Everybody and their uncle (including me) has written their own static site generator too, I don't understand why this is on HN?

Have to agree - especially as my static site generator is now clearly not going to do SEO, content marketing etc for me and I am heading back off to running a WSGI server

Isn't OP's "static site generator" run on a WSGI server?

Not sure - I meant a dynamic server.

I'd be more interested in this project if it had documentation.

Still I have spent several hours over the past few days reviewing various static site generators to see how well they fair against my own personal requirements:

* https://github.com/skx/static-site-generators

Perhaps unsurprisingly my own tool https://github.com/skx/templer is currently in the lead. The biggest failure I've seen with most of the static site generators I've reviewed has been complete failure to successfully handle symbolic links.

Templer looks like something I could understand - I am not a programmer.

Suppose I had a tree of html snippets built up over time, and I added a new snippet and then issued the templer command. Would the html files of unchanged snippets have been 'touched'? Would 'assets' be re-copied?

I use lftp to selectively upload pages and their associated assets to a remote server.

I'm attempting to find something better than the bash script hacks I currently use...

I'm pleased that it seems understandable.

Templer has two modes of operation:

1. Files are processed/copied from beneath the input tree to the output tree.

2. Files are worked on "in-place".

In the latter case only files that need to be rebuilt are ever touched/modified/updated.

The same applies with the "input/ -> output/" mode of operation, but there is the caveat that I expect people run "make clean", or similar, every now and again. If you didn't, if you just re-ran it every now and again only the things that had changed, or were missing, would be updated.

(Assets are actually copied each time, if you're not running in-place, but we use tar to do it intelligently, so timestamps, etc wouldn't be changed unnecessarily.)

"...if you just re-ran it every now and again only the things that had changed, or were missing, would be updated."

Exactly the behaviour I was hoping for. I've 'cloned' your repository and will proceed to play. Thanks for making this generally available.

As per the other link there are a lot of these systems out there, each with different pros and cons. If you spot bugs or have suggestion please do drop me a mail/raise an issue.

RIP June 5, 2012. No descriptive Readme or whatsoever!?

The docs are bit short for the moment. But simply put, montag is a bit like hobix and relies on the command line and yaml/json files. Docs will be added as the project gains momentum

Speaking for myself, and with my NIH syndrome firmly in place, unless you have documentation you'll find momentum is going to be very slow to build.

I assumed the project was an art joke or something until I discovered some code

No docs aside from some cryptic nonsense in the readme, why is so much stuff like this hitting the frontpage on HN lately?

Skip this and use Pelican - http://docs.getpelican.com/en/3.1.1/ which supports reStructuredText, Markdown, and AsciiDoc.

No doc, no README ? it's maybe too early to promote it.

How does this compare to other static site generators,lets take hyde for example, any particular advantages?

What hilariously cryptic commit messages...

What a short markdown processor!