Whereas adding instruments express a sum, grouping

From b844bc18e6369d23c3a3c18954b26cb159e8b828 Mon Sep 17 00:00:00 2001
From: Joshua MacDonald <jmacd@users.noreply.github.com>
Date: Mon, 15 Mar 2021 14:21:03 -0700
Subject: [PATCH 13/14] Update specification/metrics/datamodel.md

Co-authored-by: Reiley Yang <reyang@microsoft.com>
---
 specification/metrics/datamodel.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/specification/metrics/datamodel.md b/specification/metrics/datamodel.md
index f507ab4ea27..f40d0705050 100644
--- a/specification/metrics/datamodel.md
+++ b/specification/metrics/datamodel.md
@@ -160,7 +160,7 @@ three categories:
   to grouping instruments are generally not. Adding instrument values are always
   parts of a sum, while grouping instrument values are individual measurements.
 - Monotonic vs. Non-Monotonic. The adding instruments are categorized by whether
-  the  derivative of the quantity they express is non-negative. Monotonic
+  the derivative of the quantity they express is non-negative. Monotonic
   instruments are primarily useful for monitoring a rate value, whereas
   non-monotonic instruments are primarily useful for monitoring a total value.
 

From aef7a65411b99a27f526f69344e9df83df770d43 Mon Sep 17 00:00:00 2001
From: Josh Suereth <joshuasuereth@google.com>
Date: Tue, 16 Mar 2021 14:08:05 -0400
Subject: [PATCH 14/14] Updated from review.

---
 specification/metrics/datamodel.md | 19 +++++++++++--------
 1 file changed, 11 insertions(+), 8 deletions(-)

diff --git a/specification/metrics/datamodel.md b/specification/metrics/datamodel.md
index 2968f03ee69..69f08235ea7 100644
--- a/specification/metrics/datamodel.md
+++ b/specification/metrics/datamodel.md
@@ -99,20 +99,23 @@ breadth of OTel metrics usage.
   temporality for a stateless client, stateful server:
     - Collector re-aggregates into 60 second resolution
     - Collector converts delta to cumulative temporality
-3. A number of OTel SDKs running locally each exports 10 second resolution, each
+3. OTel SDK exports both 10 seconds resolution (e.g. CPU, request latency) and
+  15 minutes resolution (e.g. room temperature) to a single OTel Collector.
+  The collector exports streams upstream with or without aggregation.
+4. A number of OTel SDKs running locally each exports 10 second resolution, each
   reports to a single (local) OTel collector.
     - Collector re-aggregates into 60 second resolution
     - Collector re-aggregates to eliminate the identity of individual SDKs (e.g.,
       distinct `service.instance.id` values)
     - Collector outputs to an OTLP destination
-4. Pool of OTel collectors receive OTLP and export Prometheus Remote Write
+5. Pool of OTel collectors receive OTLP and export Prometheus Remote Write
     - Collector joins service discovery with metric resources
     - Collector computes “up”, staleness marker
     - Collector applies a distinct external label
-5. OTel collector receives Statsd and exports OTLP
+6. OTel collector receives Statsd and exports OTLP
     - With delta temporality: stateless collector
     - With cumulative temporality: stateful collector
-6. OTel SDK exports directly to 3P backend
+7. OTel SDK exports directly to 3P backend
 
 These are considered the "core" use-cases used to analyze tradeoffs and design
 decisions within the metrics data model.
@@ -189,10 +192,10 @@ three value types:
 This model may be viewed as an idealization of
 [Prometheus Remote Write](https://docs.google.com/document/d/1LPhVRSFkGNSuU1fBd81ulhsCPR4hkSZyyBj1SZ8fWOM/edit#heading=h.3p42p5s8n0ui).
 Like that protocol, we are additionally concerned with knowing when a point
-value is defined, as compared with being implicitly or explicitly absent. There
-is no delta temporality in the timeseries model. To precisely define presence
-and absence of data requires further development of the correspondence between
-these models.
+value is defined, as compared with being implicitly or explicitly absent. A
+metric stream of delta data points defines time-interval values, not
+point-in-time values.  To precisely define presence and absence of data requires
+further development of the correspondence between these models.
 
 ### OpenTelemetry Protocol data model