One thing I noticed is that pandocs output isn’t stable, so different versions of it may produce slightly different output, which makes tracking in git annoying.
That’s how https://www.oilshell.org/ started out, with a shell script and the original markdown.pl. Then I added a Python script that wraps CommonMark, but it’s all still controlled by shell (and Make, which I want to get rid of).
lol, your site looks awesome. In the past, I used hexo to build my blog, but hexo is too heavy, so I write inkval script to help mine and other people just like small pages.
Thanks, I also like yours! I can totally see the “heavy” argument - I rarely touch my site, so I’m okay that build times are actually not the fastest, but it’s very important for me that if I end up doing it, everything works and bash kind of delivers that (except on Windows, cough).
I had to laugh, as this is how my site works (kind of): https://github.com/skade/yakshav.es/blob/gh-pages/build.sh It’s a little shorter, but I enjoy not having a huge thing transforming my pages.
One thing I noticed is that pandocs output isn’t stable, so different versions of it may produce slightly different output, which makes tracking in git annoying.
That’s how https://www.oilshell.org/ started out, with a shell script and the original
markdown.pl. Then I added a Python script that wraps CommonMark, but it’s all still controlled by shell (and Make, which I want to get rid of).lol, your site looks awesome. In the past, I used hexo to build my blog, but hexo is too heavy, so I write inkval script to help mine and other people just like small pages.
Thanks, I also like yours! I can totally see the “heavy” argument - I rarely touch my site, so I’m okay that build times are actually not the fastest, but it’s very important for me that if I end up doing it, everything works and bash kind of delivers that (except on Windows, cough).