The return of the blog
Tue, 5 Dec 2023
Meta content
======================
Hi, it has been a while hasn't it?
For the uninitiated: I used to run a regular html/css blog on my
personal home server. This `server' was in fact just a raspberry
pi stuffed in a shoebox which lived underneath my parents'
ISP-provided modem/router. I always knew that this system would
fail some day, and living more than one thousand kilometers away,
I would have no way of fixing things. This happened way back in
May of 2023. I fixed the issue on a visit this summer, only for
the SD-card to break itself again not a month later. All this
time, I have add the itch to write, but there were two problems:
First, my server was fried, and second, my static site generator
was on that sever.
Yes you heard that right, I was a total [REDACTED] and kept the
only version of my software on a device prone to breaking. I
suffered and I learned. My old static site generator was an
incredibly efficient Makefile, consisting of less than 50 lines
of code. Despite its brevity, it handled: indexing, rss, styling,
special pages, and more. Re-writing this system has been the
main holdup for me in terms of writing.
Now, I hear you thinking: `it was only 50 lines of code, it could
not take more than an afternoon to reproduce'. This is entirely
true, however there is one other little snag: my laptop gave up
life. For the record: I know what is wrong, and I know how to fix
it, but the process is rather too-involved for my current
schedule. As such, I have been reduced to my mobile, and the ipad
I historically only used for annotating pdf's, and dealing with
modern society (in the form of banking apps, identification,
loyalty cards, etcetera).
As you may know: ipads are not exactly great for programming.
Granted, there are some really impressive apps out there such as
ish which emulates a full x86 alpine environment, and a-shell
which aims to bring a unix environment to the ipad (this is
actually what I am writing this post on). These apps are
incredibly impressive, and a-shell is the sole reason I can even
afford to leave my laptop unimpaired for now. Nevertheless, the
programmings-side-of-things is currently too clunky, making a
re-write of the static site generator a much greater effort than
it should be (For context, currently in a-shell, there is no way
to loop over a file in it's port of dash...).
All of this nasty weather comes with a massive silver lining
though: not having access to my usual way of doing things has
exposed me to many new systems which I wish to write about. That,
and the fact that I have about one year worth of posts to catch
up on should make for an interesting year.
With all of that preamble out of the way, let's talk about the
new blogging system. It is INCREDIBLY simple, and very scuffed,
but I must say, I find it has a certain charm to it. Basically, I
have a folder containing plaintext files. The first line is a
title, the second the date, and the third the tags (though not
even this is fixed, these three lines can contain whatever).
These files are pushed to a GitHub repository where a small
script is ran on them. I cannot stress enough: this script is
SMALL. All it does is list the files in reverse chronological
order, and grab the first three lines from each using unix head.
I use some sed magic to get urls working in html, but that is it.
ls -rc posts/* | xargs head -n 3
# the rest is commentary
I was inspired to do things this way largely from my explorations
of gopherspace. For those who do not know: Gopher is a very
simple internet protocol which was around before the modern web.
It essentially serves plaintext files much like this one, and has
some special syntax to deal with links and such things. Gopher
has been largely phased out by the growth of the web, though --
as with all things -- a bunch of stubborn nerds refuse to let it
go. In case it is not clear: those are terms of affects, and I am
such a suborn nerd.
I severely disliked gopher when I first learned of it. I saw no
reason to distribute information over protocols which required
specialized software to retrieve. That is not to say I liked the
modern web mind you, but gopher (and gemmini for that matter)
seemed like they would just require yet another application to
exist on my system. This was especially frustrating because I
read only one gopher phlog at the time, and I considered
switching to a different application just for that a waste of
time. I also hated that there were no RSS feeds (which I still
consider the only saving grace for the modern web), thus
separating these phlogs even more from my other content. (Just
before anyone asks: yes I will eventually get rss working on this
site as well, but writing is currently more important to me than
programming).
All of that is to say: things have changed, and I now spent a
significant portion of my internet time on gopher. I became
inspired by gopher phlogs such as hoi.st by Luxeferre, someone I
know from my days working on KaiOS -- and gopher.black by James
Tomasino, to create a very simple site so I could just focus on
getting some of my writing out there again.
I might mirror or migrate this blog to gopher some time in the
future. It certainly seems a lot better suited for that space,
but for now: I just want to write again, and this is the easiest
way to do that. Thus we have this weird https-plaintext, I assure
you that it is as confusing to me as it is to you.
