Marcus Rohrmoser

A pro for xml: it can be rendered to html with JS disabled. A sidenote concerning yaml: json is actually a subset of yaml, so every yaml parser must parse...

I'm curious, too (concerning https://github.com/mro/ShaarliOS). What are the primary usecases you think of? Daily? The list? Search & tags? Having all posts buffered for offline use?

the mere caching is simple – just cache the (logged in) [Atom dump](http://shaarlidemo.tuxfamily.org/Shaarli/?do=atom&nb=all). But key is what UI/UX to provide.

any news here? It looks so promising but doesn't run nicely. Such a pity. Depends on the sh version: ok: GNU bash, Version 5.1.4(1)-release (x86_64-pc-linux-gnu) fails: GNU bash, version 3.2.57(1)-release...

unasked advice from outsider: Personally, I would use yaml for the raw project data plus a single file, ruby 1.8.7-2.2+, zero dependency, 20 line script to merge it into Readme.md....

@gRegorLove thx. That's a sad surprise. Is the indieweb really microformats only?

@Zegnat > Looks like you have found yourself a nice little edge-case here! lol, indeed, but isn't the @indieweb an edge-case by it's very nature? > a problem with @mro’s...

P.S.: If my (devuan/jessie) lighttpd (1.4.45) would let me without hassle, I would return [HTTP 406](https://www.w3.org/Protocols/rfc2616/rfc2616-sec10.html#sec10.4.7) instead of the current $ curl --location --silent --header 'Accept-Encoding: identity' https://l.mro.name/ | file...

maybe related to https://github.com/indieweb/indiewebify-me/issues/78

> perhaps we just specify "If you're using HTTP-based transport, then the value MUST be via HTTPS if publicly available"? - the publisher may have good reasons you and me...