I discovered that I can’t get the hugo_jupyter module to install on my laptop for some unspecified reason due to a collision with the existing software. This is why I don’t typically upgrade but rather do a clean install.

Most things work, but between the system, homebrew, and conda, my python3 setup is not working properly. I guess a backup to bootable image and a fresh install will solve most of the issues, but in the interim, I need to get the generator fully functional.

Why did I wait so long to do this?

The initial testing cutover went well, I have to do some work to re-organize the photos, but it seems that hugo will work out for the site. I need to do some serious reading, as it is a lot different, but the theming system seems to be easier to use and the short codes look like they will be quite useful.

It seems that the previous fix was not sufficient and after a lot of attempts, I gave up on it. The issues I was having were not too easy to solve, and after hitting many dead ends and not wanting to convert things again, I decided to let it sit for a while. Apparently I get rat-holed on problems occasionally and when I know that not that many people use the site, the priority level is not all that high to fix things.

Here we are, the first quarter of 2017 has flown by and not a new post to be seen. That is actually my fault for being defeatist. The pelican generator was upgraded to v3.7 from 3.6 and some portions of my theme started blowing chunks. As a result, it was rather difficult to publish anything without it breaking and I got a little stubborn and refused to back out the upgrade.

Back when I decided to convert over to Let’s Encrypt, I was having a few difficulties with the web server and redirects. After futzing with it for a while, I decided to convert over to NGINX, as I really didn’t need all of the features of Apache any more.

This worked out pretty well, until this week when I decided to have a look at my mailman interface from the web. Nothing. Well, not quite nothing…