Breakpoint: path to milestone zero

Breakpoint: path to milestone zero

One of my favourite things about engineering and in particular, about programming, is very well summarised by this gif. I have it saved in my computer, that’s how much I like it

I also vaguely remember an XKCD comic about the same topic. Programming is, effectively, about side quests. I can’t find the comment, please shoot me an email if you know what I’m talking about.

Side quest 1: The blog setup

I’m trying to do a blog. I have a concept, I have a project that I want to write about. Great. First side quest: how do the kids blog these days? Huh. Or at least, how do software engineers do? Out of sheer accident, I found out that Git Hub has this nice static page hosting–oh, and it supports custom domains, will you look at that!

So side quest number 1.1 [1] is surprisingly short due to my lack of investment into finding a correct blog platform. Praying that won’t bite me too much, or at least too soon. [2]

On to side questi 1.2: how do I use this thing? Luckily for me, Git Hub put together a Valve-level tutorial, which was very clear, so here’s me using that: https://github.com/royalsflush/skills-github-pages.

Great, so now I know how to create the first post. I’m still aware that my resulting site looks hideously confusing (still in https://royalsflush.github.io/skills-github-pages), but I decide to let that slide because I’m sure at that point it’s not gonna be a problem. ( Foreshadowing…)

So I go on my merry way, write the README trying to think about the conceptual approach of it all. I’m pretty excited about this. I think the idea is not a bad one, and I’m definitely keen to playing with some fancy 3D visualisation tool in React to plot spinning graphs of my blog posts. That sounds pretty fun. I also dig the idea of giving a blog reader a structure to explore a story without tangling a bunch of different subjects. It feels like a cool way of consuming content that is not the complete randomness of a blog (so it also has some linear discovery), but it’s not a book. Anyway, I’m excited!

So I turn on my Git Hub pages, and it shows me the README in all its published glory. I also have a domain (breakpoint.royalsflush.com) that I want it primarily. Time for some domain wrangling.

Side quest 2: domain wrangling

I go to my domain provider, configure a CNAME record–this one I have done plenty of times. And now, I also own royalsflush.blog. Hm. That should redirect to breakpoint.royalsflush.com. And now I’m thinking: how can I do this? Normally CNAME only redirects to a specific domain, can I do a wildcard redirect? Short answer: no. [3] Slightly longer answer: CNAME only works for specific subdomain, and if you want a full on redirect, you want an HTTP redirect instead (https://en.wikipedia.org/wiki/URL_redirection).

Now I’m thinking I’m gonna have to configure an Apache server to do an HTTP redirect, and I’m sighing… until I realise my domain provider offers forwarding as a service. Side quest done!

Ha, of course not. I try to access royalsflush.blog to see my post, Chrome gives me a security warning. Me, in my diligence, put a redirect with SSL, and I quickly learned that SSL certificates are not for the entire domain (superbly explained in https://serverfault.com/questions/566426/does-each-subdomain-need-its-own-ssl-certificate). I have decided that this is not my biggest issue right now, because I try with breakpoint.royalsflush.com (which works), and nothing is showing correctly. Issue count += 2 [4].

Side quest 3: debugging Jekyll locally

Okay, so Git Hub recommends me to install Jekyll locally and try things out, which, I dig. I see I have a version of gem installed, let’s do it (I wasn’t expecting that!). Then, of course, my version is super old because it has been shipped with the dev tools of my MacOS. Okay, so gem update --system. Now it complains my ruby version is old. Side quest 3.2 it is.

https://www.ruby-lang.org/en/documentation/installation has some pretty good detailed info, but it doesn’t tell me a crucial piece of information: do I need to support multiple versions? Or can I just go the Homebrew [5] route and call it a day? I’m wary that I’m gonna some obscure thing that uses the system version. https://mac.install.guide/faq/change-ruby-version confirms my suspicion. Let’s pick up RVM and do a combo installation [6]. Which in turn requires gpg2 for keys, so I ended up using homebrew anyway.

And now, to cap this post about side quests:

gpg --keyserver keyserver.ubuntu.com --recv-keys 409B6B1796C275462A1703113804BB82D39DC0E3 7D2BAF1CF37B13E2069D6956105BD0E739499BDB

gpg: keyserver receive failed: No route to host

Filed https://github.com/royalsflush/breakpoint/issues/4. To be continued…

---

[1] Because remember, the blog itself is not the main project, Minesweeper is. But I want to write about Minesweeper, and now I want a blog to do it so, here we are. [2] It’s my current belief that no design decision survives the stress test of platform evolution, be it scale, usability, new features, etc etc. [3] I also get side tracked by the fact my domains don’t have adequate protections, so I get on to fixing that as well. [4] https://github.com/royalsflush/breakpoint/issues/1 and https://github.com/royalsflush/breakpoint/issues/2 [5] I decided to install Homebrew anyway, which I used to have and I don’t in this machine. [6] Here if you need: http://rvm.io/