Site architect

Overview
In this roundtable format we'll discuss the unique challenges of penetrating the enterprise IT market for CMS. An architectural approach is essential to overcoming these challenges. We'll discuss our experiences in delivering solutions in large enterprises and try to identify a handful of architectural best practices for positioning Drupal as the right choice.

Agenda

Technology adoption patterns within large enterprises are complex and often focus on risk mitigation over value creation and agility. Breaking out of this trap and presenting Drupal as ready for the enterprise requires some knowledge of Enterprise Architecture. We'll discuss experiences using different techniques such as

* Requirement Analysis
* Decomposition
* Modeling
* View point hopping
* Benchmarking
* Running Proof-of-Concepts

As a jumping off point, we'll reference certain tools and studies sponsored by Sun Microsystems for benchmarking and designing Drupal architectures.

Goals

Attendees will meet others interested in enterprise adoption of Drupal, and can build the community focused on this market. Ideally, a set of best practice architectural approaches to positioning Drupal in the enterprise will come out of this session.

Overview

A survey of open source tools for improving efficiency of Drupal coding, deployment, testing, and system resource utilization.

Agenda

In this session, we'll walk through some typical uses of:
* NetBeans PHP and Drupal plugins
* Deployment to cloud computing service (Network.com)
* DTrace PHP and MySQL probes
* OpenSolaris Zones, ZFS, and integrated AMP stack

Goals

Attendees will get a good intro to the open source technologies coming out of Sun that are most relevant to Drupal.

Overview

Storage and retrieval of RDF data represent significant departures from traditional relational database semantics. While many APIs exist for storing RDF data in a relational database, they tend to be slow, lacking in data integrity enforcement, or difficult to query.

In this session, we'll consider options for a scalable, easily queryable Drupal RDF system.

I'll bring some of my own ideas, but everyone's invited to bring their own, as well as RDF use cases to consider.

There is already a session about Facebook Application development. The session is scheduled for 11am on the 28th. At that time I'll talk about what a Facebook App is and tools for building them on Drupal.

This BoF will be a chance to work with the modules and ask questions specific to your own site. Perhaps even get your site running on Facebook, if you have not tried to do so already.

I'll be there to answer questions and help troubleshoot. So this is a good chance to get started if you haven't worked with Facebook before. And a good chance to ask questions if your stuck on any particular feature.

If you want to contribute to the Drupal for Facebook project, we can talk about the best ways to do that.

Overview
This talk gives an introduction to Bazaar (bzr) - the distributed source code revision control system developed by Canonical Inc.

Agenda
In this session, Lenz will provide an overview about the general concepts of distributed source code revision control and how Bazaar (bzr) fits into this picture. The bzr terminology and most useful commands will be explained as well as examples on topics like:

• Creating a repository
• Adding files
• Editing files and commiting changes
• Branching
• Plugins and extensions, Interoperability with other SCM systems

Goals
Attendees will have a better understanding on how distributed revision control works in principle and how it can be utilized with Bazaar. The basic commands needed to get going will be covered, so users can get familiar with the concept by themselves.

Resources
Some basic understanding of source code revision control systems (e.g. CVS, Subversion) is probably required to make the most out of this session and to understand why Bazaar (or any other distributed revision control system) is superior to these.

Overview
This session will be a brainstorm on VAT, taxes and price modifications in übercart.

Theres about 32123 or so different rules for taxing worldwide and each country have its own obscure rules...
And then theres the growing need for individual pricing of product to a specific client base (read role), and offcourse combinations of this
and .... [insert you own need for modification of a price ;) ]

But no more will we struggle with this, come join the battleplan + a presentation of a possible solution for these problems

Agenda
* price based on a role
* tax modification - different country different rules
* short presentation on a "modify-price-by-tax-or-role-or module"
* anything else?

Goals
To create the end all be all solution for displaying and working with price in übercart.

Overview

A "Cloud" offers a virtualized datacenter infrastructure that allows you to build your own network applications. In this session, we'll cover an approach to implementing Drupal in the cloud using the popular Amazon Web Services as the cloud service.

Agenda

What with this cloud stuff?

An overview of what clouds are and what they're good for.

Drupal on Amazon Web Services

APIs, tools, and techniques.

Persistence issues and solutions

Where to store /files & database backups.

Cluster management overview

Clusters go great with clouds. Here's some cluster tools and services, some cloud specific some not.

High availability in the cloud

The issue of reliability in the cloud, and an overview of deploying redundancy and failover

Auto-scaling in the cloud

Killer app of Clouds is being able to grow your site on demand; here's some popular ways to do it.

Future of clouds

Providers other than Amazon, DIY clouds, and more.

Goals

After this session you should have a good idea of the possibilities available to you when deploying Drupal in the cloud, and a good enough technical understanding to deploy a Drupal server in the cloud.

Resources

You can get a general understanding of cloud-based deployment with only a general understanding of the LAMP stack; to get the most out of this session you should be familiar with setting up Drupal on a fresh linux install using only the command line.

Overview

Relational databases store, index, and retrieve data using using predictable patterns. Indexing data well -- but not excessively -- requires understanding indexing overhead and usage by query execution planners.

But, even the perfect indexing plan layered on top of existing tables cannot satisfy all needs: relational databases have significant limitations for indexing data. Particularly, they cannot index data across multiple tables. While Oracle and Microsoft SQL Server provide some in-built tools (materialized views and indexed views) for alleviating this limitation, users of open-source databases like MySQL and PostgreSQL must consolidate and preprocess data to scale popular services.

The Denormalization API streamlines the process of consolidating node data into tables structured for radical scalability.

Agenda

* How is Drupal data stored?
* Overview of indexing tables
* Index data structures
* Query execution plans and indexes
* What is denormalization, and why is it necessary?
* Typical hurdles for implementing denormalization
* Using the Denormalization API

Goals

Attendees should leave with an understanding of the benefits and caveats of indexing and denormalizing data. And, if they choose to denormalize, how the Denormalization API can streamline the work.

Resources

Attendees should be familiar with Drupal's node system, SQL, and basic database administration.

Overview

The stock Drupal core has a number of bottlenecks and limitations for high-traffic and enterprise deployment. Many of the top Drupal sites maintain internal, patched versions of Drupal. While each of these internal versions is custom-built, the patches applied and techniques used to maintain the patches are consistent across these sites.

Agenda

* Popular patches used by high-traffic sites
* When these patches will be in core (or why they're not in core)
* Patch conflicts to watch for
* Deployment strategies to eliminate or minimize downtime caused by necessary changes to support the patch changes
* How to use version control systems to efficiently maintain a patched core without falling behind when the standard core applies security and bug patches. The demonstration will be using bazaar-ng (bzr).
* Other community resources for making this work easier
* Case studies in patching core for performance

Patches/modules considered

* Master/slave replication
* memcached
* Cache Router
* Removing LOWER()
* Database lock removal
* Possibly others

Goals

Attendees should leave with the following capabilities:
* How to decide whether patching core is worth the trouble
* How to maintain a patched core

Resources

Attendees should be familiar with applying patches and using version control systems.

A community effort to help a NGO get some ass-kicking internet presence

On the first weekend of October 2008, in only 2 days, the Drupal French Community will build a complete and live website for a selected NGO. All free. Free as in free beer, and free as in free speech.

A lot of the organizational details have been ironed out.

Though we have discussed a lot the question of "how the hell do we pull this off?" and have a detailed an action plan... there must be a million things we forgot.

Who should come?

So we are calling on all that have experience in organizing Codesprints/ Hackathons/ Mashpits and generally community events to join us and share their experience and thoughts.

What should come out of the session?

We would really like to hace constructive criticisms on our plan to adjust and augment it so we can put all the chances on our side to pull off this event.