T3chFest 2018 (English)

Conference itself

T3chFest conference:


Keynote, Artificial intelligence to stop dyslexia, by Luz Rello

Dyslexia is not related to any kind of intelligence. It means a very intelligent person can have dyslexia.

People with dyslexia are able to read, with the same understanding, texts with errors, without errors or with flipped characters. People without dyslexia read much worse when texts contain errors.

It wasn’t a technical talk, but lot’s of studies about dyslexia have been supported by technology: eye tracking, statistics,…

¿Einar es disléxico? Confunde mucho la a y la e al leer y le cuesta hacer dibujos simétricos (cuando para otras tareas parece que la abstracción espacial la tiene bastante buena).

The Future of the Conversational UIs, by Jorge del Casar

Several phases: design, implementation, test, deploy. Very similar to software development. Not sure why the speaker didn’t match both worlds.

The design phase is the more important one in conversational UIs, in contrast with traditional UIs. It’s very important to think about use cases up front.

  1. There must be a Persona, so users will feel the conversation
  2. Think out of the box: every user is different, every user will lead the conversation in different ways. Conversations are not as simple as Command-Response. There are several contexts, several meanings,…

There are tons of tools to create conversational UIs

The subject looks interesting, but it seems to be companies are trying to sell their cloud based tool, and that tool is basically a speech-2-text tool. Some companies are moving forward, but the core business is simple, and they’re selling it as a complex business you can’t do without them.

The speaker said nothing about Evie :P


:( Parece que faltaba un poco de preparación a la charla, los ejemplos parecían inventados en el momento, fallos en el texto de las slides (copy&paste?)

So, you think the technology is neutral?, by esus M González Barahona

Very, very interesting talk about technology, not the technical side, but the human and the social side of it. It’s a pity it’s in Spanish, the speaker provides more questions than answers. The talk is full of food for thought.

Technology enables (affords, make it possible, give power,…), but also, technology constraints (limits you,…)

Tech decissions today constraint what we’ll be able to do in the future, even more than politics.

Technology is made by humans, if we don’t like, we should be able to change it.

But that’s usually not involved in social discussions, social debates. We don’t usually ask ourselves (as a society) what technology enables or constraints.

There are three kind of communication services:

  1. Centralized
  2. Federated
  3. Peer to peer

What is the easiest to control? What enables innovation best?

Who’s able to update your car’s software? Would it be ok if the owner could update it? Replace car with refrigerator, phone, coffee machine,… How many companies know where you are right now?

The speaker mentioned the well known battle: security Vs freedom

Who’s responsible about what you can and can’t see on the Internet?

Problem is not if a company (say Facebook, Twitter, Google) censors some content. Problem is that a company, a private company can censor, that it’s legal for a company to censor, that it’s socially accepted a company can censor. Possible solution? more computer power and storage, cheaper cpu and storage, less energy consumption, more mobility. More connectivity, faster, cheaper. Better federated technologies, better p2p communications.


Code and law: historcally, law is who constraint society. nowadays, technology constraints us. Most of the time, law is behind technology

War against general purpose computer: speciallization, almost all electronic devices work with software, but you’re not able to decide what kind of software makes them work (car, coffe machine, washer,…)

Hardware + Software: knowledge that works. Code is not only the recipe, it also cooks the dish.

In the 80’s, there were people that thought software should allow much more than it did, that way is how Open Source was born

Contactless magnetic mechanisms for aerospace and robotic applications, by Ignacio Valiente-Blanco and Judit Esnoz

It was not a software related talk, but it was about engineering.

Magnetic mechanisms: it replaces parts of a tradicional mechanism (gear,…) with magnetic devices: plain magnets, materials made of iron (ferrites, soft-magnetic materials,…), materials that are source of magnetic fields.

Those devices doesn’t have the same limitations as tradicional ones: there is no friction, no contact, so lubricant oil is not needed, what means less maintenance.

Their best product is a magnetic gear. They use magnets to transfer movement from one side to another, reducing the rpm’s (revolution per minute) by 7, 50 or hundreds of times. (7 Hz to 1 Hz)

Another product, z-damper. It’s based on a magnetic break (trains break with magnetic fields through the wheels). It slows down something in movement. It’s used in airplains to reduce vibrations.

Superconductors: they are electric resistance free, some of them emit magnetic field. They did a demo where a piece of metal was at very low temperature, and it floated in the air (similar to this ).

Side-projects for fun and learning, by Marta Cruz and Daniel Villalobos - slides

It’s impossible to know everything, you have to choose. Try to find the sweet spot in: passion + creativity + try new things + enjoy

Use side projects to learn something new, to learn something quite different of what you use in your day to day job, to learn and experiment with other people, to get experience without a formal job

Side project elements:

  1. Passion
  2. Start small
  3. No pressure
  4. Make it simple

Benefits: learning, enjoy, shared knowledge,…


  1. Idea
  2. Debate, discussion: to improve the first idea
  3. Validate
  4. MVSP (most valuable side-project)
  5. Demo, share

Experiment without fear, a side project is a project where failure is wellcomed. Do it right, do it with love

Speakers talked about side projects as if they should be done as part of a team, within the company. I think they just expose their experience. It could be told from a wider perspective

Let’s understand what is reactivity, by José Manuel García - slides, code

What’s reactivity/reactivex? It’s a paradigm to work with asynchronous data streams

Any value that changes over time is a good candidate to be observed. After that, you’ll see observables anywhere, and you’re right!

JavaScript is asynchronous: callbacks, promises, generators, async/await

Observables are functions. They do nothing until you invoke them, they’re lazy. They do someting any time you invoke them, by subscribing to them

subscribe method returns a subscription object that allows managing the subscription: unsubscribe,… Useful to avoid memory leaks (BTW, I didn’t know you should unsubscribe from them to avoid memory leaks)

A subscription ends when:

  1. Data stream is over
  2. Client unsubscribes

Subject can play both roels: publisher and subscriber. The speaker built a state machine UI, publishing events in an observable, and observing them in another component inside the application. Angular app showing it

Operator: function that receives an Observable as param and returns another Observable. They can be chained. There are operators to do almost anythin. They are the interesting part of ReactiveX


:) Live (and visual) examples in browser console

:) Nice example with the lottery (slide #23)

AireMad and Open Source, by Ulises Gascón - slides

To be Open Source, a software must obey 4 laws of freedom:

  1. Free to execute the app as you want
  2. Free to study the source code
  3. Free to distribute
  4. Free to copy and share

Open Data should be promoted by goverments

It is illegal to collect personal data in Open Data, to protect people

Found issues while developing Aire Madrid

Measurement stations: new ones appear without notice, old ones dissapear, stations change location,…

The source of data use relative concept:

  1. Real time: one report every hour (in the best case). reports once a day, with data from previous day.
  2. Formats: there is no standard, even for the same data set. There are user guides to understand the data. Sometimes, there are errors in the specified format (are they created by hand?)

The developer must write lots of defensive code. Postel’s law

There are changes without notice, or even new data without documentation

People publishing the data has their problems too: the want to do better, but they don’t have the knowledge (they need developers like us), the money (no budget), boss/colleagues support,…

Aire Madrid evolved to AireMAD API. They maintain an API with curated data. They grew the data set adding new stations, new station types, new data: stations, pollution, noise, polen, flu,…

Wubba lubba dub dub!, by Jorge Barroso

The talk was kind of: What would you say to yourself N years ago?

The title makes reference to Rick & Morti series and it means: I have a pain that is driving me mad (more or less).

Do a retrospective every year, as a person, as a professional

Learn English (I believe in this tip)

Vocation is not born, sometimes, you can suddenly realize something you’re good at and you love it.

With 4 years of experience I thought I was the best developer in the world ~ speaker (or any other dev)

You should be trying to grow as a person/professional, to grow with others, never alone (if you’re the smartest guy in the room, you should exit that room)

In your free time, do stuff you like, don’t do stuff that are good for your current job

Speak at conferences, if not, at least write, and make them public

Go to conferences

You’re paid only to create software that works. To do that, you need to be good at coding (and other skills). To be good you must get some level of quality in your work. Pay attention to quality.

Ethics: do you share values with your company? if there is something wrong in your environment, you should point it out (abuse,…). If you don’t like your company, find another one!

Be excited about what you do, don’t lose that energy. Be happy at work, be proud of what you build.

How to be an investor without millions thanks to Blockchain and ICOs, by Paola García

DAO, Decentralized Axxx Organization: trying to remove toxic (too powerful) elements present in centralized systems

Before creating a blockchain or ICO, you should have a community (of users). For that, they create tokens, to provide value to people. Community is the most important behind a blockchain

ICO, Initial Coin Offering: an offer with discount to by some currency based on a blockchain

How to know if I should invest in an ICO?

  1. How much currency is allowed to exist?
  2. Is the price raising?
  3. Do you trust the origin of the idea?
  4. Team behind the project

Don’t buy/sell if you don’t understand what you’re buying/selling


:( It seems to be that the talk was not prepared enough. In the end, the speaker didn’t talk too much about how/why invest. She talked about general blockchain concepts you can find in lots of places

Missed talks