Daniel Markstedt

I was thinking along the same lines. But leaning towards a "beta" label with a notice about taking backups of data before using it. And perhaps also throw a Warning...

@NJRoadfan Pushed changes to libatalk to print a big warning message when loading the sqlite module, plus notes on the experimental state in the docs.

@NJRoadfan I'm aiming to merge this before the end of the week, unless you see any major blocker. My plan is to get this experimental backend into the wild with...

The relevant Debian dev package is `libev-libevent-dev`

We're blocked by a lack of a `event_base` data structure and `event_new()` callback support in libev.

I'm not seeing a crash anymore in `main`, but the interface auto detection is still failing.

This function is used in exactly one other place in the codebase, namely in the AFP speedtest. The same DSI code 12 is encountered here too. ``` [FPWriteFooter] get write...

Interestingly, the read/write header/footer protocol functions are not part of the AFP spec (as of 3.3). Seems to be interacting with the DSI header directly.

Ok, the problem is obviously that the offset value of 12 is assigned to dsi_code inside `AFPWriteHeader`. I can't see how this would be correct, not least because it's not...

I have postponed the dsi_code cleanup. It triggered cascading failures in the lantest module.