Bloguidien #4 - My writing setup?

I’ve talked about the bloguidien challenge in a previous post about the first prompt about my favorite blog post(s). Since day one I started doing things “wrong” (= my way) by a/ answering it in English and not French and b/ not really answering the question as simple as it was asked…

And I’ll thiskeep doing it my way by not answering all posts (which to be fair was clear for me since day 1) but also not answering in the right order :D. I’m planning to answer to the second prompt about travel, but tonight I wanted to write quickly about the fourth prompt:

What is your writing setup? software / specific hardware / location / music / atmosphere / isolation / interspersed with breaks

My setup is “simple”, I use emacs to write my posts in orgmode. Even though I have been a vim addict for too many years to count, orgmode made me switch to emacs to write blog posts and manage my work and personal notes and todo lists. I didn’t became the biggest emacs’ fan and I still don’t understand how people like (e)lisp. I don’t judge them, I just don’t get it 🤷. I still code in vim, at least mostly as sometime I do use emacs. I use emacs’ evil mode to use all vim shortcuts and avoid as emacs default keybind as possible for my wrists and fingers’ sake.

I write on my laptop, most of the time at my desk, meaning plugged to external screens and using a mechanical keyboard (which I enjoy so much more than any other keyboard :)). Sometimes I go to the leaving room and write there when my partner is also there doing or watching something I don’t care about. In those rare cases, I sit at the diner table, no external screen or keyboard used. I don’t like writing with my laptop on my lap though (ironic?).

I recently bought a desktop to move away from using a laptop 18h a day a killing it in a 3 or 4 of years as I’ve done so often in the past. It isn’t fully setup yet as it was setup only for my factorio gaming week and used by my brother. But this week I’ll finish its setup by removing the Fedora I installed for simplicity for my brother and put back an Archlinux and all my beloved tools. So soon I expect to write a lot on it too. Files will be shared with syncthing so I don’t care too much about which machine I use.

I wish I could but I just can not type lengthy text on my phone… Typing texts, signal messages or short emails are already annoying enough on a touch screen, but a blog posts or any long text is just a no go for me.

I don’t care too much about background noises, I often write in a silent environment but I’m fine with some background music or if my partner is watching something. I do feel I’m faster writing alone in silent but I’m not looking for efficiency anyway when blogging.

Until recently, I wrote 99% of my posts at night, after my partner went to sleep. This is the case for this post too. But since the arrival of a new puppy at home and being awake way earlier, I wrote a bit in the morning since then too. And I’m going to try doing it more often. I guess we’ll see soon enough if that happens!

Non technical blog posts are usually written in one go, typing stuff away until it is done. Technical posts are written in multiple times. Usually it starts as note once I’m doing or learning something. Lots of bullet points and copy / paste of commands / code I have did. Then I transform them into a post. Lengthy tech posts (like my deployment workflow or laptop setup can be written in small pieces over multiple days or weeks.

Once I finished writing a post, I activate the spell and grammar checks within emacs. They are disabled by default because my org files are usually pretty large as I have 1 org file per year for both blogging and gemlogging and they slow down emacs too much to my taste. The current file I’m typing this blog in has 6.3k lines. So for sake of speed, I only activate those once the post is done for review and fix found errors and disabled them right after.

Next step is to export the article as a markdown file that will be used by Hugo, the static site generator I’m using. This is done with keyboard shortcuts so it takes a couple of seconds: space → m → e to tell emacs to open the export screen, ctrl + s to indicate to export only the current subtree (and not the other subtrees corresponding to other posts) and then H → H to export to markdown using ox-hugo. It will read the properties in my org file and at the start of the article to export it in the right format, folder and with the correct frontmatter headers. See the screenshot bellow (font size has been reduced to show more things, it is usually a lot bigger).

Next step is to run Hugo locally (using hugo serve) and look at the post being render within my blog to verify there is no error in formatting.

Last step is committing the new file and pushing it (to both sourcehut and a locally hosted git server). Sourcehut pipeline will then run and create the artifact which my web server will retrieve and deploy. To learn more about that process, I’ve written at length about it already :).

As usual, a lenghty post even though I thought this one would be quick, but any reader here already know that short posts are very rare here :D.