This section of my blog is all about being a Geek. I develop websites in Drupal, run LAMP webservers, and enjoy all sorts of software and gadgets.
Dries Buytaert: The State of Drupal
DrupalCon Paris is underway now, with Dries Buytaert cutting a Cat5 cable instead of a ribbon to officially open proceedings.
He's now speaking in his role as the project founder and lead about the 'state of drupal'.
And first, the subject of the code freeze for Drupal 7 - he's describing the waves of drupal development work, and how it builds up towards code freeze into a patch frenzy - and how developers complain.
He shows a quote:
Paul Saffo: "People overestimate short-term benefits of innovation, but under-estimate long term benefits".
He talks about Gartner's 'hype cycle' which starts with an innovation trigger, early adopters, then positive hype, peak of inflated expectations, negative hype, trough of disillusionment, slope of enlightenment, then settles into a plateau of productivity.
He says that drupal 6 has settled nicely into the plataeu of productivity.
He says that Drupal7 needs to be at the height of the curve of inflated expectations before code freeze (which prevents new features from being added). The question is, is it there?
Dries outlines some of the changes made. One of the highlights of this is that Fields API is now in core (essentially CCK. Tacxonomy terms and node bodies are now fields. JQueryUI is in core. There are new themes, and much much more.
Dries says the code is going to freeze - but some things are half-baked like Fields API, Usability improvements, test coverage - and more importantly performance has degraded.
So - code freeze is extended one week to next Monday to allow people to work on D7 udring DrupalCon Paris. The second phase is named 'code slush', which is time limited to 5 weeks. No new features or functionality will be allowed - except 10 carefully chosen exceptions. Important API changes for existing features will be allowed. Usability, testing, documentation and performance improvement changes will be allowed. The API and Schema freeze will be on Oct 15th.
The 10 exceptions are: Imagefield, Field translations, converting profile module to field API, convert taxonomy terms to fields API, overlays, edit anywhere, shortcuts, dashboard, plugin manager, RDF/RDFa.
Then the last phase is 'strictly polish' - small usability, testing, documentation and performance changes only. This is time limited to 4 weeks.
The final phase is bugs and release blockers only. After this, Drupal 7 will be released when it's ready - hopefully early 2010.
He shows a quote from Joseph Schumpeter who says that Radical innovation drives creative destruction.
Dries says that everything goes from being an innovation to simply a service - even computer hardware is doing this. It starts as an innovation, then is available as a bespoke product, then as a general product, then as a service. That's when innovations mature.
Dries says he likes this. Drupal distributions are the way to go. He says install profiles are becoming modules. Plugin manager will also be in core - allowing easy module installationa nd upgrading. It's also worth looking at Features.
He says richness and reach equal success. Drupal has richness, but not yet reach. Distributions will achieve that reach. He believes drupal is a framework AND a product, and wants to develop both.
For drupal 8 he wants to consider make more of a distinction between the two in the codebase - so you could just install the 'framework' if you want.
He says that we need to move from simply innovation driven development to market or user driven development. Drupal 7 usability is a good example of this.
Quote: "If you're not afraid, you're not doing the right thing." But the result is some pain.
The usability created friction between designers and developers - but this shows it's achieving something - and changing drupal. Drupal is growing up, and this is part of the reality of growing up. We have no choice but to grow up.
We have to reach out to new people and learn how to interact with them.
We need to increase the reach - its our biggest challenge and biggest opportunity.
There's some discussion/argument in the questions about how the process of the usability changes was handled.
There's a question about the upgrade path, and whether it'll be easy. Dries says there'll be a data upgrade path - but API's will break - so custom modules need to be updated.
In other announcements:
NowPublic has been acquired by Examiner. An ironic new drupal adopter is the Java.net community of Java developers!
And that's the end of Dries' presentation - now on with DrupalCon!
Facing up to Facebook's Books
Are you on Facebook? Great - now, when was the last time you paid them any money?
The answer, of course, is never. As with many companies on the web they don't charge their customers for the service. Instead they have to try to find other sources of revenue, such as advertising - but that's a tough market at the moment.
However, Facebook has now reported the impressive news that it is now EBITDA positive (that is, it makes a profit before taking into account certain financing costs. It stands for Earnings Before Interest, Tax, Depreciation and Amortisation). In Facebook's case the key figure is depreciation - because it is making so much capital expenditure on the technology infrastructure to run such a rapidly growing network. Because of this it's still not cashflow positive.
TechCrunch reports:
The company is likely spending well over a $1 million per month on electricity alone, say experts we
Twitter and the Evolution of a Business Idea
The images above trace the history of Twitter, the hugely successful (in terms of fame rather than profit at this stage) micro-blogging service.
I first heard about Twitter in early 2007, and opened an account (managing to get my name without any additional numbers etc at that stage!) - but never really got around to exploring it further.
As part of the upgrade to this blog I decided it'd be good to also be able to post regular, small updates - without cluttering up the main blog. So I decided to figure out how to Tweet, and you can now see my Tweets in the right hand sidebar of this blog.
In doing that I became interested, as always, in the genesis of the business behind the service - and here is the story...
The Idea
Everything starts with one person and their idea. In this case the person was Jack Dorsey [PHOTO 1], otherwise known as @jack to Twitter folk.
Here's how he describes coming up with the original idea:
On May 31st, 2000, I signed up with a new service called LiveJournal. I was user 4,136 which entitled me a permanent account and street cred in some alternate geeky universe which I have not yet visited. I was living in the Sunshine Biscuit Factory in Oakland California and starting a company to dispatch couriers, taxis, and emergency services from the web.
One night in July of that year I had an idea to make a more "live" LiveJournal. Real-time, up-to-date, from the road. Akin to updating your AIM status from wherever you are, and sharing it. For the next 5 years, I thought about this concept and tried to silently introduce it into my various projects. It slipped into my dispatch work. It slipped into my networks of medical devices. It slipped into an idea for a frictionless service market. It was everywhere I looked: a wonderful abstraction which was easy to implement and understand.
The Sketch
Jack likes doodling on yellow legal pads, and that's where Twitter first emerged from his head [PHOTO 2].
At this stage his name for the site was 'stat.us' using the '.us' domain suffix instead of '.com' because he thought that was cool! The evolution of the brand name is one of the interesting things about Twitter - morphing through a number of versions, just like the design of teh site itself and the feature set.
But Twitter today isn't that different from his original sketch.
The Detail
Jack worked for a company called Odeo (now called Obvious after it sold the website called Odeo.com). The company was run by Evan Williams, who is a great entrepreneur and may merit his own blog post sometime.
His masterstroke was creating a company that encouraged doodling, pet projects and trying out new ideas, even if it wasn't 'on focus'.
Jack toyed with the Twitter idea some more, and wrote out the list in [PHOTO 3] in early 2006. It's not very clear in the photo, so here's what it says:
casual awareness.
"what are you up to?"multiple entry point to set status
- web
- phone
- sms
- immultiple ways to "subscribe" to status
- web
- phone
- sms
- im3 aspects
- set status
- timeline (collaborative)
- configuration
In a very simple list Jack set out what is still recognisable as the way Twitter works. It doesn't take a 50 page proposal. So far we've seen two sparsely covered pages of legal paper - and we get the idea.
Jack wrote at this time:
The 6th year; the idea has finally solidified (thanks to the massively creative environment my employer Odeo provides) and taken a novel form. We're calling it twttr. It's evolved a lot in the past few months. From an excited discussion and persuasion on the South Park playground to a recently approved application for a SMS shortcode. I'm happy this idea has taken root; I hope it thrives. Some things are worth the wait.
The Prototype
Jack had persuaded Evan to give him the chance to create a prototype of the site, and with help from colleague Biz Stone they created what you see in [PHOTO 4]. It's very basic, but it does the job, allowing teh concept to be tested. In a future blog I want to write a little bit about 'Agile Development', but one of the key concepts is to test the idea as quickly as possible in the simplest way possible. This is what Twitter did.
The Launch
This user interface then evolved into [PHOTO 5] to gain a bit of design and colour (but it's still very basic) - and then into what you see in [PHOTO 6] which is my page on Twitter today.
You can see that it really is an evolution. They get it started as quickly as possible, and then constantly refine it. I really believe that is the best way to go.
But it's rare for entrepreneurs to let the actual brand name evolve, but I think they were right to do so. Talking further about this, Jack says in a recent interview with the LA Times:
The working name was just "Status" for a while. It actually didn
Click For RSS Feed