Mythical Man-Month Analysis

--Originally published at That Class Blog

Well. What I’ve learned from this book, in fact, it’s stated from the start (I don’t remember at which point exactly, but I think it was in the beginning chapters); It’s that there isn’t a thing that causes problems and delays, there isn’t a specific thing that the book, the teacher or anybody else can tell you and warn you about at the moment when you become a software manager, or project owner, or developer.

Things just happen. A lot of them. At the start it’s just something simple, maybe something that no one thought that it could evolve to something else. Then another thing appears, now you notice but you don’t react immediately, and then another, and another, while the progress keeps getting slower and slower. And then the morale gets lower, and the performance too, and the client gets mad.

It’s a sad and stressful path. And there really isn’t something that you could know it will happen in every project. But you can try to prevent, everything, and make plans for every scenario if it comes to be, even if at the end the project goes very well.

And this is what the rest of the book is about. What is to really manage a software project and everything you can do when assuming such position. Either as a manager for the architects or a manager for the implementers, or general manager.

A basic concept for a good project to try to get to the end is communication. There is no chapter in the book where this capacity isn’t mentioned. It’s very important. At first level, there needs to be a very good communication between the architects and the implementers’ representatives. This is essential because the implementer will try to keep the architect from the skies and the architect can explain to the builders exactly what he wants. Then, problems will arise at the moment where an implementer has a doubt and doesn’t communicate it to an architect, and decides to just make assumptions. This only causes a slippery slope, where if this assumption causes a malfunction or a undesired quality of the project, a full redo must be done. Full communication should be present when talking to the clients, because after all, even if the architects are the ones who tell how they see the project, the client is the one who will have the final word. The don’t need to see the white box of things, only the black box, and if they don’t like it, everyone in the project is in problems.

Now, not every part of the communication has to verbal, in fact, it shouldn’t be. There is something important in writing things on paper. It will force the top-level architects, the implementers’ leaders and the implementers themselves to make micro-decisions at that moment. Sometimes just talking about it provides nothing useful (And if the person is like me, this happens a lot). Writing things out forces the author to state his ideas clearly. For the architects, it’s important to write on paper the specifications of the project. And those should take as many sheets of paper as necessary. Precision and lots of details are needed, even if the document becomes boring.

In the other hand, the main architect (Who is the main surgeon, as stated in the proposal for team organization in chapter 3) can’t deal with all the administrative work. Budgets, schedules, space allocation, and other organizational information should always be documented by the architect’s right hand. And why is this important? If most of the staff can work without knowing this information. Well, it’s important to not just tell on a daily basis what should the implementers do. They tend to feel more secure and have a better performance if they know the full work plan. Where are they working? How much time they have left? What is the course of action if some requirement isn’t meet on the schedule? It’s important the full staff is aware of all these questions and answers.

And finally, what should the base developer write? Like everyone else, everything. His doubts, his progress, the bugs he has found, the bugs he has solved and how were solved. This log will serve for the future because we can almost assume that those questions weren’t thought only by on developer, but by many of them, or almost everyone, especially if an architect hasn’t made a good job describing his design.

Who would have guessed that at the end, communication is the basis for a project not to fail, as it is to solve most or all of the problems in life? Always ask, never assume. Talk in the proper medium and style as corresponds to your position, talk with respect, but most importantly talk with precision and details. It will be easier for the other person to provide you with a solution to your problem, or answer your question.

Oh, and another thing I like a lot about the book is the proposal for team arrangement. Instead of having to deal with a lot of programmers that are currently working for the project (And them having to deal with each other, and a unique leader -you-), let’s have micro teams. Where each team will have a surgeon and a copilot, and they and only them will work directly with the code (In fact, the copilot can suggest changes and offers his opinion, but he doesn’t make any decision) and the rest of programmers can work on other things. There will be also a chief surgeon (Chief architect) who will coordinate with the rest of surgeons. This makes the work related to the chain of command easier. And also for each developer, because know their specific role in all the frame.

I must say that I was planning on giving an analysis per chapter, but time wasn’t enough. So I instead preferred to give this more general opinion of the book.

You can find some specific posts I did make here.

 


Mythical Man-Month Analysis

Simpson & Romberg

--Originally published at That Class Blog

Por: Arturo Fornés A01227071 y Miguel Montoya A01226045

Nota: Tanto en el diagrama de flujo, como en el código y los ejemplos, utilizaremos el método de Simpson 1/3 múltiple.

Antecedentes

Hay varios métodos para calcular integrales definidas donde se hace una aproximación por cada intervalo, como Riemann desde la izquierda y derecha, o por método de trapecios. Existen dos métodos más complejos que obtienen una mejor aproximación, estos son los métodos de Simpson (Trataremos especialmente el de 1/3 múltiple) y el e Romberg.

Simpson 1/3 está basado en la “interpolación cuadrática”, pues crea una función que pasa por todos los puntos dados e integra. La función es para dos intervalos continuos (Uniendo 3 puntos). La aplicación múltiple, es crear pares de intervalos para aplicar Simpson 1/3 a muchos puntos.

Romberg está basado en la integración por trapecios, y utiliza un método parecido a las diferencia divididas de Newton. Obtiene aproximaciones de integración con diferente número de trapecios para cada aproximación, y las mejora, hasta obtener un mejor resultado para la integral.

Justificación y propósito

Simpson 1/3 utiliza la siguiente función, para calcular la función perteneciente a los intervalos, antes de integrar.

{displaystyle P_{2}(x)=f(a){frac {(x-m)(x-b)}{(a-m)(a-b)}}+f(m){frac {(x-a)(x-b)}{(m-a)(m-b)}}+f(b){frac {(x-a)(x-m)}{(b-a)(b-m)}}.}

Debido a que siempre es la misma formula para obtener el polinomio, es fácil calcular la integral.

{displaystyle I=int _{a}^{b}f(x),dx}

Donde f(x) es el polinomio de Lagrange de grado dos. Podemos calcular una ecuación para cada intervalo, sin necesidad de calcular la integral de cada uno.{displaystyle int _{a}^{b}f(x),dxapprox {frac {b-a}{6}}left[f(a)+4f(m)+f(b)right].}

Por otro lado, Romberg mejora el método de trapecios utilizando la siguiente formula

Simpson & Romberg

Donde se obtiene una integral mejorada, utilizando otras dos aproximaciones de trapecios, de menor nivel (Uno más exacto que el otro). Se denomina como aproximación más exacta a la que utilizo un mayor número de trapecios. Este método es recursivo, si obtienes x aproximaciones de un nivel mayor, todavía puedes obtener x-1 aproximaciones de nivel mayor.

Diagrama de flujo

Simpson 1/3 Múltiple.

Simpson & Romberg

Romberg.

Simpson & Romberg

Código en C++

Simpson 1/3 Múltiple: Github.

Romberg: Github.

Ejemplo comparativo

Para Simpson 1/3 múltiple, no se utiliza el método de trapecio para intervalos extra (En caso de que el número de intervalos sea impar, el último intervalo se integra con trapecios).


f (x) = x²

Se integra f(x) de 1 a 5:Simpson & Romberg

Simpson & Romberg

Usando Simpson de 1/3 múltiple se obtiene una aproximación con una precisión de 0.00001:

Simpson & Romberg

Usando Romberg de 3 aproximaciones iniciales, usando aproximaciones iniciales de 100, 150 y 200 trapecios:

Simpson & Romberg

Romberg no fue preciso en este caso ya que este método fluctúa dependiendo de los valores de las aproximaciones por trapecios que se obtengan y qué tan diferentes sean estas entre sí.


f(x) = e^(x)

Se integra f(x) de 1 a 3:

Simpson & Romberg

Simpson & Romberg

Usando Simpson de 1/3 múltiple se obtuvo una aproximación con una precisión de 0.1:

Simpson & Romberg

Usando Romberg de 3 aproximaciones iniciales, usando aproximaciones iniciales de 100, 150 y 200 trapecios se obtuvo una aproximación a la integral definida con una precisión de 0.1:

Simpson & Romberg

En este caso ambos métodos dieron valores certeros, con el mismo grado de precisión.


Simpson & Romberg

Ted-Talks Opinion

--Originally published at That Class Blog

Let’s talk first about the talk given by Sam Richards, the one about empathy.

I liked his example, putting the US in a position like the one they make other countries be when a society or a regime gets in between a resource and the US. It’s not right, and it’s even worse when the US citizens know about how are the foreign people treated by their army, by their governments that will do anything to their citizens if the US helps them to stay in power… And won’t do anything. That’s wrong. People should at least acknowledge that what they are doing is wrong. That people are getting killed, innocent people, to maintain the American economic lifestyle.

We often forget to ask who is really paying for our comfortable way of life. And if we did that more often, and put ourselves in the shoes of the people who are paying, with their lives, not just with money for the bills. They are suffering! They are dying! Their houses are bombed every day! From their government, from other nations, from the terrorists, from whoever. And the innocents, they don’t support the terrorists, they don’t support the regime, and still, they are paying the hard bills for us.

At the end, when using empathy, we allow ourselves to see the big picture and understand the other’s perspective. In the experiment Sam Richard does, we understand how a regular person, a regular Muslim, with a regular family, who live on a regular street in Syria, Iran, Irak, Afghanistan or wherever people are being subjugated to the other’s will sees the conflict. And Sam doesn’t ask you to change your perspective, but to understand theirs.

And as I have experienced, if you understand the other party perspective, it’s more probable to solve a conflict.

And if using empathy could help us understand wars, it can help us to solve any conflict in our regular life. Understanding your boss, who is asking you to perform at 110%. Understanding your employees who can’t really give their whole life to the work. Understanding your parents when they are angry at you. Understanding your sons when going to difficult stages of change in their lives.

It’s easier… And more healthy even.

And now into the David Marquet’s conference, about leaders.

First, I like to have control too. And I know that to have control is to give control.

Second, I know that my recurrent problem is that the people who I give the power to do things, they usually do nothing with it. I guess there are some people who need to be pressured as if they were hard-labor workers, like the ones in the photo of the factory.

And I know that it isn’t a really good practice. Because at the end, who really knows how to deal with the stuff, is the one who interacts with it as a regular job, not the one person who wants a change, or an execution of the stuff. And this doesn’t mean to give all the power of control to the subordinates because the organization and performance are at risk (Like the example he uses of giving leaves to the officers).

I gues I’m giving to much control to others… But still… I really do think there are people who ar ejust lazy.

 

 


Ted-Talks Opinion

2 of Plato’s regimes – MMM Ch. 4

--Originally published at That Class Blog

Conceptual integrity is to maintain the good initial ideas, even if it implies not making that many features. This is better than having lots of uncoordinated ideas.

Likewise, functionality and simplicity. Which is better? There have been software projects that have an outstanding performance in one of both.  But we have to say that it’s a combination of simplicity and straightforwardness. Every part must use the same philosophies, same semantics and syntax and the, you have conceptual integrity.

And how, or more precisely, who will achieve that integrity? Either one mind (Surgical team), the equivalent to an aristocracy; or a handful of minds, with a division of architecture and implementation, the equivalent to a democracy.

And it’s no bad to desire a democracy, but it has several problems. The people in charge of the architecture should consider methods to implement their architecture; the implementers, when suggesting ideas must preserve the integrity of the architects.
That’s why at the end, the aristocracy (Surgical team) is better because as long as the design and implementation are clear in the surgeon’s mind, everything is ok. Or at least, an aristocracy of architects, who have a clear idea of how the system should behave, and they should decide how to maintain integrity, even if that means to leave the implementers with practically no voice, at the end, what must endure the most is the design, not the implementation.

2 of Plato’s regimes – MMM Ch. 4
Using CC (BY-SA). Published by Jocelyn Kinghorn. From https://www.flickr.com/photos/joceykinghorn/11487085615

 


2 of Plato’s regimes – MMM Ch. 4

The Surgical Team – MMM Ch. 3

--Originally published at That Class Blog

The dream: To develop a project with a small team you can know, and trust in their abilities. Forget the big teams, integrated by mediocre developers. In fact, by spending the double on a very good programmer, 10 times the performance could be expected. We can do this all by ourselves. Ah… The dream…

The reality: Big and complex problems need to be done and big teams are needed to do this.

How? Harlan Mills offers a solution. To make several teams, where only one member of each attacks the problem, and the rest only assist him in achieving maximum productivity. You keep a few number of people in the building design, but a lot of them in the actual construction. Just like a surgical team.

Here are the roles of this approach:

  • The surgeon: Define functional and performance specification. Designs, codes, tests the program and writes its documentation.
  • The copilot: Is a less-experienced surgeon. He advises the surgeon, and he can listen to him, or not. He knows the code perfectly but isn’t responsible for it.
  • The administrator: Even though the surgeon is the boss, he needs someone to be in charge of his administrative (Money, personnel, machines…) decisions. The administrator can serve 2 teams.
  • The editor: He writes the external and internal documentation. He reworks the draft of the surgeon.
  • Two secretaries: One for the administrator and the other one for the editor.
  • The program clerk:  He is in charge of the maintenance of the machine and user readable files.
  • The toolsmith: File, text and debugging services. Made fast and with quality
  • The tester: Design test cases and data.
  • The language lawyer: Master of the language selected for development. Can work with 2 or 3 surgeons.

Because the purpose of this team organization such that everyone is represented as just one, the surgeon. At the end, instead of organizing a 200 people team, you just have to deal with 20 surgeons.

The Surgical Team – MMM Ch. 3
Using CC (BY-NC-ND). Published by UCD School of Medicine in https://www.flickr.com/photos/76652722@N04/6878041625

The Surgical Team – MMM Ch. 3

Interpolación de Newton y Lagrange

--Originally published at That Class Blog

Antecedentes

Ambas interpolaciones son métodos que permiten la creación de un polinomio de grado n-1, donde n es el número de datos que se tienen. Ambos métodos asumen que no existe ruido en sus mediciones de datos, es decir, que el polinomio generado por el método pasara por todas las coordenadas insertadas al método.

A menos que se conozca la función original (Lo cual es prácticamente imposible al aplicarlos en la vida real), no hay manera de calcular algún error.

Justificación y propósito

El propósito de ambos métodos es poder generar una función para la cual se puedan introducir todos los datos originales y obtener 0 error, pues la curva se va modelando punto a punto. Al obtener una función, se puede crear aproximaciones y estimaciones.
(En nuestros métodos, no proporcionamos al usuario la función, solamente el valor f(x) para la x deseada)

El método de Newton utiliza la formula de polinomios de Newton:

N(x)=[y_{0}]+[y_{0},y_{1}](x-x_{0})+cdots +[y_{0},ldots ,y_{k}](x-x_{0})(x-x_{1})cdots (x-x_{{k-1}}).
Donde las “y” entre corchetes se refieren al calculo de diferencias dividas, el cual es un algoritmo usado para computar tablas de funciones logarítmicas y trigonométricas, usando división recursiva.

El método de Larange obtiene una misma función, pero elimina por completo la necesidad de usar diferencias divididas:

L(x):=sum _{j=0}^{k}y_{j}ell _{j}(x)

donde

ell _{j}(x):=prod _{begin{smallmatrix}0leq mleq k\mneq jend{smallmatrix}}{frac {x-x_{m}}{x_{j}-x_{m}}}={frac {(x-x_{0})}{(x_{j}-x_{0})}}cdots {frac {(x-x_{j-1})}{(x_{j}-x_{j-1})}}{frac {(x-x_{j+1})}{(x_{j}-x_{j+1})}}cdots {frac {(x-x_{k})}{(x_{j}-x_{k})}},

Los polinomios de lagrange son más fáciles de computar, pues elimina la necesidad de recurrir a métodos de recursión,

Explicación gráfica de la interpolación

 

Estas gráficas representan dado un conjunto de puntos la interpolación de grado 1 a 4 de una funciócon la intención de obtener el valor de f(10), y cómo se ve reflejado esto en una gráfica.

Para la gráfica a) se hace una interpolación de grado 4, por lo que se utilizan los 5 puntos dados y se obtiene una función que cruce por estos puntos, luego se evalúa en x = 10.

Para la gráfica b) se hace una interpolación cúbica, por lo que sólo se utilizan los 4 puntos más cercanos a 10, se obtiene una función que cruce por estos puntos y se evalúa en x = 10.

Para la gráfica c) se hace una interpolación cuadrática, por lo que sólo se utilizan los 3 puntos más cercanos a 10, se obtiene una función cuadrática que cruce por estos puntos y se evalúa esa función en x = 10.

Para la gráfica d) se hace una interpolación lineal, por lo que sólo se utilizan los 2 puntos más cercanos a 10, se obtiene una función lineal que cruce por estos puntos y se evalúa esa función en x = 10.

Por estas gráficas se puede observar, también, la influencia que tiene el tamaño de la muestra para el resultado.

Diagrama de flujo

Newton

Interpolación de Newton y Lagrange

Lagrange

Interpolación de Newton y Lagrange

Código en C++

Ejemplos resueltos

A continuación se muestra los resultados de varias muestras resueltas por ambos métodos:

Con los siguientes puntos:

x, f(x) = (1, 0), (4, 1.386294), (6, 1.791759)

Se interpola usando estos datos para obtener una aproximación de f(5). Estos datos son muestras de la función f(x) = ln(x), por lo que ya conocemos el valor de f(5) = 1.61.

Interpolación de Newton y Lagrange

Con Polinomios de Interpolación de Newton se obtienen los valores para b0 … bn y aproximación siguientes:

Interpolación de Newton y Lagrange

Para Polinomios de Interpolación de Lagrange se obtienen los coeficientes de Lagrange y aproximación siguientes:

Interpolación de Newton y Lagrange


Con los siguientes puntos:

x, f(x) = (1, 1), (2, 0.5), (2.5, 0.4), (4, 0.25), (7, 0.14)

Se interpola usando estos datos para obtener una aproximación de f(3). Estos datos son muestras de la función f(x) = 1/x, por lo que ya conocemos el valor de f(3) = 0.33.

Interpolación de Newton y Lagrange

Con Polinomios de Interpolación de Newton se obtienen los valores para b0 … bn y aproximación siguientes:

Interpolación de Newton y Lagrange

Para Polinomios de Interpolación de Lagrange se obtienen los coeficientes de Lagrange y aproximación siguientes:

Interpolación de Newton y Lagrange


Interpolación de Newton y Lagrange

Building the software: Ch. 10 SG

--Originally published at That Class Blog

Software architecture provides the technical structure for a project. The time in which the software is mapped to diagrams and prototypes, and the program is partitioned into systems and subsystems.
Everything is described in the software architecture document, that contains the software organization, support for changes, reuse or purchase of components, and design of functionalities.

This document has the following categories:

  • System Overview: Describe the system in general terms. Help build the developer the idea of what is he going to build. As well as a description of the design alternatives that were considered, and why were or weren’t chosen.
  • Conceptual Integrity: Descriptions and diagrams of the almost obvious solution to the problem. Simple, short, and easy.
  • Subsystems and Organization: Here the major clusters of functionality must be defined (Output formatting, data storage, analysis, user input…). Usually a complete system hast five to nine subsystems. The responsibilities of each subsystem are stated, and how are they intercommunicated and restricted.
  • Notation: Describe the notation that will be used for the pseudocode and diagrams (Such as UML)
  • Change Scenarios and Change Strategy: List what parts of the software will need (Most likely) to be changed, and how the team must approach this changes.
  • Reuse Analysis and Build vs. Build Decisions: Define what components are going to be developed from scratch, reused (If available) or bought.
  • Approach to Standard Functional Areas: Functional areas that the architecture must address:
    • External software interfaces: How is the software expected to communicate with other software.
    • User interfaces: How is it isolated, so that the changes made to the interface won’t affect to the system (And reverse).
    • Database organization: Organization and content of the DB.
    • Data storage: Major data structures and what isn’t stored in the DB.
    • Key algorithms: Defined, or for later definition.
    • Memory management: Allocation of memory.
    • String storage: How is the text stored and retrieved.
    • Concurrency/Threads: How is the concurrency going to be handled if the software is multithreaded.
    • Security: Is the software going to operate in a secure environment?
    • Localization: Is the software going to be translated for other countries or regions to use?
    • Networking: Network operations.
    • Portability: Handle other environments (Android/iOS, Windows/UNIX)
    • Programming language: Specific language?
    • Error handling: Strategy to handle errors.
  • Requirements Traceability: Assign each of the requirements to a subsystem.
  • Support for a Staged Delivery Plan: Describe how the project will adjust to support staged delivery. What is going to be delivered on each stage.

Building the software: Ch. 10 SG

Quality Meter: Ch. 9 SG

--Originally published at That Class Blog

Let’s do this fast folks, because this chapter was looooooooong.

So, first things first. What is quality? We have to thank McConnel for this simple definition: “the degree to which the software satisfies both stated and implied requirements”. Easy to understand. Perfect.

But why is this important? Because keeping defects controlled affects the development speed, cost and characteristics. A delivered low-quality software increases the cost of end-user support, among other things. McConnel emphasizes in remembering that the end-user tends to forget the delivery time of the software, but not if they liked using it. They forgive taking even taking more time than expected if they will enjoy using the software.

And how to keep the quality up at all times? Using the Quality Assurance Plan. The team must commit to the QA Plan. The activities of the QA must be planed and committed to writing, the QA must be established at least at the same time as the project requirements, a group that will take care of the quality must exist, with capable people, and finally, all the QA activities must be well funded.

And which activities conform the QA Plan?

  • Defect Tracking: Keep record of every defect that is found during development. The moment of detection, and the moment where the defect was resolved (And how). Its important to keep the information public, so the team can adjust the estimates, and see progress.
  • Unit Testing: Testing made by the developer who wrote the code. The unit can refer to any programming entity. It’s informal.
  • Source-Code Tracing: Carried out also by the person who wrote the code. Consist in going through the code, line-by-line, using a debugger.
  • Technical Reviews: This are reviews made by the peers of the code author. Usually are given by the full team, and the QA staff just makes sure this reviews are being carried out. Technical reviews have a pattern:
    • Notification and distribution: The code’s author notifies to the QA team that the work is ready to be reviewed. The material is distributed to the corresponding people.
    • Preparation: Reviewers review the work, using checklists of common errors. They schedule a review meeting.
    • Review meeting: Evaluate the work alongside the developers and a moderator. The main objective is to detect defects.
    • Review report: The results of the meeting are written to paper, committed, and added to the defect tracking.
    • Follow-up: The developer applies the changes, which are evaluated again, and if approved and passes the review, the material is added to the list of materials that have been successfully reviewed.
  • Integration Testing: Test the code developed with already tested code, which has been integrated to the complete system.
  • System Testing: Execute the complete software to find defects. This is by no means a complete way to find all of the defects.

Sometimes companies also use Beta Testing. Even if the book doesn’t recommend it as much as the other tactics (Which are internal), it might be helpful to consider using it, because of the great final-user feedback the developers might receive.

And that’s it. Let’s not stop the testing, until we run out of cake…

Quality Meter: Ch. 9 SG
Used under CC 2.0 (BY-SA). Published in: https://www.flickr.com/photos/chripell/3409348666

Quality Meter: Ch. 9 SG

The end is nigh.

--Originally published at Ce qui est chouette

This post will cover Chapter 17 & 18 & 19 of Software Project Survival Guide and Chapter 8 of The Pragmatic Programmer.

Let’s get to wrapping. To wrap up first hold a change board meeting at the end of each stage that includes the whole staff to evaluate changes that were deferred during each stage. Wrap-up is time to recalibrate estimates, see if milestones were met, whether necessary tasks were omitted, and in general if the project scope still aligns with the estimated at the start.

Evalutate performance against the project plan, whether the team performed technical reviews, kept track of progress and follow the plan. If the plan wasn’t followed, see why, if the plan resulted impossible to follow, replan.

The end is nigh.
The end by Christ Lofqvist on flickr under a CC License

To fully conclude a stage archive the environment that was used to create the software, in the future the team might need to recreate the software at a certain stage, and collect data into logs to serve as reference for future work and estimates. This is stored in Project History.

At the end of the project as a whole, gather all the data you can, e-mails, summaries, and most important, subjective opinions of team members. Hold a meeting, why not? Everybody loves meetings. In it, team members will discuss candidly their insights; a questionnare can be an alternative to this meeting.

To complete the Project History Document include both objective, quantitative information about the project and the team’s subjective, qualitative impressions. Now this document will serve as reference. For full reaping of this document in the future, conclude it in the following two ways: create a planning checklist based on the work done for the project and add the major risks to a general Top 10 Risks List template.

To fully conclude, a list of Do’s.

  • Create and follow a Software Development Plan.
  • Empower project personnel.
  • Minimize the bureaucracy.
  • Define the requirements baseline, and manage changes to it.
  • Replan when necessary.
  • Reestimate system size, effort, and schedule periodically.
  • Define and manage phase transitions.
  • Foster a team spirit.
  • Start the project with a small senior staff.

Now onto the Don’ts:

  • Don’t let team members work in an unsystematic way.
  • Don’t set unreasonable goals.
  • Don’t implement changes without assessing their impact.
  • Don’t let additional complexity creep in.
  • Don’t overstaff.
  • Don’t relax standards.

Onto the last Pragmatic advice. This time it’s pragmatism for the whole team. Teams should be divided based around functionality, some team will cover a certain subsystem defined in the architecture, another will serve as the change board, and so on. Automate everything that must be done systematically, people aren’t as efficient at standard procedures as one may think. Leave testing and installations to machines. Design a test case that can be performed without any human intervention, this way test can be performed more frequently.

No phase can be counted as done until it passes every test designed for it. A fun but risky way to assure a module, class or routine is fully done is to design a project saboteur, to branch out their version of the current software and introduce bugs to see if the automated testing will catch it. Regarding bugs, once a developer finds one, they should make it their mission that no one else finds that bug, put it under change control to be dealt with.

On documentation, it should be done in parallel to development. Each module should be accompanied by its respective documentation, in code and in a separate markup language document, signed by its developer in some way, be proud of your work, developer.

This is it, the about guy

 

 


The end is nigh.

It is time. Or is it?

--Originally published at Ce qui est chouette

This post will cover Chapter 16 of Software Project Survival Guide.

The entire team must gather around a common goal: driving the software to a releasable state. By doing this, quality is ensured by focusing on polishing key factors of the code at each stage. Developers may begin to work on the next stage’s detailed design, when they’re ready to release the current stage, but must be ready to drop everything and go back to correct any defects.

It is time. Or is it?
Noel Gallagher by Peter Hutchins on flickr under a CC License

As to when to release, there are some techniques, the most basic one consists on defect counting if the defect count is under a certain number by the end of the stage, it is determined to be ready for release; defect density consists on averaging out the number of defects per line of code; defect pooling, for this keep two pools of defects with arbitrarily separated defects, report defects into the first and report into the second, the number of unique defects is given by the union of both pools, compare this against the approximate total number of defects given by the product of both pools divided over their intersection; defect seeding consists in planting errors in the code and having another team find them.

The development team along with the testing team should keep a public note of the defects detected and work being done on them. A public space such as the break room, an office or meeting room is usually OK, a private website could also work.

Keeping a release checklist can help review the release and on being prepared for it. It should consist on the critical release activities the team wouln’t like to forget in the heat of the moment. Similarly a release sign-off form ensures that every member of the project team is in accordance to releasing the software at each stage.

I’m now singing In the Heat of The Moment, your wonderwall guy.

 


It is time. Or is it?