Anton McConville
Digital Designer & Developer
Beyond Vegan Food

Last year seemed a sort of turning point for vegan food consumption, because in no small part of the phenomena known as ‘Beyond Meat’. Beyond Meat is an engineered, entirely plant based replacement for hamburgers, sausages and even ground beef in burritos.

It is cropping up at restaurants like A&W, Mucho Burrito, Tim Hortons, Burger King and more as well as in chilled cases at pretty much all of the grocery stores. It’s a good news story. Maybe even a great one I think. At least the A&W burgers taste great.

As a vegan, I was as excited to taste it, cook with it, and learn more about it. I like it, and I like that there are so many more options for vegetarians and vegan fast food these days too.

Plant Based Economy

I’ve been vegetarian a long time, and over the years my diet has shaken out a few favorite substitutes for ‘regular everyday food’ too … for example:

I’ve been taking more interest in the economy around plant based alternatives recently, and started looking at some of the companies behind these products. I was interested thinking about possibly ethical investments, but also curious to understand motivations and ambitions … here’s what I found from a little surfing …

Califia Farms

I drink plant based milk everyday - in tea, or shakes, or over cereal. Recently I’ve been drinking a lot from Califia Farms, in part because I their milk has been discounted a lot in the stores I shop in, but mostly because I really like the taste … especially their Barista Blend in tea or coffe.

I found this press release about them which show investments from companies in Belguim, Singapore, Hong Kong, Canada Qatar, Brasil … which I found interesting, given the very Californian name of the brand! In that press release they talk about disruption of the trillion+ dollar dairy market.

Plant based milk is controversial too, in that the growing demand for it has resulted in over farming in drought struck California. I’ve excerpted this from an article in the Guardian:

Pre-shipping, the carbon created by a litre of semi-skimmed (1.67kg) is far higher than that of almond milk (360g). “But what people don’t know is the environmental damage almond plantations are doing in California, and the water cost. It takes a bonkers 1,611 US gallons (6,098 litres) to produce 1 litre of almond milk,” says the Sustainable Restaurant Association’s Pete Hemingway.

Over 80% of the world’s almonds are grown in California, which has been in severe drought for most of this decade. Hemingway describes a situation in which farmers are ripping up relatively biodiverse citrus groves to feed rocketing demand for almonds, creating a monoculture fed by increasingly deep water wells that threaten statewide subsidence issues.

My verdict - I love this milk, and want to keep drinking it. I have concerns about the gold rush for non-dairy alternatives, and a little suspicious of the global interest in this company. I also feel guilty contributing to the overfarming in California. Sometimes it seems there is no right answer. Will keep watching.

Fieldroast - MapleLeaf Foods

Fieldroast Chao cheese is the only product I eat from them - because it is the closest I remember cheddar cheese slices feeling and tasting.

A little bit of research revealed that Fieldroast are now owned by mega meat processors Maple Leaf Foods. Who are evolving to become a ‘protein’ company as opposed to meat company.

My verdict - I worry about this product most of all, because of the parent company. I support the product, in the hopes that it offers more people a path to veganism, but I’m kinda looking for a decent alternativea


Vega One is my goto quick/convenient and refreshing protien packed start to a day. I love it.

A little research reveals that vega is now a part of Danone - a global dairy brand. Like Maple Leaf Foods, they’re betting on both animal and plant based products.

Last year I converted to veganism because I’d learned how abusive the milk industry is - so it’s a struggle to me that giant dairy makes my vegan shakes.

*My verdict - am looking around for a good alternative, but not in a huge hurry right now.

Sol Cuisine

Sol Cuisine make delcious beet burgers - I feel at least as tasty as the beyond meat ones.

Sol Cuisine seems privately owned by Canadian equity firms who invest in agriculture.

I feel good about Sol Cuisine at this point :)


Tofurky seem to be privately owned. It has an interesting background, and seems genuinely earth loving :)

I feel good about Tofurky products


I really didn’t mean to write so much in this blog post - just intrigued about the backgrounds of the companies making all of this processed vegan food these days. It is worth thinking about.

In general I eat vegan blts maybe once a week. I eat a veggie burger at home once a month maybe. I drink plant milk, and a vega shake almost every day. I eat vegan cheese once in a while.

Like any processed foods - too much of it is not a good idea. I want to focus more on naturally made things starting with real plants, as opposed to the processed kinds.

I’m glad that so much more attention is being placed on vegan diets now … but conscious that this food isn’t what being vegan means.

Kubecon North America

Writing this from San Diego, at the end of another intense KubeCon conference. I was here as part of IBM’s staff, to present a couple of sessions at the mini theatre, and to take a shift at the IBM booth, but when not taking my turn at the booth, I attended as many talks as I could. Here are my notes …

General themes - use of service ServiceMesh ( mostly Envoy from the talks I was able to attend ), economics of cloud services, and machine learning - both in terms of devops, and in terms of building ML on Kube.

Measuring and optimizing ML Workloads at Lyft

The first talk I attended, and I think the most inspiring one that I attended … definitely has me thinking about a lot of different things.

The essence of the session was about how at Lyft they’ve begun using Machine Learning to understand their infrastructure bill better. They build on AWS. Estimating, and justifying cloud usage fascinates me. From this talk I feel we’re still at the finding a way of accounting, and navigating the economics of cloud computing.

Themes that stood out to me - attribution schedules, allocation of resources, versus efficiency of use. Different workloads have different needs, at different times.

Aside from the content of the talk, the presenter was outstanding - although just a recent hire, his clarity of communication, and depth of understanding of this topic was super impressive.

How Spotify Migrated Ingress HTTP Systems to Envoy

There seemed a lot of talks about Envoy this conference. I’m personally still trying to understand the differences/similarities between Envoy and Istio

I loved hearing these insights from some of the companies that serve millions of users …

Spotify serve 8 million requests per second. Their system serves 79 markets, 248 million monthly users and is built on 1200 microservices - almost all on Google Cloud.

This was a well delivered talk that walked through the problems they were attempting to solve, why they chose Envoy to solve them.

Before switching to Envoy, they felt their old gateway Was

  • underinvested
  • had service discovery headaches
  • had high operational conversations

With Envoy …

  • a cloud native perimeter
  • vibrant community
  • service discovery

They implemented and moved in a way where product teams needed to understand as little as possible about the switch. They built gradually, with minimum disruption to their dev teams and end users. They facilitated a quick fallback in case of problems.

Moving from Legacy Infrastructure to the Cloud in a Government Organization - Chris Carty, City Of Ottawa

Happy coincidence for me that a senior engineer from my own city’s municipal government was presenting their story of evolution to the cloud.

He did our city proud, with his talk too - very well presented and organized story - a checklist of considerations for any government or any organization in moving from mainframe based solutions to kubernetes.

In Ottawa’s case the city employees 17,000 people, has 120 lines of business ad 400 apps to support ( Java, DotNET and Perl ). Some of the mainframes are end of life, and rather than renew them, they’ve begun migrating solutions to Microsoft Azure.

I recommend watching his talk, or asking for his slides - he has really beaten a path for others to follow if they’re embarking on a similar journey.

One point of interest for me, and I raised a laugh in asking, tongue in cheek … was what this means for me as a tax payer. I am curious about how they evaluate costs … so they are no longer paying for mainframes, but does that mean they work with the same budget for cloud resources? Have they modeled or estimated their use. Is this a net savings, or a net add? The presenter said they were not there yet - in estimating or projecting.

Flyte: Cloud Native Machine Learning & Data Processing Platform - Ketan Umare & Haytham AbuelFutuh, Lyft

I’d heard about Flyte and wanted a bit more insight.

Lyft wanted a way to build reliable, scaleable, reproducable models … especially when it came to reporting on business results internally every quarter. They had lost a key data scientist who had a lot of intellectual property on their own laptop, and had been the only data scientist using it at the time.

This talk was a walkthrough. I am super interested in machine learning, and really want to get into it, but have far too many interests, and now that I’m managing a team, still creating case studies, and still learning more about the cloud, I need to be selective about where I invest my time and how. I’m not sure Flyte is my entry point … I want to learn more about the basics of tensorflow and models. However, I think this is a good overview if yoy want to learn more check out the video when it is up.

How Yelp Moved Security From the App to the Mesh with Envoy and OPA - Daniel Popescu, Yelp & Ben Plotnick, Cruise

A really well presented talk about Yelp’s adoption of a service mesh to improve security of their apps. Really well presented :)

Takeaways - moving from a 15 year old monolith to microservices for authentication and authorization. Respecting developers who would consume Yelp through integrations - always considering the restful service pains they would need to suffer with any architectural decisions they made.

Again Yelp adopted Envoy in their solution for this. Although one of the engineers began the talk by wishing that Istio existed before they started their migration of services. I wanted to ask why he would have preferred that. Maybe I’ll send a message to him.

Things to look up from this talk: Open Policy Agent

Executive notes:

  • Security in service mesh makes Yelp services secure by default
  • Work incrementally ( same approach as talk above from Spotify )
  • Automate migrations
  • The use case is the north started

Security Beyond Buzzwords: How to Secure Kubernetes with Empathy? - Pushkar Joglekar, Visa

Another wonderful talk - I loved how well organized the presentation was. He based it around an acronym of security considerations - a threat model

Spoof Tamper Repudiation Information Disclosure DOS Elevation of Privileges

Where he explored what each of those meant for Kubernetes and encouraged an exploration of them as a graph.

I kind of want to follow this idea for GDPR.

Great talk.

Tutorial: From Notebook to Kubeflow Pipelines: An End-to-End Data Science Workflow

This was a very slick 90 minute tutorial from Google … the most impressive part for me was how well executed it was, in setting up Kube resources, installing Kubeflow, the data model etc.

It introduced a tool called Kale which attempts to simplify kubeflow pipelines.

It all worked well, I was able to get to the end of the tutorial - deploy a pipeline, work through the sample data … but it isn’t what I really want, which is to understand tensorflow and work from first principles with my own data.

I need to do that work myself. Otherwise - impressive tutorial and technology :)

Fine Grained Mesh Metrics for Better Visibility With Native Performance - Mandar Jog & Kuat Yessenov, Google

This session served to open my eyes to the idea of mesh metrics rather than educate me about a specific implementation - mostly because my brain was getting kind fried this late in the week, and after the kubeflow tutorial earlier in the day.

All in all, I feel inspired to learn more about mesh, build more examples with it, and personally to dig into tensorflow/kubeflow in 2020. I also have a swath of ideas about exploring the economics of cloud usage - have discussed with a colleague to collaborate with in 2020 too.

Adventure Of A Lifetime

It feels doubly lucky for me this month to fulfill a dream in traveling to Colombia, South America, AND sharing a passion project with the developer community there at JSConf Colombia, Medellín. I’m going to present some project work that visualizes the psychology of an artist’s lyrics, while experimenting with an emerging CSS and Javascript technology known as Houdini.

A couple of years ago I saw Coldplay perform live in New York. I was blown away by the concert so much, that I booked to see them again a few weeks later on the same tour. The combination of color, light and sound made the concert very different from any I’d been to before. Each person in the crowd had a gadget called a ‘Xyloband’ on their wrists that lit up and pulsated various colors and patterns in sync with the music. When scaled across 60,000 people in a dark stadium, it amplified emotions and felt intoxicating.

The concert inspired me to doodle Coldplay throughout their career, for each album, and explore their lyrical changes ( like I did with David Bowie’s lyrics a few years ago ).

I find it really interesting relating the psychology of the lyrics to life events of the artist, while also looking for clues as to why the music connects with me so much. It is fascinating to see the influence of life events in Coldplay’s music too.

This time I wanted to stretch myself, and find a fresh angle with my design. Coldplay shower confetti at certain points when they perform. I wondered if I could visualize data with confetti in some way.

This blog post is just a preview of some of the things I’ll be talking about at the JS Conf Colombia. I’ll save the details, and my github repo for attendees, before writing more about it.

My talk is made up of a few parts - lyric mining with Node and Watson, drawing with SVG and layout with Flexbox, exploring and integrating with CSS Houdini.

As a taster, for now I will share my basic confetti Houdini paint repo. I used this repo to figure out how to make a custom confetti css class. During my presentation, I’ll talk through how it works, and show how I use it in the project that I put together.

I look forward to sharing more on this soon :)