I am a >Code supporter. In the latest episode, guest Jerome Hardaway said something along the lines of:

Each day I want to be 1% better than the last

I don’t think that even he understands the power of the message. Let’s make a simple graph out of it, where on day 1 we have a value of 1 for a certain skill:


1% better daily

What this means is that in less than a year you aim to be x7 better than when you started. Granted this is only a rough estimate and depending the subject, the difficulty, other issues and any other kind of ceiling a x7 result won’t come, but still, the result won’t be linear.

As the days go by, you’re becoming extremely more competent than you thought, even if today you did not make it to be 1% better than yesterday, but only 0.1% better.

The point is that you do not stop.

Small Prolog


A small post just to break the blog hiatus. Once upon a time there was a Prolog interpreter that was used in the Windows NT kernel. You can read about it here. The code is in the public domain and years ago I had downloaded from the net. It seems that it took me close to 8 years to put it on GitHub, for software archaeologists to dig into if they like.



This is a nice trick for running your 30 minute meetings. I found it during listening to the latest episode of The Engineering Commons. 30 minutes are the standard slot for a meeting it seems. So plan for a 30 minute slot, but plan to run the meeting in 22 minutes. Why? Because if you are in a series of meetings, chances are that you’re going to be on time only in the first one (and even that is debatable). So delays accumulate and you can forget ever making it to the last meeting.

So why 22 minutes? My guess is that if you have planned for 5 meetings in a row, this 22 minute window slides enough within each 30 minute slot so that you can make it in all of them.

For longer meetings you can always multiply appropriately.



Today it was my fourth (I believe) encounter with APL:

  • First, too many years ago when skimming through Tim Budd’s “The Kamin Interpreters in C++” (and the Kamin book afterwards). For the hardcore fans, Tim Budd has a book on implementing an APL compiler.
  • Next was a Dr Dobbs issue about the J programming language.
  • Some years a ago a comment on this blog about Dyalog.

Today it was Functional Geekery’s Episode 65 where I found out about the most interesting (to me) implementation of Conway’s Game of Life.



Here is an experiment: today instead of watching the news for the night, I picked up a record I had not listened to for years, Live by Jean Michel Jarre.


Live, 1989

I had gifted it to a friend who passed away 10 years ago and I do not believe I had listened to it since then.

Rendezvous Houston - Houston Festival Light show in downtown Houston

The kids were impressed by the music τo the point that they wanted to learn more about Jarre. So I headed over to YouTube and played the whole of the Huston 1986 concert (another album which I had not listened for years since I only have it on cassette). They were really fascinated by what they saw also. The laser harp and the circular keyboard made an impression.


Dad, he must be a super cool engineer!

I do not know about engineering, but design-wise he won their hearts out. We even got to listen to Ron’s piece (from Ronald McNair who was supposed to play this in space) which led to discussing about the Challenger.

Funny what feelings, forgotten memories and processes a forgotten CD can spark.

[ Mostly saving this post for posterity ]

I had heard of the language long before it was made Open Source. In order to get access to the implementation I had Timos Sellis sign an NDA with Ericsson. I had fun with Mnesia (the distributed DBMS) and started thinking of whether it could be applied to stuff I was good at at the time (namely SMTP and DNS). In fact it was. Because most of the Erlang team formed a company named Bluetail that was making software based load balancers with Erlang and got sold at $152M.

Nowadays I occasionally have “fun” with the Erlang VM whenever I have a RabbitMQ instance go mad.

But this? This is something only someone like the guy who writes RabbitMQ could discover:

Donald Knuth was the first Erlang programmer

The Fields Medal. A genius Mathematician racing against time. Will he make it before he’s 40? The journal of his adventure is what this book is about. Along with his fears, the way he measures himself against other giants (does he really have to? To the reader it may seem he does not, but to him is matters a lot; the top is the top). And yes, it is written as a diary.

I will not pretend that I knew, or still understand anything about Landau damping. Or even the Boltzmann equation. In fact I know only of a handful of people who could recognise the complete notation used. Interestingly, this does not make the book less enjoyable. It took me over a year since I first found out about the book, started reading it and then finishing it. And you know what? It was not the Mathematics that put it to a pause. It was a long chapter on music, on stuff that I could follow in complete detail. When I reached there I paused in order to find the proper time to enjoy the written soundtrack. Well that time came about a year later.

Far too many top Mathematicians and Mathematical Physicists parade in the book. Perelman‘s antics and perseverance get explained. Luckily some simpler (to grasp, not solve) problems too, the Collatz for example. As it happens I too use it sometimes to block my mind from interference. But that’s where the similarities with the author stop. The other thing that felt close to heart was Coq. I always had ideas about using it in ways not expected by its creators, but my academic limited attention per suject blocked this.

All in all a fascinating read about a man, his passion and his target. You can safely ignore the Mathematics. It is hard enough for the professionals and is there to show what he dealt with. The pain is described not in algebraic notation but with plain alphabet.