Sebastian Thiel

Sounds good, and good luck! If you think docker will also do, instead of using full-blown VMs, you could try [the respective travis-ci feature](https://docs.travis-ci.com/user/docker/).

Thank you for the very elaborate and incredibly well thought-out reply ! Even though my reply cannot possibly be as well formed as I clearly am on the user-side of...

As a side-note: It would get even more interesting if this system would be composable, which is probably where the actual difficulties start: ``` Rust pipe!(pipe!(command!(ls), command!(sort), command!(uniq)), command!(wc -l));...

@ArtemGr Certainly that would be fine as well. The benefit of Macros usually is the highly optimized/specialized syntax, but in the end it can surely be handled by functions and...

This is how it looks like in action. 

And something you might find interesting: now that I have undone that change to use symlinks as marker, it became evident that sled now is way faster when iterating crate...

In the same vein, when trying to re-export the data after deleting the tree with the small objects, it consumed huge amounts of memory… …which slowly rose. The CPU wasn't...

I thought maybe something is wrong with the export/import, so I wrote a 'manual' version. It produced exactly the same result and exhibited the similar memory consumption. ```Rust pub fn...

**Important**: Here is the database I use for all of the above: https://www.dropbox.com/s/fmka504t8ykner1/criner.db.zip?dl=0 It's just 1.1GB compressed, and should help tremendously when debugging. Here is the sled version that can...

I am happy to help :)! Also I rely on Sled as there seems to be no other option but SQLite, or none I can see, and thus far Sled...