Managing This Blog — 4

I am pleased and content with the way I now manage this blog. My simple shell scripts are working well. The automation is great.

I write. I “schedule” the post by moving the file to a Pending directory. A cron job converts the text file to full HTML and updates affected indexes. The new files are synced to my local server and pushed to the online host server. The original text file is moved to a Published directory.

Basically all I do now is write and save files in the correct directory.

Focusing on content rather than user interfaces is wonderful. No bane of the web. No online editor. No Google fonts or APIs or tracking. No geek static generators.

No esoteric markup language. Just a text file using only nominal HTML tags. The scripts do the remainder of the converting.

I anticipated going into the project I would be tweaking the scripts for several weeks. There are too many unforeseen corner cases. I have been tweaking the scripts as expected. Nothing traumatic.

Originally when I designed the scripts I presumed pushing at most only one post per day. That presumption works fine for my habits, but I wanted the option to post more than one article without manual intervention.

Similar to the one-post-per-day limit, I wanted a way to test multiple scheduled posts so I could review them all as full HTML on my local server.

I added this preview ability. This new feature helped me tweak the code snippet presentation in my post. This preview ability does not touch the master web site files but instead mirrors the files in a test directory. I added a new bookmark in my web browser to view the test version of the site.

I tend to tinker with my articles a little throughout each day. I created an hourly cron job to keep my “testing” version of the site updated.

A simple proofreading technique is to view text in a different format. Moving from raw text to full HTML helps me spot content errors and glitches.

This preview ability is a welcomed feature. Previewing allows me to review each post to ensure I correctly inserted categories and tags as well as validate the desired scheduled post date.

I would like to add a site search tool. Without using a third-party tool or the bane of the web.

The script code is clean but not super efficient. There is some duplication that I could move into functions. I could add more comments, if for no other reason but to remind myself why I did certain things. These are rainy day projects. Do not fix something that is not broken. I am sure more corner case HTML rendering issues will surface. For now I am pleased and content.

At the moment the script is written for me and my use case. While I try to create generic and robust scripts, there are presumptions in the way I designed the script. Thus, probably not a good candidate for other people.

Posted: October 07, 2016 Category: Usability Tagged: General, Migrate

Next: Spoofing a VirtualBox Guest BIOS

Previous: Learning Something New About The at Daemon