13 July, 2020

King

Who Is A King?

Head is the person at top of an organization. If the organization is a state, then the person is Head of State; if the organization is a department, then the person is Head of Department; and so on.

A special type of head is a king. He must have following properties:

He must be able to make organization-wide decisions on his own. If he need vote of members or approval of an executive committee for examples then he is not a king; he is a president. Its not that he should take unilateral decisions, its that he must be able to, because that power is needed in emergency and also when due to whatever reason the underlings are unable to decide on their own and the organization is going in paralysis.
 
He had delegated authority to underlings. He dont try to make all major decisions himself;  if he do then he is a dictator. Dictators are bad because they order around, not listen to and not confer to anybody beyond their immediate circle of around six people. This result in nobody taking ownership, both because they are too disheartened to and because they are no longer capable of even if they were capable before. A king dont work that way. A king make people incharge and that goes down to hundreds or thousands of individuals he personally put in charge of various parts of government. Dictator also make people incharge but those people can be counted on atmost two hands. As a result each one of those become too powerful to be replaced and too overburdened to handle his chunk of government.

A king, relatively speaking, keep very little for himself. He may appear to be getting huge amounts from treasury but he dont keep it for himself, he give almost all of it away to his officials and even to general public. A king, by definition, have to be generous; both in delegating power and in distributing wealth.  

Why Is A King Needed?
 
Kingship is the sweet spot where a head keep executives in check while at the same time treat them well enough that they happily take ownership. 
 

The Second Check

Why Is It Needed?

Second check is needed, obviously, when first check is not enough.

How Many Checks Are There?

There are only two checks.

What Is The Second Check?

Second check is check on method, its the how of an action.

What Is The First Check?

First check is check on intention, its the why of an action.
 

08 June, 2020

When To Concentrate and When To Spread

Concentrate or Spread:

    the forces
        _ Attention
        _ Men
        _ Equipment
        _ Materials.

Forces, not force. A combination is always needed, to do anything.

Concentrate when you decide to finish off. Spread when you decide to contain.

This is for all 3 levels of action (strategy, operations, tactics).

At Strategic Level


Decide to go to war, or to not go to war. Dont attempt to do both.

As an entity you are at war or you are not at war.

You allocate your resources differently when you are at war as compared to when you are not. When you are at war, allocate considerably more resources for war effort than for any other purpose. The other purposes are maintenance, reserves, growth and luxury.

This, putting considerably more in one thing than other things, is what concentration is. Note the word "considerable". Obviously when you put your resources in multiple things some of them get more than others and one of them may get highest. Its not concentration though. Concentration is when one thing get considerably more than other things, like double or more, than its usual share. Usual share, as in non war time; its when you are just guarding your assets, not defending your assets against enemy attacks, and not doing your own attacks on enemy assets.   

04 February, 2020

Side Paths, Complications and Short-Term Memory

        Each side-path you take from your main path, results in adding complication; but only when you intent to later come back to your main path and for that need to set aside precious little short-term memory in your brain for the necessary shift back.


There are ofcourse 2 ways to prevent this addition of complication (as should be deducted from the last line above):
    First is to not come back to the main path once the side-path is ran its course. The so-called side-path become the main path as soon as you enter it, and if you later move to another path that path become your main path. Since you never intent to move back you do not set-aside short-term memory to remember where to move back to, therefore no complication is added.
    Second way is when you do intent to move back but not have to set aside short-term memory, instead its long-term memory where information is stored. This switch from short-term memory to long-term memory dont come instantly, it needs repeatition - if you do a thing for a while your long-term memory take over and your short-term memory is left spare, thus complication is reduced.

       Note that complication increase only when you use more short-term memory, and reduce only when you use less short-term memory.

Short-Term Memory
    Short-Term Memory is very fast, almost instanteneous; its like when it looks like you already know the answer, the answer is at your "finger-tips" and requires no conscious thinking at your part.
    Short-Term Memory is fast because its very little - not a lot can be saved in it. Since its very little no conscious linking is required to recall it and therefore you appear not thinking when recalling it.

Long-Term Memory
    Long-Term Memory is opposite of Short-Term Memory. Its slow but vast. Since its vast it requires linking and thus thinking. Since it requires linking it requires repeatition of information to be saved in it.    

What Goes In Short-Term Memory
    Its not just information about previous-path-to-move-back-to that goes in Short-Term Memory. The other thing that is stored in it is values. These 2 are the only types of things that goes in Short-Term Memory: address of values, values.
      
        Remember that wherever there is a division, at top-most level there is always division in 2 parts. In data its values and addresses-of-values.

How To Reduce Complication       
    [Consider this as your take-home lesson from this article]

    Since complication increase only when usage of short-term memory increase and reduce only when usage of short-term memory reduce, the only way to reduce complication is to reduce usage of short-term memory.

    Since both values and address-of-values goes in short-term memory reducing either reduce  complication. Following are some tricks:
  • Dont try to remember too many things at a time (reduce number of values in short-term memory). Make each process you use require as little values as can be managed, or make groups of values so at top level you not have to remember too many things at once but dont make more than 1 level of groups. Note that remembering 3 things at once when there is no grouping is already hitting limits of human brain, and with grouping you can increase that to 6 values at most.
  •  Dont go into another side-path (another level of indirection) when you are already inside a side-path. Once you are done processing the side-path you are already in, go back to main path and from there move to the next side-path. Make your processing structure accordingly. This is to follow only when you are not already doing compound side-pathing for a while, because if you are used to this i.e. have practise then its your long-term memory that is storing values - not short-term memory so number of address-of-values is not a concern.
  • Use external memory if you must absolutely have to go inside deeper and deeper levels of side-paths and/or absolutely must have to remember more than 6 data-values at once in short-term memory. Use pen-and-paper or use software, but using pen-and-paper is usually much less complicated because most of the people are used to it much more than they are used to saving short-term stuff in software notes.
  • Utilize another person. If work is really too much complicated then utilize services of another human being. Delegate tasks.

17 January, 2020

Change Has To End



[Following applies to all systems, and also to all phases of a system (be it design-phase or operational-phase), and even to all sub-phases of a system (be it discovery and laboratory-experimental sub-phases within design phase or young or matured sub-phases of operational-phase etc)]

A change in a system is needed only in either its Functionality (what a system do) or in its Mechanism (how a system works).

If its functionality that has to be changed then its either to expand the system (by adding more features to cover more environmental or input conditions) or to contract the system (by reducing number of features to optimize for a cost or output variable) (cost variable may be time or money or human effort or material etc, output variable may be effectiveness or availability etc of the system).

If its mechanism that has to be changed then its either a change in level of technology (upgradation or downgradation) or a change to a parallel level of technology (side-stepping instead of moving up or down, to better match resources available) (sometimes you do have to downgrade technology level because newer technology is too complex and thus have too many requirements).

A change is a process i.e. it has steps. The steps, by definition, have to occur in sequence (not in parallel).

(Multiple things may and should happen at parallel, but its still just one step).

First step in the process is a plan. Its a meta-step because in this you decide what to do in remaining steps. You also decide when to do those whats.

Define scope and stick to it or make only minor changes. In scope decide what (parts of system) to change.

(During implementation of plan, minor changes in scope is ok, may even be necessary because ofcourse you cannot plan for everything, but if you find yourself touching entire classes of areas you never thought of making changes into then you need a new plan and new implementation, judiciously undo some or all of the changes made so far and repeat the entire exercise).

Second and last part of planning is to make a sequence. Having already decided what to change in first part now is time to decide when to make those changes. Its here where you look at dependencies and layers in system. 

Your plan has to be in written form, for it to be a plan.

Dont be afraid to make changes in majority of parts in a system (be it functionality or mechanisms). Different sequences of changes in a system should be very away from each other in time. Nobody like frequent and seemingly never-ending series of changes. At end of each sequence of change a system should be well-suited to situation at hand as well as to foreseeable situations in future.

(A situation is a set of all environmental and input conditions at a time).  

A system is a set of interacting parts. If parts dont interact then its not a system. By definition, all functionalities as well as mechanisms of a system interact with each other. Therefore, when you decide change in some parts of system do take into account intended and not-intended effects on other parts of system, do consider your scope very carefully.

[If you can remember only one thing from this post, be it neither that parts of a system must be interacting nor that change in a system is only in functionality and mechanism, make it that all necessary changes in a system must happen in one sequence. Let all sequences of changes stand on their own, i.e. after each sequence system work exactly as it should be and not depend on further changes to handle current or foreseeable situations. In other words, each change must have an end]