Intro
I started the static site checker a while ago. It’s a utility to validate static web sites, such as this one. I started it because I found most of the existing HTML checkers of the time were ridiculously simple and under powered: they’d not spot essential things like broken links, they’d not check attribute values.
I don’t know if things have improved since then, but I don’t really care either.
I attempted to push SSC to the OpenBSD packages, but the one person who looked at it disliked it because it correctly pointed out that unclosed elements were inefficient, as is stated in the specification that defines HTML, the Living Standard. I guess wanting to build a website with correct HTML is a minority sport. It’s ironic, because, as SSC pointed out, there were then a good number of errors in the backwoods of the OpenBSD website.
I’ve not left SSC alone; I keep improving it! It’s my old man’s trainset.
Recent new features
I recent added some features to the static site checker:
- Signing: output can be signed, so changes to the output after SSC has been run can be detected; this was at the request of a client.
- Censorship: when processing server side includes to create a much faster website to publish, specific data can be sought and suppressed. For example, I don’t want my DoB to be accidentally included on arts & ego, so I’ve used this new feature to replace any occurrences of it with something suitably daft. Having said that, the feature needs refining.
- MathML Core, from the recommended snapshot published by W3 in June.
- WebVTT: verify VTT files, and verify that AUDIO and VIDEO elements refer to them.
I’m currently reviewing the Living Standard to ensure that I’ve not missed anything too glaring, and to extend SSC’s analysis of potential booboos (Later: see SSC 0.2.19).