Prabros. logo

Prabros.

Growing Software Tools

25 January 2016

1 / Past

For a major part of our career as designers, we dedicated our time towards designing visual interfaces. We worked with dozens of teams over half a decade designing UIs and if time and budget permitted, we illustrated it’s workings with an interactive prototype. When it was time to turn these visual artifacts into a usable product, we would handoff our work to developers’ hands and a pretty interesting process unfolds.

After some iterative development cycles and much back and forth, the result would range somewhere between outright unusable mess to delightful products. Delightful products were almost always a result of great engineering and insightful business planning but failed ones resulted from a myriad of reasons ★1: due to miscommunication, lack of understanding of the domain, misalignment with business goals, and so on.

Both of these outcomes, resentment resulting from inferior products and curiousity at how the proficient engineer brings about delightful software impelled us to learn the craft of software engineering and thus enabling us to be fully accountable for our creation. Thus about 4 years ago, we decided to learn how to program a computer.

There were moments of joy, frustration, and glimmers of enlightenment before the next problem brought us to a grinding halt in this journey. Some of these experiences lead to a clear view of the lay of land allowing us to solve certain problems but most aspects of how to overcome failures on our way to build robust systems remain elusive. Looking back at our past experiences and drawing inspiration from a range of domains, here we try to distill an outline of our approach towards our next projects.

2 / Computer is a Construct

Computer lends itself to be shaped into a multitude of models. Alan Kay identified computers as the first meta-medium — that it could incorporate all other media★2. This “protean” nature of computers means that it could be anything it is shaped to be. But this flexibility comes with a consequence: what the computer has come to resemble at this point is nothing but one realization amongst many it can be. What it is now, is a resultant of thrusts from various fronts that have been exerted over the course of it's history.

A recurring observation in our exploration is that in this tug of war, a lot of what got adopted were imagined and implemented at a time a lot different than ours. Our current technological and mental models originated at a time when hardware had many practical limitations, a different knowledge context than ours, and certainly a very different audience for computing. Many interesting ideas of yore that were abandoned and can now seen to be resurfacing ★3 in Computer Science since the technology and people around it are becoming more favourable for them to flourish.

An understanding of history is crucial towards understanding any technology. A lot of concepts in programming has a long backlog behind them. In order to understand why a certain technology is used in a problem space, it is crucial to understand it’s historical trajectory. A great many lessons can be learnt by examining these ideas (and their predecessors) and the philosophy possesed by the minds behind them.

A large part of our undertaking in building software systems is to dedicate ourselves to unravel the towers of abstractions before us and to investigate ones that have disappeared. We hope it will lead us to a coherent view of computer science, the lack of which we faced first hand when we started out.

3 / Authenticity

We have always had immense respect for the people who build products drawing experience from their diverse backgrounds; the ones who have their creative vision expressed over all aspects of their creations. One such niche where small groups of people with multidisciplinary backgrounds expressing their software craftsmanship is the Demoscene. The idea is to build a so-called “demo” which is an tiny program limited to kilobytes in size that executes to generate an audio-visual world. Even within it’s tiny confines, the creativity expressed by the artists knows no bounds ★4.

One exceptional talent who we have come across in our journey is Charles Moore. He is responsible for building not only his own programming language called Forth, but went on to design the hardware to run it on (he designed this in a software written in his own language!) . He not only has an astounding array of systems he has built in his life but also has an pretty interesting philosophy driving in his work ★5 that is derived from deeply understanding the machine.

This joy of learning about new things from diverse backgrounds is one of our primary drives. It grants an opportunity to have more than one perspective towards product making. It facilitates taking apart the elements of computing and putting them together to create novel combinations. We have felt that this exposure to a wide variety of domains, not only guides us in new creations, but also helps to develop an authentic style.

This of course comes with the cost of having to split your time and mode of thinking between abstract nature of software logic and subjective cultural symbols for expressing it. But if we could successfully sythensize software in this manner, all the components of the software would have a sort of coherence displayed across it’s marrow and joints.

4 / Software Farm

It’s not just intellectual gratification and authenticity that inform our approach but also an overarching idea of creating an ecosystem of software. The metaphor we have adopted to see our ventures collectively is by understanding them as a farm. As rightly revealed by the metaphor, some of our products will emerge from cross polination of ideas, are prone to wither on the wine, and most importantly will need constant maintenance. It’s the slow and steady work leading to fruition over a long period of time that we wish to remind us with this metaphor ★6.

This in operational terms means creating applications that interoperate with each other. My undergraduate thesis was on creating a software ecosystem for students. Though I had ambitious goals with the project, the progress made over an year was pretty meagre. Eventhough from whatever headway I could make, I was able to see how enriching such an ecosystem of applications could be.

What we have in our minds, is a software suite that works as an ecosystem sharing their data in visible ways. Time tracked in your project management application coming up in your calculator. A calculation done on it piped in to your note taking application and so on. In this way, each application in the setting would be enriching to the other ★7. This is obviously not a new idea, but so far, we have only seen such system of applications expressed with limited patency.

5 / Wonder ➝ Blunder ➝ Refactor ➝ Repeat

Nature of software is something that has kept us wondering. From what we could gather, software building stands at the cross section of logic, engineering and human psychology: math, machine, and man. Software is a tool that is borne by shaping our thought in to the machine which inturn results in shaping our behaviour. This means that when developing software we need to take these factors into account in our feedback loop. That is the code should be correct, performant, and understandable.

We admit that we only have an inkling on how to go about implementing any of the systems that we have envisioned. But we are going forward because no decision or tool made at the end of the day is perfect. It seems like a reasonable decision since we are in the infancy of understanding these systems and it still gives us a kick to explore new things in programming. We will be constantly trying to refine our approach as we go forward and trying our hands at different things to gain more experience ★8. There’s so much more left to learn and practice.

Software has a unique characteristic that lends itself to make it better. Known as refactoring, it is the practice of making the code better in small transformative steps which allows for better management of it’s complexity. It is much like how constant rewriting leads to better writing. This practice of code reorganization leads to a better understanding of the code and many a time unearths bugs which would be tough to spot otherwise.

Seymour Papert in his seminal work, Mindstorms borrows the concept of debugging from Computer Science and applies this notion to epistemology. He eloquently writes ★9 on how your brain can be debugged of it’s misconceptions and incorrect intuitions. Our learning theory is based on this idea that you can debug your own thinking by refactoring your mental models. It’s a very empowering notion to adopt for it enables new ways of seeing.

6 / Setting sail

Don’t feel as if the key to successful computing is only in your hands. What’s in your hands, I think and hope, is intelligence: the ability to see the machine as more than when you were first led up to it, that you can make it more.Alan Perlis

With this vision, we are choosing to first build small systems of our own and use them in helping us carry out our daily activities. We have tentatively titled them as Decent Tools. Although, the bigger projects that we have envisioned under this idea are ones that we think we will come to their full fruition only 5 years down the lane or so. We are only at the stage of taking baby steps converging towards this goal. It’s simply because of time required to understand and explore breadth and history of these areas that we have identified. They warrant thorough exploration to come up with something meaningful.

Our intent is to stay small, create products that can fit both of our minds and use this as a constraint towards assembling this ecosystem. But this approach we have been outlining so far is by no means to be taken up as a principle set in stone for we are iterating as much on our approach as we are involved in building our software tools. If anything, it’s only a rough sketch outlining our vector of progress; an in medias res peek at our evolving philosophy.

A lot of planning and a few prototypes have been done in past years which have strengthened our current view. But it remains to be seen if it would continue standing the test of time. We will be constantly putting it under scrutiny for determining it's fruitfulness as an approach towards growing software tools.

So here's a start to our journey, setting sail towards unseen shores.

7 / Bibliography

People

Picture of alan-perlis

Alan Perlis

April 1, 1922 - February 7, 1990

Known for his pioneering work in programming languages and to an wider audience through his epigrams on Programming. Received the first ACM Turing Award for excellence in Computer Science.

Appearence: Setting Sail

Picture of christopher-alexander

Christopher Alexander

Born: October 4, 1936

One of the most important architects of 20th Century who influenced a wide range of domains with his work. Even with having read only his Notes on Synthesis of Form we plan to soon follow up on this with his other books.

Appearence: Sidenote 1

Picture of seymour-papert

Seymour Papert

Born: February 29, 1928

Mathematician reknown for his work in AI, learning theories, and for building educational programming language LOGO for children.

Appearence: Repeat

Picture of charles-h.-moore

Charles H. Moore

Born: September 1938

Charles Moore is best known for his invention of Forth, a stack based programming language that has even found its applications in NASA's space mission. He has an incisive view on Computer Science that cuts across all boundaries and is explicit in his language design.

Appearence: Chuck Moore

Picture of alan-kay

Alan Kay

Born: May 17, 1940

Originated Object Oriented Programming and Window based GUIs at Xerox PARC. His writings and talks distill knowledge from vast variety of domains into overarching narratives. Be sure to watch at least one of his talks. We recommend Normal Considered Harmful for starters. Drawing Courtesy: Jack Shaedler.

Appearence: Protean Machine

Picture of bret-victor

Bret Victor

Born: July 12, 1977

Explorer of new ways of thinking with Computers. His showcased work and writing at his website are treasure troves for software designers.

Picture of karsten-schmidt

Karsten Schmidt

Born: July 12, 1977

Karsten Schmidt is a creative coder who merges art, code, and music in his creative explorations. He had his early roots in the 8-bit demoscene and has been actively designing audio-visual experiences under his banner, Post Spectacular.

Appearence: Sidenote 4

Picture of john-carmack

John Carmack

Born: August 20, 1970

Pioneered many 3D game techniques in popular games by him such as Doom and Wolfenstein that were major stepping stones towards the current gaming scene. He now works at Oculus bringing VR to the world.

Appearence: Blunder

Teams

Logo ofMercury

A Demogroup dedicated towards consistently churning out high quality demos.

Appearence: Sidenote 4

Logo ofFarmer and Farmer

A series of vignettes on human-technology interaction. Jointly written by Sep Kamvar and Jonathan Harris.

Appearence: Sidenote 7

Logo ofCommunications Design Group

Research group aimed at inventing new ways of thinking with Computer. Recruited by the group comprises of Vi Hart, Toby Schachman and many more talented people.

Appearence: Sidenote 3

Books

Notes on the Synthesis of Form

1964

Great treatise on how to solve complex design problems and create forms that exhibit good fit with their context.

Highly recommended read to any problem solver.

Appearence: Sidenote 1

Mindstorms

1980

Seymour Papert’s work on designing programming environments that foster learning and exploration.

A great read to that opened up our world to philosophy of learning.

Appearence: Repeat

Writing

Computer Software

1984

Alan Kay's thoughts on Computer Software that appeared in the Scientific American.

Appearence: Protean Machine

Jacobs Ladder of Coding

2015

An autobiographic post by visual artist Karsten Schmidt. He looks back on his journey that spans bare metal programming, demoscene, programming languages and many more.

Appearence: Sidenote 4

Deep Thoughts: Engineering Lecture Series

2015

Internal lecture given by John Carmack at Facebook.

Published by

Appearence: Blunder

Videos

Future of Programming

2013

Bret Victor's look at some of the rich ideas from the past that never made it mainstream. He also gives an analyzes the context that lead to the germination of these ideas.

On

2015

Mercury's fascinating demo. Captivating shape transformations with on-key music.

Appearence: Sidenote 4

If you any have comments, feedback or kudos on this post, we would love to hear it.Converse with us on Twitter.