Talk:Project Management

Initial Discussion

• We should start eating our own dog food. --Benkoo (talk) 05:47, 10 July 2021 (UTC)

• Just to clarify on your point about Jenkins. Jenkins is about automated project workflows so that is focused on the user side of the PKC? And then kubernetes is focused on resource automation that manages the resource workflows, and the dockerized tools we employ? --Baliforester (talk) 05:53, 10 July 2021 (UTC)

• Jenkins is used for automating the overall software and content migration activities on the server-side. Users usually don't have to do much about it. Kubernetes is dedicated to ensuring all the deployed services are running in the right state. It works with docker's container standard to deploy the docker images that we deployed for certain functionalities added to MediaWiki and others through adding extensions. --Benkoo (talk) 06:03, 10 July 2021 (UTC)

• So basically we use Jenkins to build and integrate technology while kubernetes is mostly for the deployment of the technology making sure it’s reliable and secure.--Baliforester (talk) 07:48, 10 July 2021 (UTC)

• It is important for us to have a template for every process that matters. As we are aware that we have regular meeting every 2 days, we should start building correspondent templates containing key-value pairs. As more data flow and fill in a specific dictionary during our usage, a more complete and precise namespace will be constructed. Meeting Template Design. KevinTung (talk) 04:12, 11 July 2021 (UTC)
  • Input source:wechat
  • Hey I agree --Baliforester (talk) 05:50, 13 July 2021 (UTC)

• As we have an expressive system, a HUGE amount of the data(content X operational) generated along our project can be expressed. So we should try to express and it serve as the metirc of system design.--KevinTung (talk) 04:34, 11 July 2021 (UTC)
  • Input source:wechat summary of Ben's message

• The Annoki project has addressed some issue of collaboration in MW and propose some systematic solutions. We can learn from these people who have similar goal. --KevinTung (talk) 04:48, 11 July 2021 (UTC)

• I create an input-trigger section to contract our action space. In the entrance of our system, we decide our action based on the trigger condition. The input-trigger namespace can be futher constructed. --KevinTung (talk) 05:00, 11 July 2021 (UTC)

• Ben: No accountability for neither one of us has been considered and planned. --KevinTung (talk) 07:20, 17 July 2021 (UTC)
  • Accountability vs Responsibility

• Ben: The structure of relations in all actions, and all resources, should and must be connected using "Category" tags, and this insight was already very well stated in earlier conversations. Category Tree provides a way to manage the hierarchy of concepts. I also explicitly stated that all Inputs and Outputs are resources, and I even documented a list of resources in the Logicl Model's Template Links. --KevinTung (talk) 07:54, 17 July 2021 (UTC)

• How do we describe the management for each resources in a high-level?
  • How do we define "Resource Management" ? --KevinTung (talk) 17:11, 18 July 2021 (UTC)

• Ben: All data dictionaries must follow the ideal of being Sound, Precise, and Terminable. --KevinTung (talk) 05:06, 19 July 2021 (UTC)
  • Criteria:Soundness
  • Criteria:Precision
  • Criteria:Terminable
  • BOOK:Static analysis by abstract interpretation of functional temporal properties of programs
  • The notion of soundness and completeness is applied as attributes of logic system. Precision is used on logical reasoning and other domains. Termination is the most prominent liveness property in a program. There are also safety and liveness in abstract interpretation. How do they generalize to our theory of UDA?

Ron: XLP & Rare Earth Metals consolidated initial literature search (>90 citations from a set of ~850 refs): http://chemconnections.org/XLP/SciFi-Reference_09_02_2021_123540.pdf

Logic Model workflow design

• We should design forms to fill out the outputs and then forms can generate timelines and necessary inputs based on the phases needed to fulfill the outputs needs. So we work backwards no forward --Baliforester (talk) 11:05, 11 July 2021 (UTC)

Basic question of writing a Logic Model

• If there are multiple Processes between Input and Output, how should I specify the Logic Model? A set of Goal, Context, Success Criteria, Boundary conditions with a list of Input, Output, Process? --KevinTung (talk) 07:27, 20 July 2021 (UTC)
• In PKC Workflow project, there are lots of types of resources as Input and Output of the project. However, they are types instead of instances. In PKC specification, we should write the resource instance instead of types, right? For example, current PKC project uses Jenkins and TLA+ which are automation software and formal analysis software. --KevinTung (talk) 07:37, 20 July 2021 (UTC)
• How do we model the system using Logic model when some parts are stateful or recursive? Note: TLA framework use a loose but expressive expression which is the state transformation.

Universal Project Abstraction

• This is about creating a unifying understanding of the project scope and direction --Baliforester (talk) 03:48, 13 July 2021 (UTC)

Logic Model Context

老师想问一个问题: context原本是定义为causal relationship在time space的映射，但若观察一个普遍的创造性工作（比如一篇论文或项目），它的context都是以backgrond来表达，这除了包含时空，也包含了与其他事物的关系，以及推理，来说明这个purpose的Why。 所以我在想，是不是可以映射到除了时空以外的space？那我们应该怎么去表达？ --KevinTung (talk) 12:46, 30 August 2021 (UTC) 没有时空背景的概念，例如抽象的数学概念，为了能被联系到其他的数学知识体系，一定是在一个严谨的命名空间之中，而这些命名空间，基本上充满了人名 、地名、或是机构/现象的名称，所以，天然地跟时空背景发生了关联性。换言之，没有什么计算机制可以跟时空完全分离的。 (Ben)