Autosave: 2024-10-22 16:25:49

neuron/6675f329-3b33-497e-accb-2333f4cb63df/385af4b4_43ea_4a5f_85a3_5791d95109aa.md

@@ -1,17 +0,0 @@
----
-tags: [fleeting, networks, ARPANET, packet-switching]
-created: Tuesday, October 22, 2024
----
-
-# 385af4b4-43ea-4a5f-85a3-5791d95109aa
-
-## _Paul Baran's concept of a distributed network and influence on architecture of the ARPANET_
-
-Baran worked for RAND not ARPA and his work predated the commencement of the
-ARPANET. However his model of a distributed network was highly influential on
-the design of the ARPANET. He served as a consultant to ARPA during the initial
-construction of the ARPANET.
-
-Unlike ARPA, Baran was concerned with command and control and how a
-communication network could be resiliant enough to withstand nuclear conflict.
-He analysed the weaknesses in AT&T's telephone network.

neuron/8e6dee36-9f5f-47c8-b9f6-5160559ef056/385af4b4_43ea_4a5f_85a3_5791d95109aa.md

@@ -9,8 +9,8 @@ created: Tuesday, October 22, 2024
 
 Baran worked for RAND not ARPA and his work predated the commencement of the
 ARPANET. However his model of a distributed network was highly influential on
-the design of the ARPANET. He served as a consultant to ARPA during the initial
-construction of the ARPANET.
+the design of the ARPANET and served as a consultant during the initial
+construction.
 
 Unlike ARPA, Baran was concerned with command and control and how a
 communication network could be resiliant enough to withstand nuclear conflict.

neuron/8e6dee36-9f5f-47c8-b9f6-5160559ef056/index.md

@@ -11,25 +11,24 @@ is a [Zettelkasten](https://en.wikipedia.org/wiki/Zettelkasten) work in progress
 comprising notes from my self-directed study of software engineering and
 computer science.
  
-**Build ID:** 6675f329-3b33-497e-accb-2333f4cb63df
+**Build ID:** 8e6dee36-9f5f-47c8-b9f6-5160559ef056
 
-**Published:** Tue 22 Oct 2024 08:25:11
+**Published:** Tue 22 Oct 2024 16:25:47
 
 ### Recent edits 
 
-- [[385af4b4-43ea-4a5f-85a3-5791d95109aa]] 
-- [[861cc26e_5f59_4615_821d_fe3ee5e1014d]] 
 - [[385af4b4_43ea_4a5f_85a3_5791d95109aa]] 
+- [[861cc26e_5f59_4615_821d_fe3ee5e1014d]] 
 - [[cfbef1c4_95ea_4fd1_85b8_536492e45e31]] 
 - [[Compile_Python_app_to_single_executable]] 
 - [[VirtualMemory]] 
 - [[Zero_property_of_multiplication]] 
 - [[Working_with_CSVs_in_Python]] 
+- [[Variables_and_data_types_in_Bash]] 
 
 
-### All notes (465) 
+### All notes (464) 
 
-- [[385af4b4-43ea-4a5f-85a3-5791d95109aa]] 
 - [[385af4b4_43ea_4a5f_85a3_5791d95109aa]] 
 - [[861cc26e_5f59_4615_821d_fe3ee5e1014d]] 
 - [[API_Gateway]] 

zk/385af4b4-43ea-4a5f-85a3-5791d95109aa.md

@@ -1,64 +0,0 @@
----
-tags: [fleeting, networks, ARPANET, packet-switching]
-created: Tuesday, October 22, 2024
----
-
-# 385af4b4-43ea-4a5f-85a3-5791d95109aa
-
-### _Paul Baran's concept of a distributed network and influence on architecture of the ARPANET_
-
-Baran worked for RAND not ARPA and his work predated the commencement of the
-ARPANET. However his model of a distributed network was highly influential on
-the design of the ARPANET. He served as a consultant to ARPA during the initial
-construction of the ARPANET.
-
-Unlike ARPA, Baran was concerned with command and control and how a
-communication network could be resiliant enough to withstand nuclear conflict.
-He analysed the weaknesses in AT&T's telephone network and proposed a more
-robust and resiliant network model.
-
-If we think of the nodes of the phone network as being loci for the transferral
-of data, the phone network was highly centralised. The message would go from its
-transmitter to a central switch which would route it to the intended receiver:
-
-// Diagram
-
-The problem is obvious: if the central switch went down, all data transferance
-stop.
-
-In fact this is an over-simplification. The phone network at the time was
-actually decentralised:
-
-// Diagram
-
-This system is an improvement because each region has a switch and not all nodes
-are dependent on the central switch at all times. If the central switch goes
-down, the peripheral switches can still manage connections within their regions.
-It just means that trans-regional messages are not longer possible. Conversely,
-if one regional switch goes down, the other regions are not affected and can
-communicate between themselves and each other. So there is greater resiliance
-but still points of failure - at the central and regional switches.
-
-Baran conceived a _distributed_ network as his model for the most resiliant
-communications network. Instead of centralised or decentralised switches, nodes
-would connect to eachother in aggregate:
-
-// Diagram
-
-Under this system, every node can eventually access every other node through
-intermediary nodes, without going through a main switch that establishes the
-connection.
-
-This is far more resiliant than the previous two network types because if a
-single node or multiple nodes go down, the message can be re-routed through
-others.
-
-Furthermore, by distributing the connections between nodes, the overall transfer
-of data can be more efficient as there isn't a bottleneck of centralised
-switches that can block up at times of high network activity. If one pathway is
-blocked, another route can be taken.
-
-The distributed model also opens up the opportunity to optimise transmission.
-For a given node in the network, it becomes possible to find the shortest route
-to another node. This would be very important in designing the IMP nodes of the
-ARPANET.

zk/385af4b4_43ea_4a5f_85a3_5791d95109aa.md

@@ -5,13 +5,60 @@ created: Tuesday, October 22, 2024
 
 # 385af4b4-43ea-4a5f-85a3-5791d95109aa
 
-## _Paul Baran's concept of a distributed network and influence on architecture of the ARPANET_
+### _Paul Baran's concept of a distributed network and influence on architecture of the ARPANET_
 
 Baran worked for RAND not ARPA and his work predated the commencement of the
 ARPANET. However his model of a distributed network was highly influential on
-the design of the ARPANET. He served as a consultant to ARPA during the initial
-construction of the ARPANET.
+the design of the ARPANET and served as a consultant during the initial
+construction.
 
 Unlike ARPA, Baran was concerned with command and control and how a
 communication network could be resiliant enough to withstand nuclear conflict.
-He analysed the weaknesses in AT&T's telephone network.
+He analysed the weaknesses in AT&T's telephone network and proposed a more
+robust and resiliant network model.
+
+If we think of the nodes of the phone network as being loci for the transferral
+of data, the phone network was highly centralised. The message would go from its
+transmitter to a central switch which would route it to the intended receiver:
+
+// Diagram
+
+The problem is obvious: if the central switch went down, all data transferance
+stop.
+
+In fact this is an over-simplification. The phone network at the time was
+actually decentralised:
+
+// Diagram
+
+This system is an improvement because each region has a switch and not all nodes
+are dependent on the central switch at all times. If the central switch goes
+down, the peripheral switches can still manage connections within their regions.
+It just means that trans-regional messages are not longer possible. Conversely,
+if one regional switch goes down, the other regions are not affected and can
+communicate between themselves and each other. So there is greater resiliance
+but still points of failure - at the central and regional switches.
+
+Baran conceived a _distributed_ network as his model for the most resiliant
+communications network. Instead of centralised or decentralised switches, nodes
+would connect to eachother in aggregate:
+
+// Diagram
+
+Under this system, every node can eventually access every other node through
+intermediary nodes, without going through a main switch that establishes the
+connection.
+
+This is far more resiliant than the previous two network types because if a
+single node or multiple nodes go down, the message can be re-routed through
+others.
+
+Furthermore, by distributing the connections between nodes, the overall transfer
+of data can be more efficient as there isn't a bottleneck of centralised
+switches that can block up at times of high network activity. If one pathway is
+blocked, another route can be taken.
+
+The distributed model also opens up the opportunity to optimise transmission.
+For a given node in the network, it becomes possible to find the shortest route
+to another node. This would be very important in designing the IMP nodes of the
+ARPANET.