domingo, 12 de abril de 2015

Symbolic Analysis & Symbol-Driven Engineering

Authors introduce a framework to connect the core areas of computer science.

It would be interesting to apply this framework, semantics, to administrative IT. As the algorithms are relatively simple, we eventually could prove programs correctness.

See also Semantics.

domingo, 5 de abril de 2015

O maior obstáculo da TI é a própria TI

Artigo do Cezar Taurion, excelente, como sempre. TI tem que se repensar. Tem toda a razão.

Eu só acrescentaria que TI tem que de alguma forma sair do modelo administrar pacotes de fornecedores - o que não é mais TI, é departamento de compras. E voltar a desenvolver soluções originais. Desenvolver.

sábado, 4 de abril de 2015

7 reasons why frameworks are the new programming languages

This paper is a confirmation of the tendency to eliminate the universal machine which is the computer revolution breakthrough. We went from programming language back to special machines.

Remember the accounting machines of the first half of last century? Not universal, only accounting machines. Or the unit record equipment. Special machines, fixed cycles, we could only "program" a board. We could change the position of a result in a printed form. Or add two numbers during the fixed cycle. No real programming.

Frameworks are special machines. Powerful, but specific, not universal. As the paper shows, programmers using frameworks do a little coding. Almost no algorithm. They call a lot of APIs.

Note: APIS. Not functions, not procedures, not routines, not objects. So, no modular programming. Just small changes to a fixed program machine.

This is what I call the "un-universalization" of software. Or more precisely, the end of software.

To be able to produce good software, innovative, creative, software which makes a difference, software which produce competitive advantage, we must maintain the universal machine.

Which means maintain programmability.