:::: MENU ::::

Personal stats from 4 years at phpList

As announced today on phpList.org, soon I shall be leaving phpList.

For fun, here are some geeky statistics from my last four years leading the company:

  • Total emails sent from ‘sam at phplist dot com’: 5,070 (instant messages discluded)
  • My total commits across 21 phpList repositories: 1,726
  • My participation in the phpList community forum: 502 days interacted, 932 topics read, 701 posts created
  • Weekly all-hands meetings led: 200 (out of 205 weeks)

Note to self: code/hack less!


Introduction to Open Source presentation updated

We had some new starters join phpList recently, and as is sometimes the case, they did not have a long history with Open Source or a rock solid grasp of its heritage. Therefore I dusted off an old slide deck which I last used in a speech in 2012 in Liverpool, when I was recovering from a car accident (hence the black eye and crutches).

One quick review and spring clean later, the slides were used in an internal company presentation in our Tirana office during my last visit, and have been used since by Mariana Balla, our Community Manager, with new staff who’ve joined since.

In case it’s useful to anyone else, here are the slides embedded below. It’s high level, and emphasises the factual, historical, and legal roots of Open Source and the Free Software movement. I feel this is necessary in our modern climate of ‘open source planning’, ‘open source recipes’, and the general dilution of associated terms.

2019-05-07-open-source-intro-presentation


Open Source Underdogs Podcast Interview

It turns out that there’s a fascinating interview series which has interviewed tens of Open Source business leaders (a rare and rarely colocated breed), called Open Source Underdogs. The host and producer is a new friend of mine, Mike Schwartz — CEO at Gluu. CEOs of Canonical, Automattic, and Cloudera are counted among the interviewees.

Open Source underdogs podcast interview with Sam Tuke

And so it was that when we had the opportunity to record an interview about phpList at OSCAL Conference last month, we did so, and the resulting episode was published today. Listen to my interview here.

Mike has a small team which did a great job of editing our discussion. Two particularly interesting themes that came up were the Ethics of Open Source (an old but still appealing subject), and the potential for data-based business models for Open Source. Maybe I’ll explore these topics in more detail in dedicated talks sometime. Let me know if you’re interested!


Open Source Product Development: from research to release

Later this month I will be speaking at the OSCAL 2019 conference about Open Source product development. Here’s the abstract:

In 1974 Fred Brooks argued that an app requires nine times extra work to take it from merely being functional, to being a usable product. This challenge is as real now as it was then. What is this extra work, and can you do it using pure Open Source?


At phpList we manage both hosted and downloadable products, from conception to release, using Open Source tools. That includes usability, acceptance, and A/B testing, project management, and more. In this talk Sam will provide a high-level overview of powerul Open Source systems for managing modern software products, including a specific workflow which uses them.


Whether you are creating a web app on a budget, or wanting to make an Open Source project the best that it can be, gain insight into how Open Source can support and empower your workflow.

The slides from this presentation are embedded below.

2019-04-05-sfk-presentation-public


Software Freedom Kosova 2019 Conference

One of the most unusual cities I’ve presented in, Pristina in Kosovo again hosted the Software Freedom Conference (SFK), thanks to the youthful, hardworking volunteers at the Free Libre Open Source Software Kosova (FLOSSK) organisation.

This year I participated in a panel, and spoke about Open Source Product Management workflows. The panel proved interesting, and was entitled ‘Doing Business with Open Source’, attended by local software business leaders spoke on Product Management workflows using Open Source. The Kosovan Minister of Innovation sat in the front row, which made for a more diverse audience. As an outsider I felt the need to correct the other speakers on a few occasions, particularly when one said that Open Source was known to be less secure than proprietary software (a long debunked idea), and also when another said that Open Source was rarely used in Kosovo (patently not the case).

The spirit of the conference was fantastic: lots of energy, many students from different faculties, some academics from local universities, with diverse interests and conversations. The talks were held in a classic cinema, recently refurbished, smack in the center of the city, and the stage was managed by television professionals. Combined with talks being held at night, it made for a very different atmosphere to other tech conferences, and a wonderful cosy and relaxed vibe. It helped that free beer on tap in the foyer was served throughout.

And so I was delighted to have participated again, meet fascinating people (including one ex-Kosovan refugee who had traveled back from the United States to attend), and shared some ideas.


Reading PHP serialized data quickly in a Linux Terminal

Here’s a quick script for reading a data file containing serialized data generated by PHP, and outputting it in a human readable format to a Linux command line interface. A short web search found no existing utilities for achieving this, so hopefully it will be useful.

You must use absolute paths (using tilde also expansion works, of course). Save this as a file called deserialize-php, make it executable, and move it into your /usr/bin folder to make it universally accessible.

#!/usr/bin/env php
<?php
 
// Enable error reporting
error_reporting(E_ALL | E_STRICT);
 
// Get command line arguments
$userArgs = $argv;
 
// Check that an argument was provided
if ( !isset( $userArgs[1] ) ) {
    echo "\nNo path submitted: a path to a serialized PHP data file is required'\n\n";
    die();
} elseif ( '/' == substr( $userArgs[1], 0, 1 ) ) {
    // The supplied path is absolute
    $serializedPath = $userArgs[1];
} else {
    // If the supplied path is relative
    echo "\nRelative path supplied but not supported, please provide an absolute path\n\n";
    die();
}
 
// Get the serialised data for processing
$serialized = file_get_contents( $serializedPath );
 
// Unserialize the data
$unserialized = unserialize( $serialized );
 
// Print the data array in readable form
print_r( $unserialized );

Then you should be able to use it like this:

$ unserialize-php serialized-data.php

“Nobody said Open Source makes you rich” and the economics of Open Source

Inside the historic pyramid building in Central Tirana (more recently museum to the nation’s life-long dictator), I concluded my speech on economic challenges to Open Source companies, and raised my eyes from the notes on my tablet to see a sea of hands.

I felt both nervous and exhilarated at the idea that those before me were about to challenge the assumptions underlying the mostly pessimistic conclusions I had drawn about the immediate future of Open Source businesses. Belonging to other conference speakers, those hands prodded at something else entirely however.

“Nobody ever said Open Source would make you rich”, one hand-waver said. “You should create Open Source because it’s the right thing to do, not for the money” said another. Initially I was speechless. As I stood there, under aggressive lighting rigged for TV confounded by this reflected message which I had never considered.

I realised that I had overlooked something fundamental in my presentation, something that was now difficult to concisely convey. So I shall elaborate here as briefly as possible.

Assumption 1

  1. Start-up businesses are a valid means of developing Open Source software

More tenuous than I first realised, but bear with me.

Reasoning 1

  1. A fair day’s work deserves a fair day’s pay
  2. People who labour for the common good are no less deserving than those who labour for themselves
  3. Great achievements require many hard days work
  4. ∴ Achieving great things fairly requires a considerable amount of pay

While it’s true that many Open Source applications are the by-product of other activities, including proprietary applications, that’s not a model I want to explore here.

To draw the same conclusion with alternative reasoning:

Reasoning 2

  1. Software startup economics demand both ‘hyper-growth’ and the investment required to fuel it
  2. Being runner up in a market subject to strong network effects is tantamount to failure
  3. ∴ Major growth targets and major investment are mutually dependent requirements for software startups

In other words, significant investment, revenues, and growth, are all necessary for a successful Open Source startup.

Dirty money?

“Wealth” and “significant investment and revenues” are for many people interchangeable labels. But clearly organisational wealth, and personal wealth are entirely different things. Despite the distrust of Corporate America by generations of Open Source advocates, faith that well-funded firms may conceivably  work for their benefit is reasonable and necessary. One can “be big to do good”. I believe it’s our best hope.

As somebody who has been advocating Open Source for half my life I assume that others like me see wealth, and particularly organisational wealth, as a means to increase Open Source adoption. The sharp lesson learned in Tirana was that wealth is an emotional issue, even in the abstract; even when it’s harnessed for things people love, and whenever I speak about it, the first crucial step is reframing its power as an engine for freedom.


Sam Tuke presenting 'The Economics of Open Source at OSCAl 2018 conference

Sam Tuke and Suela Palushi at OSCAl 2018 conference


Sam Tuke presenting 'The Economics of Open Source at OSCAl 2018 conference


Sam Tuke presenting 'The Economics of Open Source at OSCAl 2018 conference

 
Slides from my talk are embedded below.
2018-05-18-oscal-presentation-public


A primer on Open Source business models

Ahead of the upcoming panel discussion ‘Building an open source business‘ at FOSS Backstage run by Jos Poortvliet and I, it’s useful to review the commercial models most commonly applied by Open Source projects. These can provide a framework of understanding and common vocabulary to facilitate effective debate. Originally published on the FOSS Backstage website.


Among the hundreds of Open Source businesses operating today a handful of commercial models are reused. Erik S. Raymond described thirteen models in his seminal 1997 publication The Cathedral and the Bazaar. While these models were largely theoretical, hist list is comprehensive and insightful.

Models described by Erik S. Raymond
• Cost sharing • Free the Future
• Risk spreading • Sell the Present
• Loss-Leader/Market Positioner • Free the Software
• Widget Frosting • Sell the Brand
• Open a Restaurant • Free the Software
• Give Away the Recipe • Sell the Content
• Accessorising

A more scientific approach was taken by Linus Dahlander in his 2005 paper Appropriation And Appropriability In Open Source Software. The models were reduced to five, grouped into three types of offering (‘Software as a Service’ does not appear in Linus’ original table and has been added).

Simplified models by offering
Type Sub-category Explanation of how it works in OSS
Products Licensing Licensing the right to use the software, i.e., adding a proprietary part to the open code or allowing the customers to use the source code as they wish
Black-boxing Bunching several pieces of OSS in a hardware solution
Services Consultancy Consultancy work based on an area of expertise, be it a product that the firm releases or a community-established project
Education Education based on an area of expertise, be it a product that the firm releases or a community-established project
SaaS Software as a Service, access-based pricing to software products, providing recurring revenue
Support Support Support based on an area of expertise, be it a product that the firm releases or a community-established project

Because most software companies rely on controlling access to the binaries and source code of their applications, and Open Source firms necessarily give away this control voluntarily, an existential question for these firms is how they capture, or appropriate, the value of their offerings. Linus summarises the range of ways Open Source businesses extract value from the software they produce.

Means / protections for appropriating value
Category Sub-categories Explanation of how it works in OSS
Patents (IPR) Institutional protection in terms of temporary monopoly granted to novel, useful and non-obvious innovations. Often granted to algorithm in software, but functions is also used and heavily debated
Copyright (IPR) Institutional protection that grants creators exclusive right to reproduce, prepare derivative works, distribute, perform and display the work publicly
Secrecy Keeping secrets within the firm, primarily by closing the code
First-mover advantages Network externalities, first-mover advantages Early entry to the market, which can create advantages by acquiring superior resources and capabilities
Complementary assets Getting a large user base using complementary assets such as distribution, marketing in conjunction with the innovation

Taken together these commercial models and means of appropriation form a convenient framework for understanding the strategies of leading Open Source firms such as these.

Models and means of appropriation of Open Source firms
Company Offering Model Means of appropriation
Red Hat Service Support Complementary assets; First-mover advantages
Jenkins Service SaaS Secrecy; Complementary assets
phpList Service Saas Complementary assets
NGinx Product Licensing Secrecy; Copyright
Nextcloud Product Licensing Complementary assets; First-mover advantages
Mautic Service Saas Complementary assets
openNMS Service Support Complementary assets
Collabora Service Consultancy First-mover advantages; Complementary assets

For a different take on these models, consider Mozilla’s ‘A framework For Purposeful Open Source‘ – a 40 page document published last month.


Can I make UK phone calls using Amazon Echo or Google Home? And other questions

Update (March 7th): Google Home devices can now make free UK calls

This Christmas an elderly relative has requested the gift of hands-free, voice-activated calling via a home virtual assistant. Much has been written about Amazon’s Echo and Google Home, and the tech press are awash with conflicting information and feature speculation. As Christmas approaches and confused buyers respond to Amazon’s advertising, I’ve provided answers for shoppers with elderly or manually impaired gift recipients in mind

Disclaimer: Both devices run proprietary software which transmits personal information into the cloud under terms which present a threat to privacy.

Calling

Q: Can Amazon Echo devices make calls to UK mobiles and landlines?

A: No. Currently Echo devices can only call other Echos, and people who have the Alexa App installed. Contacts must also be synchronised with the device doing the calling before a call can be initiated. Amazon announced in September that free calling from Echo devices to mobiles and landlines is now available to customers in the USA, Canada, and Mexico, but not the UK. Presumably this will come later.

Q: Can Google Home make calls to UK mobiles and landlines?

A: No. Google Home can currently only make calls from the USA and Canada. Calling has reportedly been activated and deactivated by Google for some users, but Google has yet to officially launch it and enable it for all customers.

Standalone operation (without a smartphone)

Q: Can Amazon Echo devices be used without a smartphone?

A: Yes, but only following initial setup and synchronisation with a smartphone and the Alexa App. Once setup is complete, a smartphone is not required. However to update contacts or reset the device a smartphone with the app will again be necessary.

Q: Can Google Home be used without a smartphone?

A: Yes, but similarly to Echos, it must first be setup and synched with a smartphone, and resynched to updated the contact book.

Conclusion

As things stand, neither Google nor Amazon offer proper calling of any kind for UK customers, nor standalone operation. With assistance during setup, and assuming that important contact numbers don’t change, Amazon Echo devices can provide limited app-enabled calling. Personally I wouldn’t rely on this as a primary communication channel.

Over the next few months we can expect that both companies will add free UK calling to their devices. It’s unlikely that their dependence on smartphones for setup will change however.


Free Software as a Service presentation at 33C3 Hamburg

What are the unique challenges and opportunities for businesses which sell Free Software as a service? This was the question that I tackled in a presentation at the Chaos Computer Club Congress this month. Following a period of reflection on the economics of different Open Source business models, and which requires the lowest risk with the highest chance of success, I wrote this talk and spoke in Hamburg on December 28th.

sam tuke presentation at 33c3 hamburgsam tuke 33c3 presentation slideshow

The audience was brilliant: the half hour Q&A which followed was wide ranging with 6 participants identifying themselves as Open Source business owners themselves. One comment which I’ve dwelt on since was “what’s the best way for the Open Source community and your community of customers to interact?”. Getting the answer to this right is both hard and important — it’s easy to keep these two groups of key stakeholders separately siloed, and lose the great synergy they could create. What could be the benefits of these two groups, both experts on your product, meeting and sharing knowledge and ideas?

33C3 was my first Congress and won’t be my last. After eight years of speaking at technology conferences it’s rare to have an entirely new experience, but Hamburg was that. Congress provides a sort of global leadership for the hacker scene; it both reflects and sets the focus of a wide range of technology communities, and is of unique benefit to Germany as a beacon of IT innovation and expertise.

33C3 Hamburg conference venue

33C3 Hamburg venue at night by doctore_