Think Week - 2024 continued

 Can't really call this think week 2025 as that is next December lol... so here goes think week 2024 - continued

I found would could amount to the "best tool ever"  at least for uncluttering a local setup. 

The kind load balancer is amazing:

https://kind.sigs.k8s.io/docs/user/loadbalancer

And it just so happens to be built on top of one of my other favorite products i never really got to use in anger.  Envoy

This makes exposing services in the local address space effortless.  Just start the binary... and change services to type LoadBalancer then use the configured port to access from OUTSIDE the cluster... 

Think week 2024

 Its been a busy year, I am able to practice on my local course that I created for GSPro in my simulator now.  It's really helping my yardages.

As for think week.  ChatGPT has made developing prototypes much easier.  Instead of hunting for tutorials that do "almost" what you need.  ChatGPT can come close to generating a complete customized tutorial for your use case.

In light of this, I have been wanting to create a prototype that would demonstrate:

1. A system that could take arbitrary json payloads and trigger pipelines to perform automation work based on payload content
2. A front end for such a system could be developed in Retool

I plan to publish Github Repo when completed and will update this post - here are some screenshots of the configuration

First Completed Retool App:

Think week 2023 (actually think weekends :))

In keeping with the tradition of "Think Week" during my holiday vacation, I am preparing to make progress on my GCPro golf course design.  Note: I am finding it easier to be creative when not tied into constraints of using Think Week for work related subjects.  In my previous iterations, I never knew if any of my efforts were actually going to be used.  This time it's different and I am finding it easier to be productive.  Think week has been changed forever!  I think this is more in tune with what Bill Gates might have had in mind actually we he came up with the idea.  Not that previous think weeks didn't help me, just being 100% in control of my destiny here is definitely leading to more productivity and energy.

A few private repos have been created:

• OPCD Blender Buddy:    Plan is to automate bunker modifications to create some variability in the edges.  Working of example from C.E.Simgolf and trying to make basic modifications quicker for your medium level courses... Not going to replicate full mastery from C.E but hoping to get a "light" version that at least is quick and easy to make bunkers look a bit more natural
  
  
• UnityAction:  Plan is to create a scoreboard to pull data from Firebase DB that will be populated from apps created in the FlutterAction project
  
  
• Flutter Action:  Plan is to create flutter application that will enable score entry from Windows, Chrome, Android and iPhone.  Dedicated to the memory of Mr. K - my friend and former employer who introduced me and subsequently all of my friends to magic of "1 down automatic press".  This is truly the "Cadillac" of golf betting similar what Texas Holdem represents in poker.
  
  
• Shader:  After sending my first finished holes through the OPCD process and planting trees, I realized that getting colors / textures correct is going to be a major part of the process.  This has let me into researching Unity Shaders and diving into Materials Creation
• a 3D image is not just one file, but a composition of several file that the shader will then use to create a lifelike representation.
  
• The learning curve here is extremely steep but it's fascinating stuff.  
• I did make a breakthrough in discovering this free Materialize product: https://boundingboxsoftware.com/materialize/
  Not only does this software make creation of the different maps that are needed for 3D rendering, it goes a long way in showing what the difference is between the maps themselves.  
  
  
  
  
  

              

• 
  

Golf Course Design

 Adding a new hobby.  Looking forward to getting my Golf Simulator setup running, so I am creating a local course I play so that I can practice.

First attempt at using tools like Unity, Inkscape, and  Blender.  

It's a good relaxer from coding helps to reset and recharge.

Below is just the first few holes in Blender.  

Finally took another look at MPS after years in Eclipse Ecosystem

 In 2007, i attended a No Fluff Just Stuff conference where Neal Ford did a presentation on Domain Specific Modeling.  I was mesmerized.  From that day, I decided to learn what I could about this discipline in an attempt to raise my levels of power and productivity.  

Anyway, for some reason, I decided that I would try to use the Eclipse Modeling tools for DSL generation instead of MPS.  It may have been because I had tried doing Eclipse RCP and EMF as early as 2003 and the eclipse ecosystem seemed like the place to be.

Eclipse was an exciting ecosystem for quite some time.  The guys at Itemis and Sven Efftinge developed XText and XBase -- it was great.! Eventually though, I moved off Eclipse as IDE to IntelliJ and by then working in Eclipse seemed like a forked effort.  I wrote some integrations with projects like Papyrus but documentation was always hard to come by.

15 years pass and I come across this post from Markus Volter

https://www.youtube.com/watch?v=R6uoaxVV_2s

Amazing!  AND there is literally tons of documentation and videos out there on this!

This actually makes quite a lot of sense.  ChatGPT should be able to learn a simpler DSL language vs general purpose language.  Which leads to interesting equations

Developer -->. Force Multiplier 1 (DSL) ---> Force Multiplier 2 (ChatGPT). 

In other words, if you can first get DSL to write code for you (Code Generator approach) and then get ChatGPT to write code that writes code, it should prove to have a positive effect on productivity.

More to come!..   Kind of wish i had followed Neal into the MPS world but I guess I finally got here.

Update 7/22/2023 - I would highly suggest taking the JetBrains MPS for the Impatient Course.  Mind blowing how much is packed into this product.

https://cogniterra.org/course/28

Argo CI CD for PRs using Tekton for testing

 The idea is to use both Argo and Tekton for testing PR with e2e tests.  

I will start on Ubuntu and then try to get it running on a Mac Silicon machine.  Note above Bitbucket has been replaced by GitEA

Install requirements

kubectl
kind
postgres client
PostgreSQL on Kubernetes
Argo on Kubernetes
GitEA on Kubernetes

PostgreSQL install (for Bitbucket) :  https://phoenixnap.com/kb/postgresql-kubernetes

Argo install:  https://shashanksrivastava.medium.com/install-configure-argo-cd-on-kind-kubernetes-cluster-f0fee69e5ac4

Both of the above installations are simple via Helm charts.

 

In terminal with psql client:

postgres=# 

Install DBeaver database GUI

sudo snap install dbeaver.ce

Create gitea database

Install GitEA from Helm Chart

https://gitea.com/gitea/helm-chart/

kubectl create secret generic gitea-postgres-secret \

  --from-literal=POSTGRES_USER=<postgres-username> \

  --from-literal=POSTGRES_PASSWORD=<postgres-password> \

  --from-literal=POSTGRES_HOST=<postgres-host> \

  --from-literal=POSTGRES_PORT=<postgres-port> \

  --from-literal=POSTGRES_DB=<postgres-database>

persistence:

  enabled: true

  existingClaim: <existing-pvc-name>

database:

  type: postgres

  host: <postgres-host>

  port: <postgres-port>

  name: <postgres-database>

  user: <postgres-username>

  password: <postgres-password>

helm repo add gitea-charts https://dl.gitea.io/charts/

helm install gitea gitea-charts/gitea --version <chart-version> -f gitea-values.yaml

Create a user in GitEA with a testRepo

Pixie

 

Pixie seems be pretty easy to install on minikube in ubuntu.   

https://help.ubuntu.com/community/KVM/Installation

https://github.com/kubernetes/minikube/issues/7677

https://docs.pixielabs.ai/