Skip to content

Latest commit

 

History

History
78 lines (33 loc) · 4.58 KB

File metadata and controls

78 lines (33 loc) · 4.58 KB

Meeks Conjectures:

Conjecture:
noun
"an opinion or conclusion formed on the basis of incomplete information."

Meeks Software Architecture Conjectures

  1. The source code may (or may not) be a full implementation of the desired capability needed by the business - but is more likely just an approximation (constrained by permitted time, allocated budget, and available skills/talent of the team involved). Therefore it should not be confused with the actual or desired (or envisioned) design - that may require multiple years to achieve - of which the current source code may only reflect a partial (and incomplete, or inaccurate) representation.

  2. For any non-trivial system, its behavior is probably not deterministic (as you may erroneously believe), but is more likely - stochastic.

  3. Distributed Architecture is hard...and the complexity of the web of external dependencies is far beyond the ability of most teams to comprehend...

  4. Assuming an ∞ budget - many fail to recognize that the business has little tolerance for ∞ rework.

  5. "Change Blast Radius" is an important factor to consider when defending against the impetus to "code first, design later".

  6. "Unintended consequences are inescapable".

Meeks Enterprise Architecture Conjectures

  1. If you build an Enterprise Architecture practice along the lines of an "Indy 500 Pit Stop service crew", you'll have a better chance of providing the business with an "acceleration factor" - and achieving an agile architecture operational model, than if you are still working to implement EA practices along the lines of a construction permitting process.
  1. Much goodness would ensue, if we just got rid of the name EA.

  2. More enterprise projects fail due to an excessive degree of optimisim - than skipticism and pessimism.

  3. If the actual data, from every major business that has attempted to follow the Agile Industrial Complex's prescriptive guidance were published and made available for aggreation and analysis - the entire edifice would immediately implode and spontaneously combust.

  4. At some level of scale [n], very large organizations find benefit in using longer-term roadmaps to align planning of resources (people, time, money) - to coordinate development efforts (larger than [x]). This is often even more critical - when the organizational units are distributed geographically, have greater autonomy (e.g., budget planning), and may also include required external coordination with [m] external partners.

Meeks Tool Conjectures:

  1. Some of the most expensive mistakes you will ever make - will be when you try to perform a task without the proper tool. This will be especially true when you have a one-off task - and think you can get by using the incorrect tool.

  2. Selecting a tool based on an industry analyst magical report - without actually doing a proof-of-concept evaluation - will usually lead to failure of your initiative....and massive rework.

  3. There is rarely a direct correlation between the price and value of a tool. However, there is a strong correlation between a poor tool choice - and its cost.

  4. Just because you can build your own tool - doesn't necessarily mean you should.

  5. Just because there are suitable tools for sale - doesn't mean that you shouldn't build your own.

Meeks 2024 AI Conjectures:

2024-AI-01: "Most business people driving enterprise-wide GenAI adoption in their company - are unqualified to have an opinion on AI."

2024-AI-02: "In ~18-24 months, 90% of those GenAI intiatives will be shutdown - and those same GenAI promoters in your company, will be gone (voluntarily, or involuntarily)."

Meeks 2025 AI Conjectures:

2025-AI-01: "As the proliferation of AI Agents increases within an enterprise - the resulting explosion in IT costs - will drive enterprises to return to on-premise computing."

2025-AI-02: "All of the ornate work to craft guardrails, prompt guidance, etc. - can be undone by a non-visible exception - such as exceeding the maximum allowed context storage limit. And thus, what you might have believed to work / be true yesterday - may not be true tomorrow - and you will likely have no indicators of the breaking changes."