Converting postcards from the bleeding edge

My old blog I started back in 2001 (I couldnt get permission to publish then though, it wasn´t until that job was nearly at end end when I said screw it and started venting publicly again).

The code predates titles, tags, and other standard blog features.

It predates git, and for that matter markdown. The iphone was just a pipe dream, and it wasn´t all that obvious in 2002, that linux was going to be a success in any embedded system at all.

In doing this conversion I have saved what is left of my public canon from the dustbin of history. It was a pita - at one level Ive spent 4 years getting this converted, at another, I was merely waiting for the technology to catch up so I could spend a bit of vacation time getting it to go again.

Bits of the conversion were really tedious.

The jeckyll importer ate tags in general.

Pandoc ate cite tags. It also completely ate anything inside a ul block. It did weird things to li tags - Whole sections of text just vanished. Worse, it treated

as nothing at all, killing all my paragraph markers.

All this I fixed with some simple sed scripts, but they took a while to find.

took .04 seconds to convert.

I clearly remember it being 940+ posts, but I only have 823 files to edit. Where did the other 100 go?

didn´t grab the date

Taht vs Tqht

title: ! ‘My most interesting failures, #3: VOIP’

hugo server | wc -l 119

sed -i ’s/ ! / /’ *.md

got rid of 115 errors. Sigh but nothing rendered.

sed -i ‘/layout:/d’ *.md

strip out the date from the title

get monthly summaries

Paragraph breaks

Fixing the helicopter

rss/the-edge.xml

2003-06-22-on-writing-i-had-no-ideas-as-to-what.md:title:

February 17, 2015
296 words