Sidebar Menu

Projects

  • Dashboard
  • Research Project
  • Milestones
  • Repository
  • Tasks
  • Time Tracking
  • Designs
  • Forum
  • Users
  • Activities

Login

  • Login
  • Webmail
  • Admin
  • Downloads
  • Research

Twitter

Posts by stumathews
Stuart Mathews
  • Home
  • Blog
  • Code
  • Running
  • Gaming
  • Research
  • About
    • Portfolio
    • Info

A strategy for systematic thinking

Details
Category: Blog
By Stuart Mathews
Stuart Mathews
19.Jan
19 January 2019
Last Updated: 20 January 2019
Hits: 2448

I've had to re-evaluate how I deliver my papers to my tutor due to the feedback I've received recently.

Particularly that my writing is not structured and analytical enough. Taking these points on board (and the penalty in my last submissions) I've put some ideas that are leaning toward a model for academic thinking.

Here are some general ideas:

  • We want to be in the process of the discovery of essence.
    • To find the real problem, which is hidden behind concrete solutions and already chosen and discriminated implementations.
  • Need to establish/generate abstractions, which are ideas that represent the essence that underlies something.
    • This is devoid of implementation details, technology or anything that can become outdated.
  • Use a work context diagram to understand the WORK now.
  • Consider the whole end-to-end process that something makes up
    • Its likely that it lives within a category (which is an abstraction)
    • Consider the effect something has on other parts or components in the process.
    • Consider the wider scope of any particular thing, put it into context within the process it lives around. Step back and consider why the component/detail exists as this might release the shackles that constrain the properties of the current components.
  • Identify solutions or decisions already made.
    • Solutions contain implementation details
    • These identify the acceptance of assumptions or choices of ideas and disregard others.
      • You can reverse engineer a solution into its essence by removing its detail, stepping back and looking at it systemically.
  • Generalizations and ideas that are not constraint to a particular constraint last/live forever.
  • The idea is to start off with a blank canvas and then add abstractions the cover the spectrum/ecosystem of subject to establish an objective, clean and transparent representation of the subject under discussion or analysis.

Good ideas 

  1. Produce a model as structure + strategy to guide your thinking and reasoning
  2. Objective measurement criteria represent an easy way to determine the "value
  3. Value of something can be broken down into 3 parts
    1. The reward for achieving
    2. The penalty for not achieving
    3. Cost of implementing
  4. Convenience, pleasure and prestige drive people in wanting things.
  5. If you’re looking to innovate, identify and remove constraints

My preliminary strategy for establishing a structured piece of academic writing currently is:

  1. The idea is to start off with a blank canvas and then add abstractions the cover the spectrum/ecosystem of subject to establish an objective, clean and transparent representation of the subject under discussion or analysis.
  2. Set the objective or purpose or claim of the writing
  3. Set the intended strategy/means how I'd like to model the claim or delivery thereof
    1. Develop a model
  4. Set the context, background briefly but accurately
  5. Decompose the strategy into higher-level themes that serve as container ideas or headings
    1. When deconstructing, consider the whole spectrum in which ideas fall
    2. Order by the determined value (Importance)
      1. Determine how to determine what is more valuable than something else - this might be a calculation or a constraint imposed upon you by the request for this writing
      2. This can be used to help balance points potentially - to ensure objective measurement
  6.  Layout the delivery of ideas within the themes determined previously
  7. Reason systematically about ideas, connecting ideas with evidence
    1. Merrit and usefulness of idea/abstraction (why is it worth considering?)
    2. Use Evidence to back ideas, logic etc
    3. Draw conclusions based on previously established content(which is already backed by evidence)
      1. Use only content already backed by evidence and logic to derive other content/reasons/ideas

 A preliminary strategy for analysing already written matter:

  1. Remember that content is within an ecosystem or spectrum of abstractions - determine it completely.
  2. Establish what are the main ideas of the content and into which component/abstraction/idea in the full spectrum does it fit
  3. Look for solutions as they have made choices to formulate the solution that disregards other choices.
  4. Piece together a context diagram that identifies the context/work from its adjacent components, identifying events.
    1. Consider the influence components have within the context diagram

 In Summary:

Start With a blank canvas and aim to fill it up with the objective representation of the subject under discussion.

Establish the full spectrum/ecosystem that represents the subject.

Establish the components of the spectrum - Work Context Diagram

Establish a model to represent your delivery.

Establish what value is to the article and order by it.

Make it easier for your reader to follow.

Reason upon interplay and merit of evidence-backed generalisations.

 

 

 

 

Zero point four

Details
Category: Blog
By Stuart Mathews
Stuart Mathews
13.Jan
13 January 2019
Last Updated: 13 January 2019
Hits: 2540
  • Impressions
  • Programming
  • Investing

I didn’t want to get up to go to the gym on Saturday morning - I’m not exactly sure why but I was unusually tired. I did go to the gym later that morning after I woke up more easily. I did my usual back and bicep training and felt quite strong on the pull-up bar.

Had to adjust my routine somewhat because there was a couple working out together who used the Smith machine for most of their workout. I’m surprised they weren’t taking selfies of each other or snapchatting. It’s the January crowd – I guess they should be gone within a couple of weeks.

There are far too many people in the gym that late in the morning on a Saturday. I’ll try not to repeat.  I’ve picked up some weight and my runs are a little slower but that’s to be expected however I’m doing pretty good otherwise. My midweek run was quite enjoyable because I just took it easy and enjoyed the process. Sometimes I can force myself to run a certain way and then the whole run is just about that. This time it was just about a comfortable momentum – but decent sub 5 min/km pace where otherwise I’d be aiming for 4’30 pace.

I’m still reading more about software engineering and professional development learning theories as part of my coursework. There is a lot of stuff there and sometimes It seems like is the only thing I’m ever doing. Sometimes it’s, well, just stuff that I’ve got to write about and I wish it was less theoretical. What I like about ‘theory’ is not what most other’s like about theory: Report writing, simulating being a project manager etc…

My tutor thinks I’m dyslexic because the presentation of my recent assignment was so mixed up, she said that she had to check the university system to see if I was officially registered as Dyslexic. I’m not, but perhaps my motivation in the last assignment made me a bit sloppy, I don’t know. I passed by the skin of my teeth. I also passed my other assignment but again, it could have been a lot better.

I’ve put it aside in last week because my interest has waned somewhat and instead, I spent the train journeys writing extensions to my investing application. I’ve added a few noticeable improvements.

For one, I’ve successfully implemented a framework to handle two classes of errors – JavaScript (like ‘function is not defined’) and server errors – like 404 and 500s. This feature, or lack thereof, while not critical has been annoying me for a while, I’ve just endured with not having it. The upshot of this work is that I don’t have to inspect the JavaScript console to find a logged error message – it goes to is very own error page – very nice. Also, errors could be missed in the console if they weren’t causing any noticeable issues in the program. It’s useful to now see your errors explicitly. Anyway, this basically entailed implementing and ErrorHandler class in Angular.

The other thing I’ve implemented is the ability to define the datatype of a custom entity, this is with the view that later on when defining a new or interpreting an existing custom entity, I’ll be able to determine if it’s a number, a string, a list etc. This makes it easier I think down the line when I want to write analysis queries that interrogate an investment’s custom entity. That query/analysis feature, however, is still currently in the clouds of my mind, forming and desperately trying to be defined.

I’ve also implemented a notification service whereby I can issue notifications throughout the applications and they then all manifest the same way – as a popup at the top of the applications that disappears after a set time-out or is acknowledged with a tap. That’s a pretty cool I think anyway.

I also finally figured out how to populate a drop-down list from enum values in angular (surprisingly not easy) as well as implement paging for some of the data views that are getting a bit too long. 

I’ve also implemented a ‘switch views’ in the investment details page that alternates viewing all the investment’s details in either a compact view (with tabs) or a ‘by headings’ view.

I have realised that my CSS is not as good as it could be and this is something, I’ll invest some time in the future. For now, bootstrap is covering my ass big time

I went to the doc the other day and got a referral to see an ENT specialist, which I’ll probably do. Apparently, I have asymmetric hearing and a difference of 40db around the 4kh frequencies range in the right ear.

I’ve written some theory about my experiences learning functional programming in C# using Language Ext specifically from a point of view of never really needing it and being asked to use it. I worked on it mostly on Friday at work. A lot of my readings and experiences about it have been from this perspective. Kinda like trying to figure out not the merits of it, but how they actually translate into actual C# and into my thought processes as a developer.

I watched Blade Runner again today.

 

CMaking good progress

Details
Category: Blog
By Stuart Mathews
Stuart Mathews
30.Dec
30 December 2018
Last Updated: 30 December 2018
Hits: 3246
  • holiday
  • C++

Since A lot of doing a lot, I've made some significant progress with my C++ library. It compiles in Windows(Win10) and in Linux(Fedora 29, and WSL but you can't really call that Linux can you? Not according to Richard Stallman!)  and theoretically, it would work fine in Mac too(But I don't have a mac. maybe I'll send it to my dad for him to try).

I decided this time to rather than use Autotools as I did with my C library, I'd embrace CMake this time. 

CMake comes with Visual Studio support and supports creating shared libraries really easily. The major motivator behind this is actually twofold. I want to reuse some of the Math I've learnt while studying 3D programming in Direct X (which is only for Windows). The other part is I'd like to do a better job of stulibc as explained in Kim Kardashians butt and other stuff. 

Like my Autotools projects, I also have a suite of tests that will run that ensures that the library doesn't regress. The cool thing about this library is that it can be consumed by any C/C++ program as I've got two interfaces to the library. I've put all the C++ code into classes and I've used extern "C" {} code to export specific C functions which behind the scenes call the C++ classes. So this actually means that I can move a lot of my C code from stulibc and put it into  Stulibcpp

I was thinking that I'd probably spend a lot of time on this but I found resources on the net that showed me how it's done and I just improvised the rest. This one resource was about integrating leap motion(I remember the first time I saw one of these was when my colleagues at AppDNA brought one in) with Unity/UnReal.

One thing I did have to tweak was in Unix you can get the address of a function in a shared library with dlsym() like this as shown here:

# define GETFUNC(funchandle, lib, funcName) (*(void**)(&funchandle) = dlsym(lib, funcName))

// and call it like this:

int (*sub)(int, int);

GETFUNC(sub, "stucpp.so", "sub");

sub(10,5)

But this wasn't working for me, it was always complaining about how it cannot implicitly convert a void* to something else. This made sense as you can't do that in C++, though existing code online seemed to be using it. Anyway, to solve the problem I found this online which suggested doing this, which works great: 

int (*sub)(int, int);
reinterpret_cast<void*&>(sub) = dlsym(lib, "sub");
std::cout << sub(10,5) << std::endl;

I also re-used some of the ideas online to write C and C++ specific tests to ensure that the library would be able to be called correctly. For example, I've got a test that calls C-Style GetProcAddress() in windows and one test that merely links to the library and uses a compiled library class.

I'm relatively happy with that progress - now what needs to be done is to finish writing the bulk of the library.

As I explained in Kim Kardashians butt and other stuff, I'd like to do something similar to my C library but this time I'll be adding more Math functions, mostly around vector and matrix manipulations. I'll probably also throw in some Geometry and maybe even some collision detection stuff later one (Diff Calculus) which hopefully will allow me to link this stuff into any new C/C++ app.

Only one more day left of my holiday and this is about as far as I'd wanted to be! 

I think the initial plan will be to slowly port over some of the data structures in my C library or re-implement them with STL and then start putting together some easy file handling interfaces for dealing with saving config files etc as well as saving records to files etc.(in a generic fashion that is). 

I've limited my C++ to ISO 1998 so until I need things like lambdas and auto, I'll leave it as such. 

In other news, I met up with an old AppDNA/Citrix colleague and we compared notes around our existing projects.

One take away was that I should invest in learning google charts and Bootstrap material design for my front-end layout, particularly when it comes to visualizations.

I was actually pleasantly surprised how appealing some of his charts on his front end were. He'd done a great job of quickly developing a .Net MVC/Razor app that calculated financial metrics around companies' financial statements. It's a bit of fresh air really because when I try to develop a nice looking chart like this Running, Coffee, Chocolate and Pizza - it takes forever to be productive...a bit like C/C++!

I'd like to incorporate a few more visualizations in my Angular front end for finance tracker. Anyway, that's another story, which is further down on the list...

Now lets get cracking!

To infinity...and beyond!

 

 

More Articles …

  1. A lot of doing a lot
  2. DirectX math, coursework and some films
  3. Kim Kardashians butt and other stuff
  4. The value of life
  5. A dangerous miracle
  6. Software change, its needs and the rise of the MD
  7. Perhaps Coincidental, Perhaps Not
  8. Running, Coffee, Chocolate and Pizza
  9. Army Curious
  10. A commentary on Leadership
  • 26
  • 27
  • 28
  • 29
  • 30
  • 31
  • 32
  • 33
  • 34
  • 35

Page 31 of 182