Oscar: thank-you! Gocommerce is exactly the sort of thing I was hoping to find. Those Netlify devs are doing some amazing work. I hear you loud and clear about self-hosting. I used Woocommerce on a small eCommerce store many years ago and thought it was great at wrangling the monster that is Wordpress into a slightly different beast. I had not realised that WC could be used programmatically. Although it's not quite right for this project, I'm keen to experiment with that option in the future. Thanks again for your ideas.

Hi @Edouard, I don't really have a set amount of time for daily learning; sometimes it will be hours on a course and other times it will be just 15 minutes at the start of the day. About a year ago I realised that the process is what is important, progress follows along if you get the process cracked. I've often felt that my skills were weak, and being honest I don't think that ever goes away. Someone pointed out to me that you NEED to be intimidated by the road ahead, feeling overwhelmed is part of the learning process. I visualise learning as a dense fog that covers a city view, and as I understand the concept better, the fog lifts. I also realised that I tend to worry about how complicated a topic is before I understand the boundaries. As a result of this, I now try to define the limits of a new topic before I start trying to understand it. Building stuff is the second best way to learn (teaching is the best apparently). For that reason I think side projects are really important, and getting starting is the key. You might have to force yourself to do it but you'll be glad you did!

One man comes to mind, I'm intimidated even typing his name: Fabrice Bellard. His Wikipedia page An article about him His personal website He built his own 4G basestation which ran on a PC. He created ffmpeg. He wrote a linux emulator in Javascript. He created his own image format ( BPG ). These are the tip of the iceberg. How one man is capable of achieving so much is quite astounding. Here's a quote from the article above: "If there’s a secret to this superhero-level productivity, it appears to have less to do with comic-book mutation and radioactivity, and far more with discipline, confidence, rigor, and many years of practice."

Hi Edouard, this is a good question. I remember hearing some advice that stuck with me: a developer's skills are not improved simply through the work done on the job. The proven way to improve and expand is through deliberate practice. It sounds as though you have already found that this is easier said than done! I have a formula that works for me and it comprises of four main tactics. I schedule in time for practice and learning sessions just as I would my normal work tasks. I deviate from this schedule as little as possible. For daily practice I like sites such as codewars. I take longer courses with deadlines and assignments to keep me on track. For example, I've taken Andrew Ng's Machine Learning class which had weekly programming assignments alongside the lectures. I set distance goals for learning broad new skills and track the progress. E.g.: I want to understand an implementation of Deep Learning by the end of January. I will periodically review my skills and try to objectively decide where I have weaknesses or gaps in my knowledge. Building in periodic revision is extremely useful, particularly for skills/topics that you don't use in your daily work. I've taken more than 30 courses this year and some of them have partially evaporated because I haven't taken the time to refresh my memory of the concepts. Finally, I've found that more than many other disciplines, we require an attitude of "always learning". The moment you decide that you know "enough", or "all of it" is the point you consign yourself to mediocrity. The speed of innovation, the rate of change, the pervasive nature of software systems and the sharing of knowledge; these all contribute to a profession that will never be boring. I hope this helps, good luck!

Previously I had been using Bitbucket because of free private repos and I preferred the UI. Now I've moved over to Gitlab because feature-wise of they offer lots for free: private repos, CI runners and hosted pages. Their platform is open source and the management seem to be very transparent. Now the not-so-good about Gitlab: regular deploys that seem to coincide with me trying to leave work and go home, meaning I can't push my code for the next 30 mins. The flip side of this is at least they are continually improving their solution. I read a post by the CEO on hacker news being quite open about the fact that they were moving toward a liquidity event, which may impact Gitlab's future direction. As protection towards the community being neglected by new interests, they have "hard coded" a set of promises.

Thanks for your response. I've read some of the fabric docs and seen a few examples but was deterred due to the lack of python3 support. The simplicity it offers over tools such as Ansible were attractive, however. Based on your answer I'll reconsider its usefulness!

The videos from Djangocon are online! Here's the service worker presentation I mentioned.

There was a fantastic talk by Adrian Holovaty at Djangocon Europe about his use of service workers for his music tuition platform, Soundslice . It was the last talk of the conference but made me sit up and say "Wow!". The videos from Djangocon will be available online soon apparently, but in the meantime here's one of his tweets with some useful links .