The musings of an aspiring carver of space

Privacy-friendly community

I do apologise for missing yesterday. It wasn't that I was burning the midnight oil or some such. I just needed an earlier night, and my daughter wasn't happy sleeping in the warm temperatures we're experiencing in the UK right now. (Tonight is going better, fingers crossed.) She did go off eventually, bless her, after using up my blogging window.

This post will be a bit of a smorgasbord. (I won't litter it with horizontal rules, though.) I found the XXIIVV webring a few days ago, and I've been having fun exploring.

Fascinating designs: XXIIVV personal wiki

I've even put in for membership, although it seems like they've got a few pull requests piling up.

Regardless, it led me to some great sites: I found out about Aether through Romain's list of privacy-friendly products. I found a basic org-mode tutorial through Konstantine's link summary for February of this year. I also found a really heartfelt post about grief, which led me to another that I really identified with; I've blogged about it before: losing friends, and reliving the decisions leading up to that.

It's probably because I'm so focussed on it lately, but I've found at least three digital gardens/personal wikis recently; seemingly by chance. Maybe I should be publishing the work I'm doing in Obsidian; it would certainly alleviate the problem of forgetting what I've blogged about versus what I've just noted. I posted a few entries from zettelkasten in Aether yesterday, as part of a comment. Further thought required, I think. If nothing else, I'd be a bit embarrassed by my threadbare process at the moment; really need to go through some of these tutorials, before the inevitable clean-up becomes a task of epic proportions.

End of Day 45

jlj@Fosstodon #100DaysToOffload

I'm writing this as part of the 100 Days To Offload project; join us at: https://100daystooffload.com/

The Salisbury Poisonings, and revision

Note: no spoilers follow.

Lots more revision for my Splunk Enterprise Certified Architect exam today. I postponed it again as well, to a week's time. I'm just not getting through the material quickly enough. The deployment plan, and the process around it, is weighted as 2% of the questions, yet I spent half a day on that easily. When most of the sections are weighted as 5% of the exam, there aren't any corners to cut, as far as I'm concerned; add to that that I'll have to do all this stuff on the job anyway — assuming furlough doesn't end with being laid off, of course — and I'd just be stealing from Peter to pay Paul, as no one says these days.

My revision document is up to 14 pages now. I think it's more legible than my Admin one, which was 23 pages by the end, its skeleton completely collapsed: I could either read it cover-to-cover, or search it; as a reference, it was, and is, useless. Everything's Google at work, but I've been making liberal use of bookmarks and shorthand references to the course docs in my Architect one, all against a straw man of the exam blueprint. It still makes sense, for the moment, although the miscellaneous blob at the end has been growing.

With each day that passes, I grok more of the utility in a zettelkasten, even as the pressure of my deadline increases, precluding trialling it, let alone running it as my main knowledge base.

In a rare burst, we binged The Salisbury Poisonings over two nights. Absolutely fantastic television. Say what you want about the BBC: I'll look to this, and Fleabag, and be very pleased.

Writing like this puts it on my must-watch list. To get the performances we saw — sometimes in such a small amount of screen time — as well is so rare. A gem. Of horrifying, utterly sad material. The directing was inspired at times too: particularly some of the point of view shots, I thought.

Everything else will suffer for having followed this, that's for sure.

End of Day 44

jlj@Fosstodon #100DaysToOffload

I'm writing this as part of the 100 Days To Offload project; join us at: https://100daystooffload.com/

Killswitch (2014): Snowden != Swartz

I finished the subject film last night. (It's neat that the Internet Archive hosts it and The Internet's Own Boy, btw. The nostalgic post-feature walk down my 90s presence in the Wayback Machine was very enjoyable.) I liked it. I thought it was well-cut and compelling. I know I'm too old for their target audience, but I was put off by the conflation of Swartz and Snowden's causes; I think it's a bit of a stretch, and, more importantly, threatens to blur the clear lines between the two associated acts.

We'll never know why Swartz pillaged JSTOR — I assume, unless information has come to light in the last six or seven years; as I said, my mental health isn't up to digging too deeply on this sad tale — but TIOB speculates convincingly on the point, and regardless, no one was actually harmed (to the best of my knowledge). I think that that's a much harder argument to make in the case of Snowden: the sheer volume of data he provided to journalists — and, therefore, to many actors, most of whom don't wish his country well — defies human consumption (let alone screening for harm) in any reasonable time frame.

Still, as implied, I'd imagine that that editorial decision went down well with the target audience. And I very much consider Killswitch to be time well spent.

One final note: they included more footage from one of the Lessig interviews used in TIOB, and I couldn't help but think that, as a whole, that material painted the man in a slightly different light: he looked broken. Completely broken. After TIOB, I felt he'd been irrevocably changed, but was still intact. I did not have that impression as Killswitch ended.

End of Day 43

jlj@Fosstodon #100DaysToOffload

I'm writing this as part of the 100 Days To Offload project; join us at: https://100daystooffload.com/

Deep Work

I'm just starting the subject book, by Cal Newport. I've been thinking about reading Newport for quite some time: CGP Grey recommended So Good They Can't Ignore You on a very early episode of HI, as I recall. Rob Muhlestein (aka rwxrob) recommending Deep Work as part of his RWX.GG knowledge project was the nudge I needed.

I begin this endeavour with more than a bit of trepidation.

Am I cut out for deep work? I honestly don't know. Back in the days of pencil and paper, I remember accomplishing most things through a facility for memorisation, and the adrenaline surge of a time-crunch. (I was an astonishing procrastinator.) I suppose, thinking about it, I did produce some Dungeons & Dragons campaigns of staggering detail, requiring many hours of concentrated work. But that isn't the same thing as pushing my brain with a heavy cognitive load for an extended period of time.

But I must try. As I've said before, my training programme is not letting up, and my current, chaotic approach is reaching the end of its utility.

I finished Notes On A Nervous Planet by Matt Haig last night. I think I mentioned that a friend had recommended it, knowing about my problems with anxiety. While I think there's a high likelihood — on any particular random opening of the book — of your eyes falling on something rather facile, I would be lying if I said that it hadn't helped. My zettelkasten, only started in earnest when I was already two-thirds of the way through the book, still has notes on half a dozen pages of Haig's thoughts.

I'm still anxious, though. The degree varies from day to day, spiking when I think about leaving the house (which I rarely do). Another source is looking at what's required of me professionally over the coming months. I'm hoping Newport will help me get a handle on that.

End of Day 42

jlj@Fosstodon #100DaysToOffload

I'm writing this as part of the 100 Days To Offload project; join us at: https://100daystooffload.com/

Happy Father's Day!

This one will be brief.

I'm a lucky man, and I've spent the day feeling very grateful for that, surrounded by my loved ones.

I've had a complicated relationship with my own father, but even in this, I'm luckier than many: he's survived to my middle age, where we've both come to realise that we have pretty much diametrically opposite views of our shared past, and that there's little to be gained by dwelling on that now.

This, in turn, has galvanised my commitment to my own children. I will make mistakes; likely many of them. But my children will know that I regret each and every one of them, and still love them with all my heart. They will know this because I will tell them. Often. Probably too often.

They will know how special they are to me. And that however badly this world and the people in it hurt them, betray them, disappoint them, I will be there to listen. And to share in what I hope will be their much-more-prevalent joys.

End of Day 41

jlj@Fosstodon #100DaysToOffload

I'm writing this as part of the 100 Days To Offload project; join us at: https://100daystooffload.com/

Derek Sivers is nice

As usual, the weekend (so far) has been filled with many lovely things that I won't be blogging about. And then we decided to finish Picard tonight as well. (I do have some thoughts on it, of course, but none worthy of the spoiler warning that would have to go with them. Drop me a line if you're interested in this diehard TNGer's take.)

I ran out of space with my rclone project: it's a hard 15GB limit that doesn't seem to have touched the estimates on my actual Google account. I can't imagine a service account I create gets its own Google Drive allocation, but I don't know how else to explain what is going on. Suffice to say that it's been fun, but I'll have to think about what I might want to use that space for at another time. For now, I've picked up a 128GB SanDisk thumb drive for £15 and copied the Timeshift snapshot to it. (It's extraordinary, to me, that it takes a third of the physical space that the 4GB thumb drive that I still use does.)

I now feel like I'm on fairly firm ground with my backups. I'll still need to test them, down the road, if I can be bothered. To be honest, as long as I have a copy of my photos, videos and documents, rebuilding just seems like a chance to practice system administration. (Some would argue that creating and restoring functioning backups is a very important skill in that suite, of course. One that requires some effort to practice.)

I started watching Killswitch (2014) last night, curious to see more of Aaron Swartz. I'll write about that once I've finished it. I will say that I've already ordered a copy of Tim Wu's The Attention Merchants off the back of what I've seen, though.

I'll close with a nod to my Now page and Derek Sivers. You may have noticed that the former is linked at the top of my site; it's a .plan file for the 21st century. And I mean that in the best possible way. I'm looking forward to exploring Sivers' body of work... (OK, I've just googled him now, and, my first thought is that I can't believe how accessible and friendly he is over email. Wow. OK, I'll probably have more to say on that in a future post; I certainly have a title for this one now.)

End of Day 40

jlj@Fosstodon #100DaysToOffload

I'm writing this as part of the 100 Days To Offload project; join us at: https://100daystooffload.com/

The Internet's Own Boy: The Story of Aaron Swartz

I watched this for the first time last night.

Somehow, I hadn't even heard of Aaron Swartz.

And, while I let that sink in, it's worse than being a fan of networked technology and not knowing: I was active in the digital rights scene back in Ottawa at the time too. I wasn't just reading Geist: we were meeting, as part of the group trying to get Bill C-61 amended. And we were all reading Lessig, Doctorow, and were intensely interested in Eldred v. Ashcroft.

I loved Creative Commons. It was all over my creative writing site at the time. I had no idea Swartz wrote that code.

I've tried to let myself off the hook a bit: maybe all these people he worked with — on RSS as well, and even Markdown, apparently, although that's from Wikipedia — were trying to keep him from the limelight because he was still a minor. I don't know.

It's a bit easier to understand how I missed his death: my personal life was a shambles in late 2012/early 2013. I was barely keeping it together, even with my head down, just trying to get by.

The film is as interesting for who it wasn't able to get, I think. I found the whole topic to be pretty upsetting, so I won't be doing any research on this or other points. For me, that they got his parents talking was enough: and that made it all the more heartbreaking, of course.

He was so eloquent! Even on live television. That, to me, speaks of his brilliance more than anything else: that he was able to answer questions so well, on the spot. Likely what I consider a home-run answer would've had him shaking his head afterwards, wondering why he hadn't said this or that. I bet his shower thoughts were world-altering.

And that's the saddest part of all: Clay Shirky referenced it in one of the few articles I read about Swartz's death: everyone will miss out, is currently missing out, on all the other things he would've done. He was just getting started. It's awful; just awful.

I'll close with Sir Tim Berners-Lee's beautiful epitaph:

Aaron dead. World wanderers, we have lost a wise elder. Hackers for right, we are one down. Parents all, we have lost a child. Let us weep.

RIP Aaron Swartz 1986 – 2013

End of Day 39

jlj@Fosstodon #100DaysToOffload

I'm writing this as part of the 100 Days To Offload project; join us at: https://100daystooffload.com/

Is Python as esoteric as it seems?

First, the new colours: Awful? Nice? Strange? If I'm risking the little traffic I'm getting now with this, I'd appreciate a quick heads-up/@mention. ;–) I was playing around with Khroma late last night, and I'm still sleepy now, so I'm willing to accept that my judgment may be flawed. [Edit: I'm going back to the drawing board.]

My zettelkasten project is off the ground: I've been using Obsidian for a few days now. Just for note-taking — modelled roughly on the workflow outlined in Effective Remote Work's video on the topic — because working with new software while also learning the nuances of a more involved workflow seemed audacious at this stage. If there's one thing I've learned from blogging over the years, it's just bash it out: get something down, and keep that ball rolling; you can edit — heavily, if necessary — later.

One stumbling block to that process has already surfaced: I thought I'd blogged about all this, because the interface and thought process I use for both are, well, identical. But, no, a few keyword searches later revealed nothing on the blog; this is another process I'll need to refine as my slip-box grows.

That rclone process I talked about yesterday is still running, by the way; some 32 hours later. This is both surprising and fine: I'm really hoping that most recovery scenarios will involve using local snapshots; this is yet one more nightmarish aspect of the nightmare scenario of a house-fire or something, I guess.

I've talked about how I'm thinking about learning Python. I'm questioning the wisdom of that lately: every time I try to play with some code, or even get someone else's up and running, I end up on Stack Overflow nine or ten times an hour. My thought process, in no particular order, goes:

  • Should I set up a virtual environment?
    • I'll run pip when I want pip3 then; better do it.
  • Have I activated it yet?
  • How should I install this? Easy install?install setup.py .? With a whl file instead?
    • Is using pip ... setuptools --upgrade associated with the former? Is upgrading wheel all I need to do to get the latter working?
  • Is this pyyaml error related to the pip install, or to a missing dependency in my wider Linux installation?
  • What's Anaconda?
  • Do I still need to worry about generating a requirements file?

On and on. Worse still, nothing sticks because I'm not working with this stuff enough. And, yes, I'm exaggerating a bit to make a point. Maybe it comes down to what you learned on. When I'm compiling C code, I'm looking to the developer to point out dependencies; I'm happy pulling down libraries based on their documentation, or on subsequent build errors, even linking libraries symbolically as required. That may make as little sense to someone else as the various building blocks of any Python environment do to me, I suppose.

End of Day 38

jlj@Fosstodon #100DaysToOffload

I'm writing this as part of the 100 Days To Offload project; join us at: https://100daystooffload.com/

Remembrance, versus reverence and normalisation

Following on from the news about terminology changes in the codebase of both GitHub and OpenZFS, I joked in our company's chatroom about how our new hires like me might be the last contingent to test on Splunk's current cluster terminology. To my surprise, a colleague quickly replied, saying that such changes are indeed afoot.

I have to admit, the urge to roll my eyes was strong. In fact, my response was, “Satire is hard these days.” But, on second thought, there is something to be said for these changes, if less about the motives, so publicly expressed, behind them. A few years ago, I wrote about the Confederacy statues being demolished in the American south. To paraphrase, there is a very important difference between remembrance and reverence. We are not cleansing history with these acts, as some have argued, but failing to revere them any longer.

Now, few would call referencing the terms of slavery — in software and code, devoid of any other context — reverence. I could see an argument being made for calling it normalising, though. And, to go further, that the outcomes from both reverence and normalisation are exclusively somewhat to extremely negative.

As I get older, I realise how difficult it is to objectively assess things that have been a part of one's experiences and observations for most of one's life. Whether something came to be five years before your birth, or 50 years before, will be indistinguishable to you. (Well, let me quickly clarify that that is the perspective of a cis white male, very obviously, and, far from coincidentally, likely also the perspective of the majority of eye-rollers right now.) But that fact alone is poor rationale for it to outlive you as well.

I did have some success with the home set-up today too: rclone is currently backing up a Timeshift snapshot to Google Drive using a service account I set up for the purpose. I wonder how long the term clone will be in common use.

End of Day 37

jlj@Fosstodon #100DaysToOffload

I'm writing this as part of the 100 Days To Offload project; join us at: https://100daystooffload.com/

Timmy's ain't been quaint for a long ole time

My son decided that 5am was go-time this morning. My Splunk revision has suffered as result, I think it's fair to say. Still, I feel more comfortable with search head clustering now, so it wasn't a total loss. For completeness, in a disaster recovery scenario, whether the remaining search heads are an odd or even number isn't relevant (as far as I've been able to discern): it's whether 51% of the original membership remains. If so, the election of a new captain will proceed without issue. If not, manual intervention is required; likely in the form of designating a static captain temporarily.

Fascinating, I know.

Don't even get me started on the criteria under which a master node will rebalance the indexer cluster, or what triggers a change in its generation (ID).

But you didn't come here to revise Splunk. (I hope — may God have mercy on your soul otherwise.)

I'm Canadian. I think that's come up before, but I wanted to put it out there again, as Tim Horton's is a national institution to our older generations (of which I am most certainly a member). I mean, it's been an American company for, oh, twenty years now or something, and stopped selling doughnuts with coconut shavings thirty years ago, so, really, why would anyone care? ;–) The fact is, “double double” is (still?) part of our lexicon (meaning coffee with two portions of cream and two portions of sugar, I know you were dying to know).

And Tim Horton's is in the news. Well, in the Financial Post. For avidly tracking the users of its app. I was shocked by the bit on server-side location processing, or the tools at the command of third-parties in the event of you preventing an app from accessing your GPS data. The type and number of events generated around user movements — including confidence factors around why the user might be where they are — is chilling. I mean, this makes loyalty cards mickey mouse; an increase in customer data of, what, tenfold? More?

And so, folks won't actually be rolling up any cup rims in the future, it seems. (Plenty of Canadians' favourite new car dream.)

End of Day 36

jlj@Fosstodon #100DaysToOffload

I'm writing this as part of the 100 Days To Offload project; join us at: https://100daystooffload.com/

Enter your email to subscribe to updates.