STARTECHCONF Santiago Chile 2011

Everyone wants to jump into HTML5 but how do you use the cool features of this new specification while ensuring older browsers render your web pages as expected? This is where polyfills and shims come in. In the session, you’ll learn how to use specially crafted JavaScript and CSS code that emulate HTML5 features so that you can take advantage of HTML5 today without breaking your sites in older browsers.

The JVM has traditionally been associated with one language: Java. Most applications and libraries that run on the JVM are written in Java, and most of the tools available focus on Java. But that's changing. There are dozens of languages implemented atop the JVM, with several of them heavily used by real-world developers today. How did we get here, and how hard or easy is it to target a language to the JVM?

This talk will go over the strategies needed to implement languages on the JVM. We'll talk about the JVM's built-in capabilities like GC and JIT and how you can monitor and tune them. Well talk about code generation and how the JVM's stack operates. We'll look at how the JVM optimizes code. And we'll look at two languages, JRuby and Mirah, as case studies.

HTML5 and friends have been getting implemented in browsers at an impressively quick pace. But that leaves us as web developers wondering, “Great, but how am I supposed to build cross-browser applications with these features when I still have to support IE”.

We’ll break down HTML5 into categories of features: some implementable today, others reserved for hackers. We’ll go into how to handle older browsers by providing them with today’s features or discussing how to convince your clients to let the experience be different.You’ll walk away knowing how and what to implement from HTML5 confidently, and have the tools and resources necessary to develop to equip your team as HTML5 heroes.

Throughout the years, the Swiss Army Knife has been the trusted companion of scouts and explorers alike, and for front-end developers, CSS has been a trusty, if sometimes frustrating, companion. And just as blades, scissors and sundry tools have been added to the Swiss Army Knife, with CSS3, we have new tools and implements of creativity, and some tried and true tools have been honed and sharpened. Of course the key to success is knowing which of the many tools to use and how to wield them in a given situation. Join Stephanie Rewis as she explores some shiny enhancements to favorite old tools like backgrounds and borders, as well as slices and dices with new tools like CSS masks and more!

This session will look at the process of creating a web app that feels more like a native app, from how to approach style, CSS, what JavaScript APIs are available and look at developing apps across iOS, webOS, and Android.

One of the cornerstones of disciplined Agile software development isproper test coverage. Does your implementation match its specification? Have you considered edge cases? So-called "highceremony" approaches advise stringent devotion to writing test code first. In this talk, Obie considers his over 10 years of experience with these thoughts and philosophies and how they work in practice, from the enterprise to startup worlds, and muses on future progress in this area.

Since the early days of the web, the only reliable way to get movement on your site was through Flash, or more recently, Javascript. But now, with WebKit and Mozilla leading the way, transformations and transitions can be done with pure CSS, even on mobile devices. And for those in need of even more movement, CSS3 provides for keyframe-based animations. In this session, we¹ll take a look at all of the possibilities and explore what works and where ‹ from the simplest effects, to creative usability enhancements including the combination of CSS with mobile Javascript frameworks.

To Relate or not to Relate, that is the question raised by the NoSQL movement. There is a lot of buzz about Couch, Casandra, MongoDB, and other non relational databases, and at the same time there are decades of hard work that's gone into optimizing databases built around the relational model.

I would actually argue that there is no such thing as a NoSQL database -- there are a variety of compelling options to relational database -- each of which have different features and different performance characteristics. So no one-sized fits all comparison will do. So, I'll try to outline a general taxonomy for persistence mechanisms, and then proceed to comparing relational DB's to their new friends in practice.

The talk will contain quite a few stories from the trenches with CouchDB, MongoDB, MySQL, Postgres, Tokyo Cabinet, ZODB, and other databases, and hopefully will help you think about the data storage needs of your applications in new ways.

In our quest to reach the holly grail of the realtime web at Yahoo!, we have invested heavily in technologies that can scale and improve performance and development drastically. One of these technologies is Node.JS, which provides the ability to execute Javascript at the server side, closing the gap between server side and client side code. In this presentation, we want to provide some insight about Node.JS and our effort to adopt it as the main power-force behind Search Direct product.

This talk aims to introduce you to the plethora of APIs HTML5 and related technologies offer us. Sure, HTML5 brings eye-candy and we'll look at that, but there's so much more to it when it comes to building day-to-day web sites and web applications.

This talk will a quick but thorough introduction to the jQuery Javascript Library. If you have zero experience with jQuery, this is the talk to start with. We will cover jQuery and its history. We'll then start by introducing basic jQuery concepts and principles. Finally, we'll conclude by building a basic jQuery plugin.

The content of this talk will be much different than the jQuery training offered prior to the conference. Because of the available time, the training will be diving deep into jQuery concepts. This talk will cover some of the same topics, but will only cover them briefly.

The way traditional businesses approach the management and organization of creative, intellectual workers is wrong. By throwing away everything that blocks productivity (meetings, deadlines, managers, titles, strict vacation policies, etc) and treating your employees as the responsible adults that they are, huge amounts of potential can be unlocked and employee happiness and retention can be at unprecedented highs.

At GitHub we've embraced a philosophy that gets things done and strips away policy and procedure in favor of smart decision making and personal responsibility. Come see how we make it work and how you can reap the same benefits in your own company.

Git is the version control system most of us use every day. However, there are some strangenesses to it. Raise your hand if you really understand the ‘reset’ command. When it comes down to it, this is one of the most interesting, fundamental and amazing commands that Git has, yet nearly everybody is afraid of it. This is just bad marketing. This talk will de-mystify the ‘reset’ command so that you are not only comfortable using it, but can do new and interesting things with it and in doing so will arrive at a better understanding of the entire Git system. We will explore the Three Trees of Git (HEAD, index, work tree) and all the cool and mind-bending fun you can have with them.

¿Por qué cuesta tanto acceder a la información gubernamental en la Web? ¿Por qué es tan difícil? ¿Por qué no se asume una estrategia más práctica, menos costosa y mucho más transparente?

La charla se concentra en una propuesta: Que los gobiernos provean acceso a sus datos a través de APIs públicas. Que no desarrollen más interfaces web horribles. De esta forma reducirán su costo y facilitarán el acceso a la información.

Implementar una API es extremadamente fácil hoy en Ruby, por ejemplo: Utilizando Rack, Cuba, JSON y Heroku. Ver ejemplo: http://censo2010.heroku.com/

xtalk

Esta charla mostrará JRuby, la implementación de Ruby para la JVM, como una alternativa ligera al desarrollo Java EE tradicional. El uso de JRuby permite aumentar velocidad de desarrollo, disminuir la tasa de errores y simplificar la arquitectura de aplicaciones enterprise, manteniendo la compatibilidad con componentes y aplicaciones Java ya existentes.

El objetivo de esta charla es motivar a los desarrolladores Java a incursionar en la programación políglota, aprovechando las ventajas que ofrece la JVM para ello.

En una perspectiva práctica, se mostrará cómo ejecutar Ruby on Rails en servidores de aplicaciones como Websphere, JBoss y Tomcat, en la reutilización de componentes Java ya existentes dentro de una aplicación Ruby y finalmente, en cómo simplificar el entorno enterprise.

A pesar de la imagen de éxito del país, nuestra industria de software está varios años atrás de industrias vecinas como la brasileña o la argentina, donde tendencias que son norma en USA como Agile han penetrado con mucha más fuerza que en Chile. Aplicando la técnica ágil de la retrospectiva "Keep, Fix, & Try" se presentará una personal visión de cómo mejorar la industria chilena de software desde la perspectiva de la Agilidad y el Lean Thinking

Multi-core processor architecture is the principal mechanism for continually augmenting CPU processor power in the near future and enabling the industry to continue with the trend described by Moore’s law. However, in order to harness truly parallel multicore systems, it is necessary to write multithreaded code. Programming truly multithreaded code often requires complex co-ordination of threads and can easily introduce subtle and difficult-to-find bugs due to the interleaving of processing on data shared between threads. Unit now, programmers are often shielded from multithreaded complexity by the simplified programming model. For example, Servlet architecture, a dominant web application programming model on Java platform, enables programmer to write applications in “singe thread mode”, serving multiple simultaneous requests where concurrency conflicts are resolved at database level. While this model can still be used, there are numerous benefits in programming for true parallelism. Thanks to its flexible syntax, Groovy programing language introduces a more intuitive and robust parallel programming models to Java platform. GPars library (now part of Groovy core) makes it easy and intuitive to write truly parallel code on Java platform harnessing the Actor and Dataflow concurrent models. The presentation will explain current issues with parallel programming in Java and how Groovy programming language and GPars library resolve these issues. Finally, a rudimentary example applicable in real-life project is coded on site demonstrating the capabilities of Grails and GPars library.

Cuando estás empezando con un nuevo proyecto y tus recursos son limitados tienes que ser eficiente. Al igual que con tu código tratas de maximizar esfuerzos, liberar releases pronto y conseguir feedback rápido, con tus sistemas puedes ser ágil y aprovechar al máximo tus recursos, así como flexible para adaptarte a cambios rápidamente.

Como desarrollador, me gustaría contar mi experiencia sobre cómo hemos afrontado la problemática de los sistemas cuando hemos empezado con proyectos nuevos, consiguiendo buenos resultados en la mayoría de ocasiones y en qué nos hemos equivocado.

We'll make a brief talk about how we managed to keep a 2GB RAM machine up and responding to about 20 million requests per day, with DDoS-like peaks every 20 minutes, by decoupling components and making everything as asynchronous as possible.

Scaling with ruby (with a hand from NodeJS).

Hey guys, take a look around... mustaches are not fashionable anymore... It's time to shave your templates!

This talk will tell you about all business and technical dilemmas while developing user interfaces and help you avoid that stupid war of designers and UI developers using MCVP pattern as an icebreaker!

Obviously, cherry on the pie will be the Shaven templating system...

Web Scraping para el Desarrollo de interfaces Móviles y Web, creación de prototipos e integración de aplicaciones]

En esta charla veremos como aumentar drásticamente la velocidad de desarrollo de interfaces web y móviles usando un proxy XSLT python para web scraping.

Se explicará como mediante la combinación de técnicas de 'web scraping' más JQueryMobile podemos reducir el tiempo de desarrollo de aplicaciones móviles de días a horas o incluso minutos, sin tocar el código fuente base de tu sitio web.

Además, se mostrará como usar web scraping para la creación de prototipos y complejos diseños para ser mostrados a tu cliente rápidamente, y cómo construir una experiencia unificada desde aplicaciones diversas y/o legacy permitiéndonos integrar distintas interfaces de usuario, y todo, sólo usando conceptos familiares de HTML y CSS.

This session describes how we developed findability standards in our organization. The search engine and UI are only two components of an effective findability system, the others being User research, content, and processes followed by creators of the content. All these components need to work well for optimal search UX. So search is not just a technology problem but also an information management one. This session will cover why the standards were needed, how they were developed, and how they are being implemented. While this is a case study of how this work was done at HBS, this session will present a search and findability framework (adapted from Morville & Callendar) which can be used at other organizations. The approach we used can be adopted/adapted for use in other organizations, and the audience members will most likely recognize their own roles in the “Who will be using the standards” section.

The key takeway of this session is that making information findable in an Enterprise setting is more than a matter of swapping out your existing search engine. Rather, the content and information also needs to be enriched and augmented and user needs analyzed in a standardized fashion to deliver a better findability experience.

Rails is not acclaimed for its performance. Neither is Ruby. The long history of "Rails doesn't scale" vs "Rails does scale" is finally coming to an end as everybody is realizing that it all comes down to you—the developer—and how you design and build your apps.

Fixing a performance bottleneck starts with actually finding it—a crucial, generally difficult task. In the past, you could use the benchmark lib. You could install RubyProf, MemProf, WhateverProf, inject code in all places you wanted to test, read huge stack traces, organize the results, and so on. If you used Rails' benchmarking/profiling tools... and then you'd be restricted to Ruby 1.8 or REE.

Not any more. RubyProf now has full support for Ruby 1.9 and a no "10-year terminal experience required" HTML output was introduced. Rails' benchmarking/profiling tools now support Ruby 1.8, Ruby 1.9, REE, Rubinius and JRuby.

So now, due to my recent contributions in Rails 3.1, you'll be able to analyze your apps' performance and receive feedback that's actually readable, allowing you to quickly spot most bottlenecks. You'll be able to keep using your Ruby interpreter of choice. You'll even be able to compare the performance of multiple interpreters, specifically for your app. And the best of it? It's damn fast, and dead easy.

This talk will walk you through the whole process: Installing all necessary gems, writing your performance tests, collecting their output and spotting performance bottlenecks. As a bonus topic, it'll also show you how you can easily assert which interpreter is faster for your app.

This talk is about using PostgreSQL together with set of tools developed at Skype (pl/proxy, pgbouncer, skytools) to achieve practically unlimited scalability of the database layer of a global scale web application. 2nd part is making this all easily available for even non-database programmers by seamlessly integrating with Django and/or Rails ORM's. Future prospects of using PostgreSQL server as _the_ application server are also presented.

The talk will cover the aspects related about the development of an open source WebConferencing system (BigBlueButton). It will begin with an overview about the project: features, architecture and integrations. The next topics will be focused on specific cases about the development process:

- Using Redis as Event DataStore and Messaging System
During the development of the record and playback feature, one of the tough decisions was choosing the system for store the data generated from a WebConference. This part focus on the methodology to use for recording and the communication of modules.

- API Development.
One of the key features about bigbluebutton is the API, which enables an easy integration with other tools and systems. This point will cover the key points like the security model, language adopted, and future plans.

The last part will be about some problems faced and solutions, in order to give and share the developer experiences, like the version control system, tools, and others.

Internet y la experiencia de usuario final en la web, han generado gran demanda de mejoras a distintos niveles partiendo por interfaces HTML enriquecidas, llamadas asincronas y optimizaciones en los tiempos de respuesta. A su vez la carga del lado del servidor exige un uso eficiente de los recursos disponibles, y cuando estos ya se acaban, tener la opción de poder escalar a nivel de arquitectura de servicios y proveer distintas técnicas como balanceo y caching.

Dado los requerimientos expuestos, a nivel del cliente HTTP se han generado grandes mejoras en temas de arquitectura, rendering y procesamiento de instrucciones como Javascript, un claro ejemplo de esto son los productos Google-Chrome y Firefox. Por el lado del servidor Apache sigue siendo la solución estable pero no óptima cuando es requerido funcionar de una forma eficiente, su arquitectura abusa de los recursos disponibles siendo una solucion genérica que pudiese abordar los problemas de carga de una mejor manera.

Como soluciones alternativas a Apache existen NginX y Lighttpd, ambos presentan una gran ventaja al momento de comparar resultados de pruebas de velocidad y consumo de memoria. Son una buena alternativa para servicios de alta producción, pero en estos tiempos existe un nuevo mercado donde un servidor HTTP es imprescindible: dispositivos móviles y empotrados, y es aquí donde los productos disponibles no son la opción deseada ya que el funcionamiento de alta producción y empotrado difieren en sus técnicas de optimización (dado su arquitectura).

Cómo construimos y hicimos crecer una empresa de consultoría Rails o "Como convencí a un científico espacial dejar el trabajo de sus sueños y construir una compañia de Rails"

Para: programadores de Rails empresarial, o interesados en construir su propio negocio.

La entretenida historia de la decisión de iniciar, y luego las subidas y bajadas de hacer crecer a Bendyworks. A través de historias de humor, compartiré las experiencias vividas y las lecciones aprendidas a través del camino.

El 27 de febrero de 2010 Chile fue sacudido por uno de los terremotos más grandes registrados. Al día siguiente a través de redes sociales se generó una convocatoria a los tecnólogos del país para aportar con sus habilidades y conocimientos en esta crisis.

Se enfrentaron diversos desafíos:

+ Durante la primera semana llegaron al lugar de convocatoria más de 100 voluntarios (entre los que se encontraban desarrolladores de software, desarrolladores web, diseñadores y comunicadores sociales) con quienes hubo que alinear objetivos y coordinar esfuerzos. y quienes en su mayoría no se conocían. La tendencia inicial fue a reunirse según profesión, lo que dio origen a los clásicos problemas de comunicación de un ciclo waterfall. Además hubo mucha rotación de personas, con el consecuente problema de gestión del conocimiento.

+ La coyuntura del pronto cambio de gobierno tensó al equipo, dado que llegaron señales de que debíamos alinearnos con una y otra posición, lo que exigió que contínuamente el equipo debía reunirse para adoptar decisiones estratégicas, con la consuecuente pérdida de foco.

Public authorities are starting to assume that they do not have exclusive use of public data. As main managers of public information, few agencies are hesitant about the fact that opening their data helps them to detect and correct errors, and also increases interoperability and create value by mixing different data sources.

Promoting the reuse 'culture' not only benefits the public sector, by improving its services and transparency, but also by means of its economic potential and job creation, business and society benefit from this.

The reuse culture will boost the appearance of many new products and information services, which will not come to light, or will not have any added value, without access to and reuse of public sector information.