I just submitted a PR to the github-pages gem to update to rouge v3.3.0, catching syntax highlighting for it up for nearly 2 years of changes to rouge.
The tests passed, so hopefully it goes through. It was pretty trivial for my part... I just connected the dots from other people’s comments on a PR with failing tests and submitted my own PR based on that suggestion.
How does something so small hold back updates to a project that supports over 100,000 docs pages and blogs?
Did a bit more work on my Terraform Provider for Ansible. Did my best not to bite off more than I can chew feature-wise for this long weekend, leaving it totally functional but not as complete as I wanted to.
Kind of glad I limited the scope of the actual feature to leave part of it for another weekend, because I realized after some release chores to make the feature available publicly that I didn't account for time to spend on documenting this new feature. 🤦♂️
I played around with my wife's Wacom tablet in Adobe Illustrator last night.
I sometimes enjoy doing a little doodling - but I also walk away from the experience with a profound sense that other people's talents often make their jobs look easier than they are. It's good to take a moment to appreciate people for the awesome work they do, that I could not.
Anyone can do almost anything with enough practice and effort. But there's only so much practice and effort that one can fit in a lifetime.
The idea was inspired in part by Greenkeeper, which can be used to keep NPM dependencies up-to-date.
I wonder how they handle this.
Granted, that’s a whole product unto itself that needs to handle more edge cases than my little special-purposed idea. Probably better just to build PR or branch names in a consistent manner so they can be parsed back to check if it has already been handled.
Thinking of doing some slightly more advanced automation than usual. We have 50+ Ansible roles that we maintain at MedStack for internal use. Some of them interact in tests, and have requirements files with release versions in them.
I’m contemplating making a little nightly task to update to the latest tagged release for all the test dependencies, and submit a PR, which would trigger tests. Seemed trivial until I considered... how to keep it from submitting basically the same PR twice?
To sum it up, my epiphany was that my work stressors map almost one-to-one with the forces that move the “systems operating point” around in this model. And that by reacting to the stressors in an emotional way, I just end up moving the operating point around wildly.
By being aware of those stressors and managing them carefully a balance can be achieved.
Last week, I learned something from a recorded conference talk that’s really helped me put some of my work stressors into perspective.
The talk was by Rich Cook in about 2013, on Resilience Engineering, and why things don’t fail as often as you’d expect.
I watched two versions of his talk, and I’m glad I did because the second one is where I had my epiphany. It was all based on Rasmussen’s dynamic safety model.
The experience is a little different. Last time, I was trying hard just to avoid an emotional crash. It was hard even to focus over the stress and anxiety - my brother didn’t agree with my reasons for leaving the farm. It felt a little like I had betrayed him, so I was processing an emotion a lot like guilt.
This time, I feel a lot more balanced from the start, but I’m trying to optimize. These days, the things that cause me anxiety are not personal - and I’m trying not to take them personally.
More than 3 years ago, I picked up mindfulness meditation as a way to cope with the emotional turmoil resulting from departing the family farm to pursue my own independent career.
It was a very emotionally challenging time, and I learned a lot about myself and what I’m capable of on my own.
This week, I’ve picked up the habit again, trying to build positivity into my life. I burned out hard last year, and I think since recovering I’ve become a bit jaded. Taking time to debug myself a little.
I wish I could have consulted for their writers. So many opportunities to explore basic computational concepts, and well within the framework of the story.
I don’t think I’m spoiling much by saying: High School kids are _in cyberspace_.
Ok... so where is the host process for _them_? When they move to any computer, do they execute as a process on that host, or are they interacting with the remote location by Internet Protocol?
I think the story would have been richer with that kind of detail.
I binge-watched “Reboot: The Guardian Code” on Netflix this weekend.
Completely factually incorrect things done for dramatic effect aside, so many missed opportunities! I mean... they never even touch on password hygiene, and it was the perfect platform to get kids started on good infosec habits.
Their only saving grace was playing to the my nostalgia for the original series.
It is proven! I am nbering on Keybase: https://keybase.io/nbering/sigchain#0e9b6b911a32a192d306e57890a78f7f3a662e6acf08eb022aaabb4cb49711d50f
Pretty cool seeing one of my projects featured in someone else's blog post. It's a nice pick-me-up finding this after dealing with someone else trying to take credit for my work earlier this week.
And... done. Shipped updates to support Terraform 0.12-beta2. The change log does not reflect well the work involved in this release.
I also added some simple bash-based test infrastructure for the inventory script to add a bit of a safety net since it’s now supporting a matrix of feature cases and Terraform versions.
I don’t really like using shell scripting for test cases like that, but pulling in a python unit test framework for a couple hundred lines of python that just manipulates JSON structures seems like overkill. So I just capture the output and compare against known good for changes. Good enough for today.